Thomas Anthony Munro

The Chemistry of Salvia divinorum
Thomas Anthony Munro
Submitted in total fullment of the requirements of the degree of Doctor of Philosophy
April 2006
Department of Chemistry The University of Melbourne
Abstract
Salvia divinorum is a hallucinogenic sage used to treat illness by the Mazatec Indians of Mexico. Salvinorin A (1a), a neoclerodane diterpenoid isolated from the plant, is a potent, selective agonist at the opioid receptor (KOR), and is the rst non-nitrogenous opioid. The plant is used recreationally as a hallucinogen, but is unpopular due to its dysphoric eects. 1a has been prohibited in Australia under an invalid systematic name. An early report of psychoactive alkaloids in S. divinorum proved to be irreproducible. Similarly, tests in mice suggesting the presence of psychoactive compounds other than 1a were confounded and therefore unreliable.
O O O
O O
2
H
8
O O R2
OR1 H
O O
R1
H R2
O R1 R2 OH OAc H
OR3 R1 28a 28b 28c R2 R3
1a
1d 1e 1f
Ac H H
OH H H OH OH Me H OAc H
In this work, an improved isolation method for 1a was developed, using ltration through activated carbon to decolourise the crude extract. Six new diterpenoids were isolated: salvinorins DF (1d1f ) and divinatorins AC (28a28c). Five known terpenoids not previously reported from this species were also isolated. 3
4 The structureactivity relationships of 1a were evaluated via selective modications of each functional group. Useful synthetic methods are reviewed, including the rst thorough review of furanolactone hydrogenations. Testing of the derivatives at the KOR suggests that the methyl ester and furan ring of 1a are required for activity, but that the lactone and ketone functionalities are not. Other compounds from S. divinorum did not bind to the KOR, suggesting that 1a is the plants active principle. The structure of the 8-epimer of 1a, reported previously without supporting evidence, was rmly established. This epimerisation proved to be a general phenomenon among salvinorins and related furanolactones, occurring via enolisation of the lactone. The more complex mechanism proposed by Koreeda and co-workers was inconsistent with subsequent data. Under strongly basic conditions, autoxidation of 1a occurred to give the enedione 59 as the major product. A previously proposed structure was shown to be incorrect. Salvinorins and divinatorins were tested and found to be inactive against insects, bacteria, fungi, HIV, tumour cell lines and protein synthesis.
O
OH O
O O
59
Declaration
This is to certify that 1. the thesis comprises only my original work except where indicated in the preface, 2. due acknowledgement has been made in the text to all other material used, 3. the thesis is less than 100,000 words in length, exclusive of tables, maps, bibliographies and appendices.
Thomas Munro
Thomas Anthony Munro
Digitally signed by Thomas Anthony Munro DN: CN = Thomas Anthony Munro, C = US, O = McLean Hospital, OU = Mailman Research Center Reason: I am the author of this document Date: 2006.09.17 18:22:07 +10'00'
5
Preface
Some of the bioassays described in Chapter 4 were performed by others. Opioid receptor binding assays were performed in the laboratories of Bryan Roth, Case Western Reserve University (Cleveland, Ohio), by Glenn Goetchius, Beth Ann Toth, Feng Yan, and Timothy Vortherms. Protein synthesis inhibition assays were performed in the laboratories of Jerry Pelletier, McGill University (Montreal, Canada). HIV replication assays (NL4.3 and AD8 strains) were performed in the laboratories of Sharon Lewin, Monash University, by Ajantha Solomon. Other HIV assays were performed at Southern Research Institute (Frederick, Maryland), under the direction of Dr Stephen Turk, U. S. National Institute of Allergy and Infectious Diseases (NIAID). Tumour cell growth inhibition assays were performed by the U. S. National Cancer Institute (NCI). Other assays were performed collaboratively: Antibacterial and antifungal assays were performed in the laboratories of Professor Roy Robins-Browne, University of Melbourne, with Andrea Bigham. Insect antifeedant assays were performed using supplies and facilities provided by David Heckel and Charles Robin, University of Melbourne. 7
8 Some photographs were provided by others, as credited. Parts of this work have been published previously: Munro, T. A.; Rizzacasa, M. A. Salvinorins DF, New Neoclerodane Diterpenoids from Salvia divinorum, and an Improved Method for the Isolation of Salvinorin A. J. Nat. Prod. 2003, 66, 703705. http://dx. doi.org/10.1021/np0205699 Bigham, A. K.; Munro, T. A.; Rizzacasa, M. A.; Robins-Browne, R. M. Divinatorins AC, New Neoclerodane Diterpenoids from the Controlled Sage Salvia divinorum. J. Nat. Prod. 2003, 66, 12421244. http://dx. doi.org/10.1021/np030313i Munro, T. A.; Rizzacasa, M. A.; Roth, B. L.; Toth, B. A.; Yan, F. Studies toward the Pharmacophore of Salvinorin A, a Potent Opioid Receptor Agonist. J. Med. Chem. 2005, 48, 345348. http://dx.doi. org/10.1021/jm049438q Munro, T. A.; Goetchius, G. W.; Roth, B. L.; Vortherms, T. A.; Rizzacasa, M. A. Autoxidation of Salvinorin A under Basic Conditions. J. Org. Chem. 2005, 70, 10,05710,061. http://dx.doi.org/10.1021/jo051813e
Acknowledgments
Thanks to the Commonwealth Government for an Australian Postgraduate Award. Thanks Mark for betting tight resources on a risky project. Im glad it paid o. Thanks to the man with the coolest name in chemistry, Leander Jerome Julian Valds III, for generously sharing ideas and advice. Daniel Siebert for helpful advice on lots of stu, and for introducing me to Bryan Roth. Torsten for introducing me to Daniel; Carl Turney for introducing me to Torsten (and everything else); and Erik for introducing me to Carl. Six degrees of separation. Mike and Heike, Antoine and Lara for being so generous. Frances for changing my life, and for the worlds coolest lab coat and bestest present. TK for teaching me NMR, from setting the trash hole to running a NOESY. Les Gamel for all that masterful glassblowing. Carl Schiesser for the radical initiator that dare not speak its name. Sammy for letting me use windoze. Danny for letting me use Adobe Creative Suite. Ben for introducing me to Hoyes NMR papers. The man with the second-coolest name in chemistry, Carlos Rodrguez, for translating Daz. Vic Iwanov for letting me use the safe, and lling out those annoying manifests. Max Hem for the photographs, glorious as always. A man with another cool name, Slava Olcheski, for the Oaxaca shot. Scott Crawford for getting those stunning shots out of the SEM. Richard Westkaemper for providing the binding model data. Tom for proofreading I owe you one big fella. Same goes for Caroline. Cheese, Gromit! Thanks Dad for keeping life interesting. Mojave, Ilulisaat, Peshawar, terror australis, silver shark, helibagging. What can I say: my dads better than yours. Thanks mum 9
10 for the sacrices you made and the love and hard slog you put into raising kids and working full time. And for helping with the move so I could keep writing this till ve days before leaving for my postdoc!
Contents
Abstract Declaration Preface Acknowledgments List of Figures List of Schemes List of Tables Acronyms 1 Introduction. 1.1 1.2 1.3
3 5 7 9 17 23 25 27 31
Botany. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Ethnopharmacology. . . . . . . . . . . . . . . . . . . . . . . . . 33 Chemistry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 1.3.1 1.3.2 Terpenoids. . . . . . . . . . . . . . . . . . . . . . . . . . 36 Alleged alkaloids. . . . . . . . . . . . . . . . . . . . . . . 43
1.4
Pharmacology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 11
12 1.4.1 1.4.2 1.4.3 1.4.4
CONTENTS Animal Testing. . . . . . . . . . . . . . . . . . . . . . . . 47 Human Testing. . . . . . . . . . . . . . . . . . . . . . . . 51 In vitro Testing. . . . . . . . . . . . . . . . . . . . . . . . 53 Mechanism: Opioids. . . . . . . . . . . . . . . . . . . . 54
1.5 Toxicology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 1.6 Social impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 1.6.1 1.6.2 Recreational Use. . . . . . . . . . . . . . . . . . . . . . . 62 Legal Status. . . . . . . . . . . . . . . . . . . . . . . . . 63
1.7 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2 Isolation.
67
2.1 Isolation Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . 67 2.1.1 2.1.2 2.1.3 2.1.4 Extraction Conditions. . . . . . . . . . . . . . . . . . . . 67 Problems Caused by Pigments. . . . . . . . . . . . . . . 69 Use of Activated Carbon. . . . . . . . . . . . . . . . . . . 72 Separation of Terpenoids. . . . . . . . . . . . . . . . . . 82
2.2 Structure Elucidation. . . . . . . . . . . . . . . . . . . . . . . . 90 2.2.1 2.2.2 2.2.3 2.2.4 2.2.5 2.2.6 2.2.7 Revised NMR Assignments for Salvinorin A (1a). . . . . 90 Revised NMR Assignments for Salvinorin C (1c). . . . . 92 Other Known Diterpenoids. . . . . . . . . . . . . . . . . 92 Known Triterpenoids. . . . . . . . . . . . . . . . . . . . . 96 Salvinorins D-F (1d-1f). . . . . . . . . . . . . . . . . . . 98 Divinatorins A-C (28a-28c). . . . . . . . . . . . . . . . . 110 Subsequent isolations. . . . . . . . . . . . . . . . . . . . 118
CONTENTS 3 Synthesis. 3.1 3.2
13 119
Known derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . 119 Epimerisation at C-8 under Basic Conditions. . . . . . . . . . . 120 3.2.1 3.2.2 3.2.3 3.2.4 3.2.5 3.2.6 3.2.7 Previous Reports. . . . . . . . . . . . . . . . . . . . . . . 120 8-epi-Salvinorins A and B (37a and 37b). . . . . . . . . . 121 Control of Epimerisation and Separation of Epimers. . . 122 8-epi-Salvinorin C (37c) and Related Compounds. . . . . 124 Chromatographic Identication of Epimers. . . . . . . . 125
Mechanism. . . . . . . . . . . . . . . . . . . . . . . . . . 126 Attempted Deacetylation under Acidic Conditions. . . . 128
3.3
Simple Derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . 128 3.3.1 3.3.2 3.3.3 3.3.4 3.3.5 Esters (46 and 47). . . . . . . . . . . . . . . . . . . . . . 128 Attempted Benzyl Ether Formation (48). . . . . . . . . . 129 17-Deoxy Compounds (49 and 50). . . . . . . . . . . . . 130 Tetrahydrosalvinorin A (51). . . . . . . . . . . . . . . . . 131 (+)-Hardwickiic Acid (ent-29a). . . . . . . . . . . . . . . 134
3.4
Modication of the Methyl Ester. . . . . . . . . . . . . . . . . . 134 3.4.1 3.4.2 3.4.3 3.4.4 Relevant Results from Previous Work. . . . . . . . . . . 134 Treatment of Salvinorin A with KOH in MeOH. . . . . . 135 O-Demethylsalvinorin A (67a). . . . . . . . . . . . . . . 144 O-Demethyl-18-deoxysalvinorin A (77). . . . . . . . . . . 150
3.5
Modication of the Ketone. . . . . . . . . . . . . . . . . . . . . 153 3.5.1 3.5.2 3.5.3 Attempted Methylenation. . . . . . . . . . . . . . . . . . 153 Attempted Direct Deoxygenation. . . . . . . . . . . . . . 154 Indirect Deoxygenation. . . . . . . . . . . . . . . . . . . 155
14 4 Bioassays.
CONTENTS 161
4.1 Insect Antifeedant Activity. . . . . . . . . . . . . . . . . . . . . 161 4.2 Eukaryotic Protein Synthesis Inhibition. . . . . . . . . . . . . . 163 4.3 Antimicrobial Activity. . . . . . . . . . . . . . . . . . . . . . . . 164 4.3.1 4.3.2 Bacteria and Fungi. . . . . . . . . . . . . . . . . . . . . . 164 HIV-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.4 NCI Anticancer Screen. . . . . . . . . . . . . . . . . . . . . . . . 168 4.5 Activity at the Opioid Receptor. . . . . . . . . . . . . . . . . 170 4.5.1 4.5.2 4.5.3 4.5.4 4.5.5 4.5.6 4.5.7 4.5.8 Other Salvinorins and Divinatorins. . . . . . . . . . . . . 171 Modication of the Ketone. . . . . . . . . . . . . . . . . 173 Modication of the Acetoxy Group. . . . . . . . . . . . . 174 Modication of the Methyl Ester. . . . . . . . . . . . . . 175 Modication of the Lactone. . . . . . . . . . . . . . . . . 176 Modication of the Furan Ring. . . . . . . . . . . . . . . 176 Incorporation into a Revised Binding Model. . . . . . . . 177 Subsequent Results. . . . . . . . . . . . . . . . . . . . . . 179 181
5 Experimental.
5.1 General Conditions . . . . . . . . . . . . . . . . . . . . . . . . . 181 5.1.1 5.1.2 5.1.3 5.1.4 Instruments and Procedures. . . . . . . . . . . . . . . . . 181 Reagents. . . . . . . . . . . . . . . . . . . . . . . . . . . 183 Plant Materials. . . . . . . . . . . . . . . . . . . . . . . 183
Assays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5.2 Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 5.2.1 Extraction of Commercial S. divinorum. . . . . . . . . . 185
CONTENTS 5.2.2 5.2.3 5.2.4 5.2.5 5.2.6 5.2.7 5.2.8 5.2.9
15 Extraction of Australian S. divinorum. . . . . . . . . . . 187 Salvinorin A (1a). . . . . . . . . . . . . . . . . . . . . . 188
Salvinorin B (1b). . . . . . . . . . . . . . . . . . . . . . . 189 Salvinorin C (1c). . . . . . . . . . . . . . . . . . . . . . 189
Salvinorin D (1d). . . . . . . . . . . . . . . . . . . . . . . 190 Salvinorin E (1e). . . . . . . . . . . . . . . . . . . . . . . 191 Salvinorin F (1f). . . . . . . . . . . . . . . . . . . . . . . 192 Divinatorin A (28a). . . . . . . . . . . . . . . . . . . . . 194
5.2.10 Divinatorin B (28b). . . . . . . . . . . . . . . . . . . . . 195 5.2.11 Divinatorin C (28c). . . . . . . . . . . . . . . . . . . . . 196 5.2.12 ()-Hardwickiic Acid (29a) and methyl ester 29b. . . . . 197 5.2.13 Oleanolic Acid (31). . . . . . . . . . . . . . . . . . . . . 198
5.2.14 Presqualene Alcohol (32). . . . . . . . . . . . . . . . . . 198 5.2.15 Peplusol (33). . . . . . . . . . . . . . . . . . . . . . . . . 199 5.3 Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 5.3.1 5.3.2 5.3.3 5.3.4 5.3.5 5.3.6 5.3.7 5.3.8 5.3.9 Salvinorin C (1c) via acetylation of salvinorin D (1d). . . 199 Salvinorins D (1d) and E (1e) via acetylation of 1h. . . . 200 Salvinorins C (1c) and E (1e) via acetylation of 1h. . . . 200 Dideacetylsalvinorin C (1h) from 1c. . . . . . . . . . . . 201 (+)-Hardwickiic acid (ent-29a). . . . . . . . . . . . . . . 202 Salvinorin A lactol (35). . . . . . . . . . . . . . . . . . . 203 (4R)-3,4-Dihydrosalvinorin C (36c). . . . . . . . . . . . . 205 (4R)-3,4-Dihydrosalvinorin E (36e). . . . . . . . . . . . . 206 (4R)-Dideacetyl-3,4-dihydrosalvinorin C (36h). . . . . . . 207
5.3.10 8-epi-Salvinorin A (37a). . . . . . . . . . . . . . . . . . . 208
16
CONTENTS 5.3.11 8-epi-Salvinorin B (37b). . . . . . . . . . . . . . . . . . . 210 5.3.12 8-epi-Salvinorin C (37c). . . . . . . . . . . . . . . . . . . 211 5.3.13 8-epi-Salvinorin D (37d). . . . . . . . . . . . . . . . . . . 212 5.3.14 8-epi-Salvinorin E (37e). . . . . . . . . . . . . . . . . . . 214 5.3.15 8-epi-Dideacetylsalvinorin C (37h). . . . . . . . . . . . . 215 5.3.16 Salvinorin B formate (46). . . . . . . . . . . . . . . . . . 216 5.3.17 Dideacetylsalvinorin C 2-O-(4-bromobenzoate) (47). . . . 218 5.3.18 17-Deoxysalvinorin A (49). . . . . . . . . . . . . . . . . . 219 5.3.19 8,17-Didehydro-17-deoxysalvinorin A (50). . . . . . . . . 220 5.3.20 13,14,15,16-Tetrahydrosalvinorin A (51). . . . . . . . . . 222 5.3.21 Autoxidation of 1a in KOH/MeOH. . . . . . . . . . . . . 223 5.3.22 NaBH4 reduction of 59. . . . . . . . . . . . . . . . . . . . 228 5.3.23 O-Demethyl-18-deoxysalvinorin A (77). . . . . . . . . . 229
5.3.24 1-Deoxysalvinorin A (81a). . . . . . . . . . . . . . . . . . 233 Bibliography 239
List of Figures
1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9
Flowering specimen of S. divinorum.1 . . . . . . . . . . . . . . . 31 Location of Oaxaca and the Sierra Mazateca within Mexico. . . 32 Young Salvia divinorum plant, Oaxaca.8 . . . . . . . . . . . . . 33 S. divinorum ower in bud (stereoview).1 . . . . . . . . . . . . . 33 Ortega et als X-ray structure of 1a (stereoview). . . . . . . . . 37 Salvinorins B (1b) and C (1c). . . . . . . . . . . . . . . . . . . 38 Known terpenoids. . . . . . . . . . . . . . . . . . . . . . . . . . 38 Clerodin (5) and standard clerodane numbering. . . . . . . . . . 39 Biosynthetic precursors of diterpenoids. . . . . . . . . . . . . . . 40
1.10 Peltate glandular trichome (SEM stereoview).39 . . . . . . . . . 42 1.11 Underside of S. divinorum leaves (SEM stereoview).39 . . . . . . 43 1.12 Activity of pure 1a versus mixtures in mice (open eld assay). . 48 1.13 Morphine (11) and derivatives. . . . . . . . . . . . . . . . . . . 54 1.14 Arylacetamide opioids. . . . . . . . . . . . . . . . . . . . . . . 56 1.15 Structurally diverse opioids. . . . . . . . . . . . . . . . . . . . 59 1.16 Relevant IUPAC fused-ring numbering schemes. . . . . . . . . . 64 2.1 2.2 Sieberts TLC analysis of crude CHCl3 extracts.40 . . . . . . . . 69 Representative major plant pigments. . . . . . . . . . . . . . . . 71 17
18
LIST OF FIGURES 2.3 Ortho- and non-ortho-substituted PCBs. . . . . . . . . . . . . . 76 2.4 Apparatus for Filtration through Activated Carbon. . . . . . . . 78 2.5 Flavonoids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 2.6 SEM image of salvinorin A crystals (blade morphology).39 . . . 82 2.7 Other salvinorin A crystal morphologies (stereoview).39 . . . . . 84 2.8 Terpenoids isolated from S. divinorum. . . . . . . . . . . . . . . 85 2.9 TLC data of isolated compounds. . . . . . . . . . . . . . . . . . 86 2.10 Isolation of terpenoids from commercial S. divinorum. . . . . . . 87 2.11 Isolation of terpenoids from Australian S. divinorum. . . . . . . 88 2.12 1 H NMR spectrum of 1a (800 MHz, CDCl3 ). . . . . . . . . . . . 91 2.13 HSQC spectrum of 1a (800 MHz, CDCl3 ). . . . . . . . . . . . . 91 2.14 Revised NMR assignments for 1a (stereoview). . . . . . . . . . . 92 2.15 1 H NMR spectrum of 1c (400 MHz, CDCl3 ). . . . . . . . . . . . 93 2.16 Revised NMR assignments for 1c (stereoview). . . . . . . . . . . 93 2.17 HMQC and HMBC spectra of 1c (400 MHz, CDCl3 ). . . . . . . 94 2.18 Single-crystal X-ray structure of 29a (stereoview). . . . . . . . . 95 2.19 (E)-Phytol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 2.20 Oleanolic acid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 2.21 Presqualene alcohol and peplusol. . . . . . . . . . . . . . . . . . 97 2.22 1 H NMR spectrum of 1d (400 MHz, CDCl3 ). . . . . . . . . . . . 99 2.23 NMR assignments and key 2D correlations for 1d (stereoview). . 100 2.24 HMQC spectrum of 1d (400 MHz, CDCl3 ). . . . . . . . . . . . 100 2.25 HMBC spectrum of 1d. . . . . . . . . . . . . . . . . . . . . . . 101 2.26 1 H NMR spectrum of 1e (400 MHz, CDCl3 ). . . . . . . . . . . . 102 2.27 NMR assignments and key 2D correlations for 1e (stereoview). . 102
LIST OF FIGURES
19
2.28 HMQC spectrum of 1e. . . . . . . . . . . . . . . . . . . . . . . 103 2.29 HMBC spectrum of 1e. . . . . . . . . . . . . . . . . . . . . . . . 103 2.30 1 H NMR spectrum of 1h (400 MHz, CDCl3 ). . . . . . . . . . . . 104 2.31 NMR assignments and key 2D correlations for 1h (stereoview). . 105 2.32 HMQC and HMBC spectra of 1h. . . . . . . . . . . . . . . . . . 105 2.33 NOESY spectrum of 1h. . . . . . . . . . . . . . . . . . . . . . . 105 2.34 1 H NMR spectrum of 1f (400 MHz, CDCl3 ). . . . . . . . . . . . 108 2.35 NMR assignments and key 2D correlations for 1f (stereoview). . 109 2.36 HMQC spectrum of 1f. . . . . . . . . . . . . . . . . . . . . . . . 109 2.37 HMBC spectrum of 1f. . . . . . . . . . . . . . . . . . . . . . . . 109 2.38 Divinatorins AC (28a-28c) and hardwickiic acid (29a). . . . . 110 2.39 1 H NMR spectrum of 28a (400 MHz, CDCl3 ). . . . . . . . . . . 110 2.40 NMR assignments and key 2D correlations for 28a (stereoview). 111 2.41 HMBC spectrum of 28a. . . . . . . . . . . . . . . . . . . . . . . 111 2.42 NOESY spectrum of 28a. . . . . . . . . . . . . . . . . . . . . . 112 2.43 1 H NMR spectrum of 28b (400 MHz, CDCl3 ). . . . . . . . . . . 113 2.44 NMR assignments and key 2D correlations for 28b (stereoview). 113 2.45 HMBC spectrum of 28b. . . . . . . . . . . . . . . . . . . . . . . 114 2.46 NOESY spectrum of 28b. . . . . . . . . . . . . . . . . . . . . . 114 2.47 1 H NMR spectrum of 28c (400 MHz, CDCl3 ). . . . . . . . . . . 115 2.48 NMR assignments and key 2D correlations for 28c (stereoview). 115 2.49 HMBC spectrum of 28c. . . . . . . . . . . . . . . . . . . . . . . 116 2.50 NOESY spectrum of 28c. . . . . . . . . . . . . . . . . . . . . . 116 2.51 Subsequently isolated compounds. . . . . . . . . . . . . . . . . . 118 3.1 Key NOESY correlations for 37a (stereoview). . . . . . . . . . . 123
20
LIST OF FIGURES 3.2 TLC comparison of epimers using vanillin/H2 SO4 . . . . . . . . . 126 3.3 Ester and ether derivatives. . . . . . . . . . . . . . . . . . . . . 129 3.4 Furanolactones 53, 54 and 55. . . . . . . . . . . . . . . . . . . . 131 3.5 (+)-Hardwickiic acid (ent-29a). . . . . . . . . . . . . . . . . . . 134 3.6 Key HMBC correlations of 59 and 60a. . . . . . . . . . . . . . 137 3.7 UV/Visible spectra of 1a, 1c and 59 in MeCN. . . . . . . . . . 138 3.8 Diol 36h and proposed autoxidation product 62. . . . . . . . . 139 3.9 O-Demethyl salvinorins A and B. . . . . . . . . . . . . . . . . . 145 3.10 Useful non-hydrogen bond donor solvents. . . . . . . . . . . . . 145 3.11 Methylenated target compound 79. . . . . . . . . . . . . . . . . 154 3.12 RP-LCMS traces of early fractions versus 81b/82b. . . . . . . . 159 3.13 1-Deoxysalvinorin A (81a). . . . . . . . . . . . . . . . . . . . . 160 4.1 Luciferase assay results for salvinorins and divinatorins (50 M). 163 4.2 (-)-Hardwickiic acid and divinatorins A-C. . . . . . . . . . . . . 164 4.3 Disk and microdilution assays for ent-29a and crude extract.428 165 4.4 HIV-1 replication assays (NL43 and AD8 strains). . . . . . . . . 166 4.5 HIV-1 replication assays (ROJO isolate). . . . . . . . . . . . . . 167 4.6 NCI 60 cell line results for salvinorins and divinatorins . . . . . 169 4.7 CNS cell line results for divinatorin B and salvinorin B. . . . . . 170 4.8 KOR binding anity and potency of salvinorin A. . . . . . . . . 171 4.9 KOR binding anities of salvinorins and divinatorins. . . . . . . 172 4.10 KOR activity after ketone modications. . . . . . . . . . . . . . 174 4.11 KOR activity after acetoxy group modications. . . . . . . . . . 174 4.12 KOR activity after methyl ester modications. . . . . . . . . . . 175 4.13 KOR activity after lactone modications. . . . . . . . . . . . . . 176
LIST OF FIGURES
21
4.14 KOR activity after furan modications. . . . . . . . . . . . . . . 176 4.15 Westkaempers original binding model. . . . . . . . . . . . . . . 177 4.16 Westkaempers revised binding model (stereoview). . . . . . . . 178 4.17 KOR binding anities and potencies of recent derivatives. . . . 179
22
LIST OF FIGURES
List of Schemes
1.1 1.2 2.1 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 Cyclization of 7 to the trans-neoclerodane skeleton(stereoview). 41
Cyclization of 7 to the cis-neoclerodane skeleton. . . . . . . . . 41 Interconversion of 1c-1h. . . . . . . . . . . . . . . . . . . . . . . 104 Preparation of known derivatives. . . . . . . . . . . . . . . . . . 119 Formation of 8-epi-salvinorins A and B. . . . . . . . . . . . . . 121
Synthesis of 37c-37h. . . . . . . . . . . . . . . . . . . . . . . . 125 Koreeda et als proposed mechanism of epimerisation. . . . . . . 127 Epimerisation of related natural products with base. . . . . . . . 127 Browns deuteration of 1b. . . . . . . . . . . . . . . . . . . . . . 128 Deoxygenation of lactol 35. . . . . . . . . . . . . . . . . . . . . 130 Hydrogenation of 1a and other furanolactones. . . . . . . . . . . 131 LiAlH4 and Li/NH3 reductions of 1a. . . . . . . . . . . . . . . . 135
3.10 Autoxidation of 1a. . . . . . . . . . . . . . . . . . . . . . . . . . 136 3.11 Proposed mechanism of the autoxidation. . . . . . . . . . . . . . 141 3.12 Unexpected oxidation product 65. . . . . . . . . . . . . . . . . . 141 3.13 Attempted reductions of 59. . . . . . . . . . . . . . . . . . . . . 143 3.14 BAl 2 and BAc 2 ester cleavage mechanisms. . . . . . . . . . . . . 144 3.15 Formation of mixed anhydride 68. . . . . . . . . . . . . . . . . . 147 3.16 Some previously reported BH3 THF reductions. . . . . . . . . . 151 3.17 Borane reduction of 67a. . . . . . . . . . . . . . . . . . . . . . . 152 3.18 Ketone deoxygenation via a tosylhydrazone. . . . . . . . . . . . 154 3.19 Formation of cyclic thionocarbonate 80. . . . . . . . . . . . . . 156 3.20 Radical deoxygenation of 80. . . . . . . . . . . . . . . . . . . . 158
23
24
LIST OF SCHEMES
List of Tables
Relative potency of 1a at cloned opioid receptors. . . . . . . . 60 Alternate systematic names for 1a. . . . . . . . . . . . . . . . . 64 Eects of solvent and temperature on recovery of 1a. . . . . . . 68 Data and sources used to identify known compounds. . . . . . . 94 Coupling constants (Hz) at H-8 for 1a, 1b and 8-epimers. . . . . 122 Some previously reported furanolactone hydrogenations. . . . . 133 Summary of results - nucleophilic cleavage of 1a methyl ester. . 148 Unsuccessful treatment of 1a with excess tosylhydrazide. . . . . 155 Antifeedant test results. . . . . . . . . . . . . . . . . . . . . . . 162 KOR radioligand and functional assay results. . . . . . . . . . . 172 Yields and TLC data (hRf ) of isolated compounds. . . . . . . . 187
1.1 1.2 2.1 2.2 3.1 3.2 3.3 3.4 4.1 4.2 5.1
25
26
LIST OF TABLES
Acronyms
AIBN Azobisisobutyronitrile [2,2-Azobis(2-methylpropionitrile)] AZT Azidovudine Borsm Based on recovered starting material CNS Central nervous system COSY Correlation spectroscopy 2D NMR (24 JHH ) DEPT Distortionless enhancement by polarization transfer NMR (13 C multiplicity) DMAP 4-Dimethylaminopyridine DMF Dimethylformamide DMPU 1,3-dimethyltetrahydropyrimidin-2-one EC50 Concentration causing 50% of maximal ecacy GC/MS Gas chromatography/mass spectrometry HMBC Heteronuclear multiple bond correlation 2D NMR (23 JCH ) HMQC Heteronuclear multiple quantum coherence 2D NMR (1 JCH ) HMPA Hexamethylphosphoric triamide (a.k.a. HMPT) HPLC High performance liquid chromatography HRESIMS High resolution electrospray ionisation mass spectrometry 27
28 hRf = Rf 100
LIST OF TABLES
HSQC Heteronuclear single quantum coherence 2D NMR (1 JCH ) IC50 Concentration causing 50% inhibition InChI IUPAC International Chemical Identier K i Receptor binding anity constant KOR (kappa) opioid receptor LC/MS Liquid chromatography/mass spectrometry LD50 Dose lethal to 50% of test animals MIC Minimum inhibitory concentration NDPSC National Drugs and Poisons Schedule Committee NMR Nuclear magnetic resonance nOe Nuclear Overhauser eect (throughspace signal enhancement) NOESY Nuclear Overhauser enhancement spectroscopy (throughspace 2D NMR) PCB Polychlorinated biphenyl PCR Polymerase chain reaction Rf Retardation factor origin] ROESY Rotating frame Overhauser enhancement spectroscopy (through space 2D NMR) RP Reverse phase (nonpolar stationary phase, polar mobile phase) SEM Scanning electron microscopy / standard error of the mean THF Tetrahydrofuran
a f
[distances of analyte (a) and solvent front (f ) from
LIST OF TABLES TLC Thin layer chromatography
29
30
LIST OF TABLES
Chapter 1 Introduction.
Figure 1.1: Flowering specimen of S. divinorum.1
1.1
Botany.
The genus Salvia, containing over 900 species internationally, is one of the largest in the family Lamiaceae (or Labiatae). The largest subgenus is the 31
32
N
CHAPTER 1. INTRODUCTION.
Mexico
Sierra Mazateca
Oaxaca
100 200 300 km
Figure 1.2: Location of Oaxaca and the Sierra Mazateca within Mexico. South American Calosphace (or Jungia),2 of which over 300 species are found in Mexico.3 Among these is Salvia divinorum Epling & Jtiva.4, 5 Discovered in the highlands of northern Oaxaca (Figure 1.2), the plant is a perennial shrub growing to about 1.5 m, preferring moist shady sites at high elevations.6 While the owers are distinctive (Figure 1.1), the plants appearance is nondescript during vegetative growth (Figure 1.3), apart from the unusual square stem. In the original botanical description, based on a dried specimen and witness reports, the ower (corolla) was erroneously described as being blue.4 This error was incorporated into several colour botanical illustrations.7 The error originated in observations by Robert Gordon Wasson and Albert Hofmann, who obtained the type specimen but were not botanists.6 In fact, the corolla is white, emerging from a violet calyx (Figure 1.4). This error has been corrected in an amended botanical description,6 which also has the advantage of being in English rather than Latin. Several others had investigated the species and collected specimens,9 but Wasson and Hofmann were the rst to obtain a owering specimen, which is essential for species identication. It should be pointed out that, although Wasson and Hofmann were credited as the collectors of the type specimen,4 it was in fact given to them by Natividad Rosa, a Mazatec healing woman.10, 11 Wasson and Hofmann have also been incorrectly credited with bringing live plants to
1.2. ETHNOPHARMACOLOGY.
33
Figure 1.3: Young Salvia divinorum plant, Oaxaca.8 the U.S.A., which was actually done by botanist Sterling Bunnell.12 Most of the S. divinorum plants in cultivation internationally are clones of this misnamed Wasson and Hofmann strain.12
Figure 1.4: S. divinorum ower in bud (stereoview).1 There has been no report of growing the plant from seed: propagation is exclusively vegetative.6, 13 Moreover, there are no published rst-hand reports of wild populations;7 despite thorough searches in the region by several workers, all known stands appear to have been planted.6, 14
1.2
Ethnopharmacology.
The leaves of S. divinorum are used as a traditional medicine by the Mazatec Indians of the Oaxaca region. The disorders treated include gastrointestinal
34
problems, headaches, rheumatism, anaemia and swelling of the stomach.15, 11, 13 An infusion is prepared by crushing or rubbing fresh leaves in water; a frothy infusion is considered a sign of potency. The leaf residue may also be applied to the patients forehead as a poultice afterwards.11 In addition, the healers (curanderos) themselves drink the infusion to induce visions.5 They believe these visions allow them to divine the cause of the illness. Hence the name S. divinorum, meaning sage of the seers.4 For these eects, larger doses of the infusion are used, or the leaves themselves are slowly chewed and eaten. Eating the leaves is very dicult, due to their sickeningly bitter taste, and often induces vomiting. First-hand accounts of the eects vary from barely perceptible (an overlay of dancing colours)5 to powerful experiences, in which awareness of reality is lost and bizarre visions are perceived as real: I saw a pulsating purplish light that changed to an insect-like shape, perhaps a bee or a moth, and then into a pulsating sea anemone. It expanded into a desert full of prickly pear cacti, and remained so for several minutes. During the rst session and throughout the night, my visions had all appeared to be something like a cross between a silent moving picture and a cartoon. I felt myself to be an observer of these mute visions, rather than being an actual part of them. Suddenly, however, I was in a broad meadow with brightly colored owers. I had just crossed a stream by way of a small wooden bridge. Next to me was something that seemed to be the skeleton of a giant model airplane made of rainbow colored inner tubing. The sky was bright blue and I could see ... woods in the distance. I found myself talking to a man in a shining white robe who was either shaking my hand, or else holding on to it. It was an amazing hallucination, as I truly believed I was in the meadow.14 This is in marked contrast to the pseudohallucinations induced by compounds
1.2. ETHNOPHARMACOLOGY.
35
such as LSD. The visions were accompanied by impaired physical coordination and slurred speech.16 Notwithstanding the intensity of some such experiences, the Mazatecs consider the plant the weakest of their visionary substances.13 They refer to it as ka Mara pastora (the leaves of Mary shepherdess), apparently from an obscure Catholic term for the Virgin Mary.17 Ott has argued7 that the lack of an indigenous name indicates that the plant is not indigenous to the region that, like sheep and Catholicism themselves, it came to the Mazatecs after European settlement. He points out that the introduced European hallucinogenic mushroom Psilocybe cubensis also has no indigenous name, unlike its indigenous equivalents, and is regarded as an inferior substitute. S. divinorum is regarded in the same way. Furthermore, the Mazatecs consider the plant as part of the same family as two introduced Coleus species. As further evidence, he points out that the Mazatec belief that drying the leaves destroys their potency has been disproven. Similarly, drinking an infusion has been shown to be very inecient; much stronger eects are produced by retaining the infusion,18 or alternatively a quid of chewed leaves,7 in the mouth. Ott argues that each of the above points suggests a lack of cultural tradition, and concludes that the plant was adopted relatively recently from another tribe. Ott has also endorsed Wassons suggestion19 that S. divinorum may represent the divinatory plant pipiltzintzintli cultivated by the Aztecs.7 The Aztecs prepared an infusion from the plant, including the leaves, and also applied it as a poultice; no other plant was used in both these ways. The only Mexican plant whose leaves are presently so used is S. divinorum. Accounts of pipiltzintzintli make no mention of seeds, unlike other Aztec visionary plants; and S. divinorum is eectively seedless, unlike other contemporary Mexican visionary plants. Here, then, we have a plant without an indigenous name, and an indigenous name referring to an unknown plant. One has no present, the other has no past, and they have several very distinctive characteristics in common. Others reject this hypothesis. Daz has pointed out that male and female gen-
36
ders of pipiltzintzintli were reported, that the plant was dried before use, and that the infusion was prepared from the entire plant, none of which is true of S. divinorum.20 Ott counters that the Aztecs may have been using gender metaphorically, as the Mazatecs do today. Drying does not aect the plants potency, and the addition of other plant parts, while superuous, would not aect the infusion. Moreover, Dazs preferred candidate, Cannabis sativa, cannot be correct, since it was introduced after European settlement.7 Valds also rejects Wassons hypothesis, preferring Beltrns proposal that pipiltzintzintli was a synonym for the morning glory, ololiuhqui.13 Ott has shown, however, that these were explicitly described as dierent plants, and that S. divinorum is thus ... the only Mexican entheogenic plant which ts the criteria for pipiltzintzintli, and ... remains our best guess for the identity of the lost Aztec entheogen.7 Both parties in this surprisingly bitter dispute concede that the evidence is inconclusive, and both suspect that S. divinorum has been used by tribes other than the Mazatecs. Valds notes early anecdotal reports of use by the Cuicatec and Otom tribes, whose lands adjoin the Sierra Mazateca.21, 9
1.3
1.3.1
1.3.1.1
Chemistry.
Terpenoids.
Isolation.
The rst compound isolated from S. divinorum was 1a, a clerodane diterpenoid discovered by Ortega et al in 1982 and named salvinorin.22 The terminology of these terpenoids will be discussed in the next section. Extraction of the dried leaves in reuxing CHCl3 , chromatography on Tonsil activated clay and crystallisation from MeOH gave 1a in unstated yield. The structure
1.3. CHEMISTRY.
O
37
O O
O O
1a
Figure 1.5: Ortega et als X-ray structure of 1a (stereoview). was elucidated spectroscopically, and conrmed by X-ray crystallography.22 Absolute stereochemistry was tentatively assigned by comparing the circular dichroism spectrum with known compounds. Subsequently, Valds isolated the same compound. Unaware of Ortegas work, he named the compound divinorin A in his thesis.15 He also isolated the deacetyl analogue 1b, which he named divinorin B. When the work was published, this oversight was corrected, and the compounds were named salvinorins A and B.23 Valdss isolation procedure was more complex. The Et2 O extract was dissolved in MeOH and washed with hexanes. Repeated chromatography on silica gel and repeated recrystallisation from EtOH gave 1a in 1.8 g/kg yield based on dry weight. The yield of 1b was much lower (74 mg/kg). Spectroscopic structure elucidation of 1a was again conrmed by X-ray crystallography. The t of the model was in excellent agreement with the diraction data (R-factor = 8.7 %), though not quite as good as Ortega et als (5.2 %).
38
O
O O O HO H H O O O O O H H O O
1b
1c
Figure 1.6: Salvinorins B (1b) and C (1c). Ortega et als model also has the advantage of including the positions of hydrogen atoms. Unfortunately, the absolute stereochemistry of Valds et als model is incorrect, although it was correctly assigned in the paper, again on the basis of circular dichroism. This stereochemistry has since been denitively conrmed by the use of exciton chirality circular dichroism,24 and more recently X-ray crystallography,25 on suitable derivatives. This absolute stereochemistry is common to all clerodanes isolated from the Lamiaceae.26 Valds et al subsequently isolated salvinorin C (1c).27 Repeated chromatography of the Et2 O extract followed by HPLC gave 1c in 78 mg/kg yield. Spectroscopic structure elucidation was supported by partial synthesis of analogues.
HO 2
H H HO 3 4 H
Figure 1.7: Known terpenoids. In addition to these new diterpenoids, several known terpenoids were detected. The monoterpenoid loliolide (2) was isolated and fully characterised by Valds, in the same manner as 1c, in 4.4 mg/kg yield.28 GC/MS analysis by Giroud
1.3. CHEMISTRY.
39
et al also detected compounds whose MS data were consistent with the nortriterpenoid stigmasterol (3) and the diterpenoid neophytadiene (4).29
1.3.1.2
Terminology.
O H H H O H
2 3 1 14 13 12 10 11 9 8 7 17 15 16
5 20 4 19 18 6
OAc OAc
Figure 1.8: Clerodin (5) and standard clerodane numbering. Clerodane diterpenoids30 are named after clerodin (5). Conclusively establishing the structure of this compound was a tortuous process, which has been lucidly reviewed by Rodriguez-Hahn et al.26 The absolute stereochemistry was initially proposed as ent-5. Compounds with stereochemistry matching 5 were therefore termed ent-clerodane for many years. Extensive crystallographic and degradation studies ultimately proved the true structure of clerodin to be 5. To avoid ambiguity, the term neoclerodane was therefore coined for compounds matching 5.31 Compounds previously known as clerodanes would be termed ent-neoclerodane. For further detail, consult Rodriguez-Hahn et al.26 Clerodanes are further subdivided into trans- and cis- varieties, according to the conguration of C-5 relative to C-10 in the standard32 numbering scheme (Figure 1.8). All clerodane diterpenoids isolated from S. divinorum to date are trans-neoclerodane.
1.3.1.3
Biosynthesis.
Terpenoids are synthesised from C5 isoprene units, derived from isopentenyl diphosphate (6).33 Assembly of four isoprene units gives geranylgeranyl diphosphate (7), the nal common biosynthetic intermediate of all diter-
40
O O 6 P O
P OH O OH OH
= OPP
OPP 7
Figure 1.9: Biosynthetic precursors of diterpenoids.
penoids.34 Formation of clerodanes begins with protonation at C-14 of 7 (Scheme 1.1). This initiates a cationic cascade leading to the labdane intermediate 8. A sequence of 1,2-hydride and methyl shifts then gives the trans-clerodane skeleton 9.
It is unknown whether these 1,2- shifts are concerted. In the formation of cisclerodanes, a discrete halimane intermediate (10) is formed (Scheme 1.2).34 These pathways have been substantiated by feeding plants isotopically labelled mevalonic acid, a precursor to 6. Notably, labelling of the terminal carbon of 7, which becomes C-18 in labdane intermediate 8, results in labelling at C-18 in trans-clerodanes,35 but C-19 in cis-clerodanes,36 as expected (Schemes 1.1 and 1.2). Tritium labelling also conrms the hydride shift from H-5 to H-10.36
Mevalonic acid was used in these studies because it was long assumed to be the sole precursor to 6. Recently, revolutionary work has overturned this assumption.37, 38 An alternate pathway, via 1-deoxyxylulose 5-phosphate, has been established. Indeed, the mevalonic acid pathway makes a negligible contribution to diterpenoid biosynthesis. Nonetheless, as reected in the above results, the pathways are not mutually exclusive. Some crosstalk occurs, which is increased by the feeding of precursors.38 Nonetheless, incorporation of mevalonate into diterpenoids is very low: below 0.01% in some cases.36
1.3. CHEMISTRY.
41
OPP
OPP
14
14
15
15
H+ 7
CH2OPP
H+ 7
CH2OPP
+ 8
18 19 18 19
+ 8
CH2OPP
CH2OPP
+
18 19
+ 9
18 19
Scheme 1.1: Cyclization of 7 to the trans-neoclerodane skeleton(stereoview).
OPP
OPP
OPP
OPP
H
10 5
H
8
10
H+ 7
H
18
+
19
+
19 18
10
Scheme 1.2: Cyclization of 7 to the cis-neoclerodane skeleton.
42
20 m
20 m
Figure 1.10: Peltate glandular trichome (SEM stereoview).39
1.3.1.4
Distribution.
In a series of simple and elegant experiments, Siebert has demonstrated that the salvinorins are not evenly distributed through the tissues of S. divinorum, but are localised in particular structures: peltate glandular trichomes (Figure 1.10).40 These are found particularly on the undersides of the leaves densely packed on newly formed leaves, more sparsely distributed on mature ones (Figure 1.11). Other types of trichome, glandular and non-glandular, are also visible. Terpenoid accumulation in glandular trichomes is typical of the Lamiaceae family.40
This nding is signicant for future isolation work. Siebert found that dipping fresh leaves in CHCl3 (30 seconds 3) gave nearly complete recovery of salvinorins.40 Powdering the leaves, as has been done in all isolation procedures to date, is therefore unnecessary (see Section 2.1.1 on page 67).
1.3. CHEMISTRY.
43
200 m
200 m
200 m
200 m
Above: immature leaf (1 mm wide); below: mature leaf (10 cm wide).
Figure 1.11: Underside of S. divinorum leaves (SEM stereoview).39
1.3.2
1.3.2.1
Alleged alkaloids.
Summary of the Original Report.
The rst investigation into the chemistry of S. divinorum was reported by Jos Luis Daz in 1975.41 The report is in Spanish, but key sections have been translated by Valds, who also added owcharts clarifying Dazs procedures.42 The report was not peer reviewed, and the workers involved were not named. Valds, who later visited the lab, reports that the chemistry was performed
44
by undergraduate biology and botany students with a minimal chemistry background.43 Many essential details were omitted. For instance, in some cases TLC data were given without the solvent system. Administration of certain extracts reportedly caused abnormal behaviour and posture in cats. These are, however, not evident in the accompanying photographs. Moreover, there were apparently great variations and inconsistency in the results, which are not specied: the described behaviour is not always present. The assay results are reported simply as active, inactive or dubiously active without further detail, and without specifying the number of subjects or trials. No positive or negative controls were used to validate the assay. Given these serious deciencies, the procedures and results will not be reproduced here. The interested reader will nd Valdss translation helpful.42 One important omission should be noted: the isolation procedure Valds describes as method 2 was performed twice: on the rst occasion, as he notes, one of the fractions was active. On the second occasion, no fraction was active, yet another inconsistency. The results can be briey summarised as follows: certain fractions of the plant extract, soluble in aqueous acid but insoluble in aqueous base, sometimes caused cats to behave abnormally. Other fractions never showed clear activity. The active, acid-soluble fractions contained at least four compounds which gave positive reactions to Dragendors reagent,44 a standard alkaloid visualisation reagent, in both standard and modied (Ldy-Tenger)45, 46 forms. On this basis, Daz concluded that several alkaloids exist in Salvia divinorum, two of them apparently psychoactive. In 1977 he reported that the structures of the two compounds were under study.20 In 1979, however, he reported that:
It has been particularly dicult to identify the substance(s) responsible for these interesting eects. There exists a great variability or instability in the constituents of S. divinorum, which has impeded the consistent reproduction of the mental or behavioural alterations, preventing the identication of the active fraction. Some
1.3. CHEMISTRY. initial observations indicated the presence of nitrogenous compounds, possibly amino acids or amines, although they now appear to be of no pharmacological interest.47
45
These remarks, while vague, are not consistent with the original report. Presumably attempts were made to replicate the original experiments, but the results proved irreproducible.
1.3.2.2
Discussion.
The phytochemistry of the genus Salvia has been thoroughly studied,3 yielding hundreds of terpenoids.2 Yet extensive literature searches48, 49, 50 revealed no report of an alkaloid from an American Salvia species. The saps of several American Salvias have given positive results to Dragendors reagent,51 but far more tested negative. Furthermore, Dragendors gives false positives with numerous non-nitrogenous compounds.44 The American species S. reexa gave false positives to several alkaloid test reagents.52 The compound responsible proved to be choline [Me3 NEtOH]+ . Being a quaternary ammonium compound rather than an alkaloid, choline could not be extracted from basic aqueous solution by chloroform.52 This is also true of non-nitrogenous Dragendors-positive compounds,44 and thus none of these compounds can account for Dazs results. Daz41 cites a secondary source53 reporting that histamine occurs in the genus Salvia. The primary source54 cited there stresses that histamine is a primary metabolite, formed by decarboxylation of the amino acid histidine. It is therefore described as a biogenic amine rather than a true alkaloid, which in the strict sense of the word are secondary metabolites. Recent work on some Mexican Salvia species has found a very close chemotaxonomic relationship with Chinese species,2 some of which have yielded alkaloids. In particular, several seco- and nor-abietane diterpenoids previously known only from the Chinese species S. miltiorrhiza have been isolated from Mexican
46
Salvias.2 S. miltiorrhiza has also yielded alkaloids.55 Thus, the presence of alkaloids in S. divinorum cannot be dismissed out of hand. Subsequent work has failed to replicate any of Dazs ndings, however. Valds found no compound in the crude extract which gave a positive reaction to Dragendors or other alkaloid-specic test reagents.56, 23, 57 He has hypothesised that Dazs group actually used mislabelled Erlichs reagent, which reacts with alkaloids but also furanolactones, and thus mistook the salvinorins for alkaloids.57 This hypothesis, however, is inconsistent with the published claims; all denitely active fractions were soluble in aqueous acid, unlike 1a. Thus, if the active fractions contained the salvinorins then not only the visualisation reagent, but the fractions themselves, must have been misidentied. Moreover, the activity reported by Daz was dramatically dierent from that later observed in cats by Valds. Daz reported intense attention, reactions of fear and attack and fury, lasting 10 minutes after intravenous injection.41, 42 The effects observed by Valds were almost the opposite. Subcutaneous injection of an extract caused erratic eye movements rather than intense attention, and loss of physical coordination (the cat could not walk, much less assume postures of fear and attack). The eects lasted over 24 hours rather than 10 minutes.58 Chemical investigations by several groups have now yielded a total of 20 terpenoids, none of which is soluble in aqueous acid (see below). Furthermore, GC/MS29 and LC/MS59 analyses of the crude extract gave no indication of the presence of alkaloids. Compounds containing an odd number of nitrogens give characteristic molecular ions and fragments, identiable by the nitrogen rule.60 In summary, the claim of biologically active alkaloids was implausible and irreproducible, and has been abandoned by its author. The claim was evidently false.
1.4. PHARMACOLOGY.
47
1.4
Pharmacology.
1.4.1
Animal Testing.
1.4.1.1
Cats and Rats.
As discussed above, Dazs tests in cats yielded no useful results. The next worker to study the plants pharmacology was Valds. Although he consumed the traditional infusion during his initial work in Oaxaca, and later tested it for activity after freeze-drying, further use of human subjects was precluded61 (i.e. forbidden by his supervisor). For the purposes of bioassay-guided fractionation, Valds therefore had to develop an animal assay. This presented an enormous challenge. Previous attempts to identify the active principles of hallucinogenic plants using animal assays, even when sustained and well funded, had failed.7 In each case the active principle was later identied, quickly and cheaply, using human tests. Examples include mescaline (from Lophophora williamsii), psilocybin (Psilocybe cubensis) and lysergic acid amides (Ipomoea and Turbina spp.)62 Similarly, initial testing of LSD in mice caused no apparent eect other than disquiet.10 To these must of course be added Dazs work with S. divinorum. Evidently, and unsurprisingly, it is practically impossible to tell if an animal is hallucinating. In Valdss preliminary trials, standard hallucinogen assays in rats and cats were not sensitive to the eects of the extract.63 However, the cats exhibited impaired motor coordination, which reminded him of the impaired physical coordination and slurred speech caused by the infusion.16 The duration of action was much longer, however (24 hours). Given the diculty of detecting psychological states in animals, Valds decided to choose an assay sensitive to this physical eect.
48 1.4.1.2 Mice.
A standard assay of impaired locomotor function in mice, the inverted screen test,64 proved sensitive to the crude extract, but insuciently so. Next, a modied65 version of the open eld assay66 was tested. The movements of mice on a printed grid were recorded over 15 minutes. Three measures of activity were recorded: lines crossed, number of rearings, and time spent immobile. This assay showed unambiguous eects, with a clear dose-response relationship. Assay-guided fractionation led to the identication of 1a as the active principle. No activity was seen with 1b, the only other pure compound isolated. However, various mixed fractions gave puzzling results.
300 lines crossed 250 rearings 200 150 100 50 0 10 20 30 40 50 100 dose (mg/kg) log scale 1a time immobile (min) lines crossed 60 50 40 30 20 10 0 10 20 30 40 50 log scale impure 1a (~10% 1c) 100 dose (mg/kg) rearings 16 14 12 10 8 6 4 2 0 10 20 30 40 50 log scale mixtures 100 dose (mg/kg) immobility
Figure 1.12: Activity of pure 1a versus mixtures in mice (open eld assay). Using the open eld assay, impure 1a (later found to contain 10% 1c) was found to be signicantly more potent27 than the pure compound.65 It was therefore long suspected13, 27, 67 that 1c was also psychoactive. The open eld assay results are shown in Figure 1.12. The potency of the fraction containing 10% 1c was at least 10 greater than pure 1a, by all three measures (Figure 1.12). Thus, if Valdss interpretation68 were correct, 1c would be at least 100 more potent than 1a, making it among the most potent psychoactive compounds ever discovered. Three other mixed fractions containing 1a also gave higher ecacies than the pure compound, despite containing as little as 10% 1a.65 At 100 mg/kg, the dilute fractions almost completely immobilised the mice, while pure 1a only reduced activity. Despite their higher ecacy at high doses, however, these
1.4. PHARMACOLOGY.
49
fractions were not in fact more potent: the lines of best t suggest that pure 1a had comparable ecacy at intermediate doses, and higher ecacy at low doses (Figure 1.12). Thus, these results are not consistent with the presence of a more potent compound in the mixed fractions, but appear to be confounded. The anomaly was not specic to the open eld assay: the same pattern was evident69 in the inverted screen assay. The confounding factor was probably dierences in absorption, caused by the unusual method of administration. The test fractions were injected in an emulsion of vegetable oil, water and Tween 80 surfactant.70 The use of oil/water/surfactant emulsions as drug vehicles has proven eective for some hydrophobic drugs,71 and 1a is indeed hydrophobic. However, to be eectively delivered in an emulsion, drugs must also be highly lipophilic: Generally the most dicult drugs are those which have limited solubility in both water and lipids (typically with log P values of approximately 2). It is unlikely that lipid formulation will be of value for such drugs.71 The solubility of 1a is negligible in hexanes, and presumably in oil (predicted log P = 1.8).72 Thus, this vehicle would be expected to give poor absorption; 1a may have been administered not in solution, but as a suspension. The 1a tested by Valds was crystalline. Two of the impure fractions were explicitly described as oily solids.73 The other two were not described;
however, they were mixtures, and were obtained by evaporation from 10% MeOH/CHCl3 .73 Evaporation of 1a from chlorinated solvents gives an amorphous solid even when pure. Hence, all four of the impure fractions shown in Figure 1.12 were amorphous. This would result in dierences in solubility; the amorphous state of a solid typically has 2-10 higher solubility than the crystalline state.74 This results from the energy barrier to dissolution of a crystalline solid: disruption of the ordered crystal lattice requires additional energy (the enthalpy of fusion, Hf ).74, 75 Also, in a vigorously agitated emulsion, the amorphous solids could form a dispersion of microscopic particles,
50
with a higher surface area than macroscopic crystals. This in turn would increase the rate of dissolution (kinetics of solubility).75 Thus, faster dissolution and greater absorption of the amorphous fractions would be expected. Other compounds present in the mixtures might also inuence absorption. Many instances have been reported of enhanced absorption of active compounds from a crude extract relative to the pure compound.76 For instance, some terpenoids are known to act as permeation enhancers,77 increasing transdermal absorption of co-administered drugs up to 90. Thus, inactive compounds in the crude fractions may have increased absorption of 1a. Recent evidence conclusively establishes that the emulsion was very poorly absorbed. When injected in solution (EtOH/surfactant/H2 O), 1 mg/kg of 1a reduced locomotor activity signicantly over 30 minutes.78 Reinvestigation by Valds et al conrmed this, showing that 0.5 mg/kg caused maximal impairment in the inverted screen test;79 quadrupling the dose did not increase ecacy. By contrast, in the original tests using the emulsion, maximal impairment only occurred above 1500 mg/kg.69 Thus, absorption of 1a from the emulsion was clearly negligible. The use of an emulsion also seems to prolong the eect of 1a; performance on the inverted screen test remained impaired after 30 minutes.15 By comparison, the eects of the solution peaked within 5 minutes, and were undetectable by 15 minutes.79 Consistent with this, a recent study found a strong analgesic eect of 1 mg/kg injected 1a solution at 10 minutes,80 while a previous study which began testing 20-25 minutes after injection found very little analgesic eect even at 40 mg/kg.81 Another study reported analgesic eects at 0.6 mg/kg, but unfortunately contained no detail on administration or timing.82 In summary, the open eld results may have been confounded by dierences in bioavailability. The vehicle used gave very poor absorption, and superior absorption of amorphous solids is typical. Thus, the dierences in apparent potency between pure 1a and the impure fractions do not provide reliable evidence of the presence of other active compounds. Evidence that 1c and
1.4. PHARMACOLOGY.
51
other compounds in the plant are inactive will be presented in the next section, and in section 4.5.1 on page 171.
1.4.1.3
Rhesus Monkeys (Postscript).
Studies administering 1a to rhesus monkeys have recently been reported.83, 84 Interestingly, even the highest dose tested (32 g/kg) produced only slight overt behavioural eects,83 described as sedation-like in both studies, conrming the diculty of establishing animal models of psychoactivity.
1.4.2
Human Testing.
As mentioned above, Valdss open eld assay results indicated that 1a was the active principle of S. divinorum. Further testing with other compounds in the open eld indicated that 1a was approximately equal in potency to mescaline.13, 14 However, this conclusion remained tentative. As Valds had written earlier,
... there is no denite evidence that divinorin A is an hallucinogen ... the results are as yet unclear. And they will probably remain so until the divinorins are tested in human beings.68 The question was nally resolved by Siebert in 1994.18 He reported that 1a was hallucinogenic when vaporised and inhaled. Activity was perceptible with doses as low as 200 g. The eects commenced within seconds, and the strongest eects lasted 5 10 minutes. No eects were detectable after 30 minutes, suggesting a half-life of under 10 minutes. Siebert also explored the eect of the route of administration. Swallowing encapsulated 1a in very large doses (10 mg) produced no eect. This also held for the plant: an infusion prepared from ten fresh leaves, when swallowed, produced no eect in any subject. The same amount, when held in the mouth for 10 minutes and spat out, produced denite eects in all subjects. Thus, absorption through
52
the oral mucosa is clearly far more ecient than through the gastrointestinal system. Nonetheless, an ethanolic solution of 1a gave inconsistent eects sublingually. The onset of eects is slower by the sublingual route, and the eects last longer. Sieberts results were soon conrmed by others.7, 85 Ott and Gartz86 also reported that sublingual application of 1a was eective in acetone or Me2 SO, with potency comparable to inhalation. Thus, far from being equipotent with mescaline as the open eld assay indicated, 1a is in fact 1000 more potent.14, 7 Indeed, it is the most potent naturally-occurring hallucinogen yet isolated.7 The rapid metabolism of 1a has been conrmed in vitro87 and in vivo.88 Threshold doses of 1a produce visions of coloured patterns overlaid on reality, reminiscent of the pseudohallucinations induced by indole and phenethylamine hallucinogens. These patterns are faint, and only perceptible under dark and quiet conditions.14 Higher doses, however, produce intense and unique eects like those described earlier (Section 1.2 on page 33). Awareness that the visions are drug-induced is lost, and true dreamlike hallucinations occur. These experiences frequently involve certain themes not found with other hallucinogens. The distinction between self and surroundings is lost; subjects often feel that they are blending into, or have become, inanimate objects.18 Similarly, the distinction between past and present is weakened; subjects will relive events, often from childhood, rather than simply remembering them. Sieberts initial reports of these unique themes have again been conrmed by others.85, 89 Further research has revealed no evidence of active compounds other than 1a. Siebert conrmed that 1b is inactive.40 He also found that self-administration of 3 mg90 of 1c, sublingually in acetone, had no noticeable eect;40 this is 10 a threshold dose of 1a by that route. Evidence that 1c and other terpenoids in the plant are inactive in vitro will be presented in Section 4.5.1 on page 171. In conclusion, although denitive proof awaited human testing, Valds nonetheless correctly identied the active principle of S. divinorum using animal as-
1.4. PHARMACOLOGY.
53
says. The magnitude of this achievement is not widely appreciated. As noted above, previous attempts to identify plant hallucinogens using animal assays had invariably failed, while human assays had succeeded, quickly and cheaply. Valdss supervisor had thus forbidden the only demonstrably eective technique available. Moreover, in this case even human testing had failed. And it had been performed, independently, by arguably the two greatest authorities in the eld, Alexander Shulgin91 and Albert Hofmann92, 10 (twice in the latter case).93 Consider also the prevailing consensus when Valds began work: the active principle, whose eects were barely perceptible, was water-soluble and well-absorbed orally, but unstable and destroyed by drying. This consensus proved false in every detail. The active compound(s) also appeared to be alkaloid(s), which was also false. Sadly, despite overcoming these myths, Valds considers the time he spent on animal testing wasted, believing it cost him priority on the discovery of 1a.94
1.4.3
In vitro Testing.
After establishing that 1a was psychoactive, Siebert submitted the compound for in vitro screening against potential molecular targets.18 The NovaScreen assay tested for radioligand binding inhibition at targets including receptors for small-molecule and peptide neurotransmitters, as well as ion channels and enzymes. No binding was detected at 10 M. Subsequently, the compound was screened by Roth et al against a much larger battery of targets, the National Institute of Mental Healths Psychoactive Drug Screening Program. This revealed that 1a bound with high anity to the opioid receptor (K i = 4 nM).67 Functional testing showed that it activated the receptor with full ecacy and high potency (EC50 =1 nM). Thus, 1a is a potent full agonist at the opioid receptor. This result has since been replicated by other groups.82, 81, 95 Further conrmation has come from in vivo testing. The eects of 1a in mice are blocked by selective opioid antagonists;81, 78, 80 rhesus monkeys trained to discriminate opioids recognised 1a as such, and
54
the eects were blocked by a nonselective opioid antagonist.83 No binding was apparent in vitro to any of 48 other CNS targets at 10 M; 1a is thus extremely selective compared to most other psychoactive compounds.67 However, the strength of this conclusion is contingent on the number of targets tested; many orphan receptors exist for which anity testing cannot yet be performed.96 However, in vivo conrmation of the selectivity of 1a is available. Siebert found that the eects of 1a were blocked by naloxone, a nonselective opioid antagonist.97 This makes it highly unlikely that a nonopioid mechanism contributes independently to the compounds eects.
1.4.4
1.4.4.1
Mechanism: Opioids.
Discovery.
The strongest painkiller available in the preindustrial era was opium, a milky secretion of the opium poppy Papaver somniferum. The isolation of morphine (11) from opium was reported by Sertrner in 1806; he also showed the compound to be a potent analgesic.98
N H
HO
O H
11 OH
N OH
O HO 12 HO 13 HO O 14
Figure 1.13: Morphine (11) and derivatives. Morphine has had an immense impact on science. It was the rst drug the rst known pharmacologically active compound.99 It was the rst alkaloid;
1.4. PHARMACOLOGY.
55
indeed, the word alkaloid was coined to describe it.98 Its structure elucidation took over a century, which is unsurprising since morphines discovery predated the concept of molecular structure by several decades. Indeed it predated the publication of Daltons atomic hypothesis, which was regarded with skepticism for decades.100 Morphines very strong analgesic eect is accompanied by serious side eects: nausea, respiratory depression and constipation.101 Intense euphoria also occurs in some, making the drug highly addictive. The pursuit of an analgesic lacking these side eects gave rise to medicinal chemistry. The rst morphine derivatives were synthesised in the 1850s;98 the structure-activity relationships of morphine have since been explored more thoroughly than those of any other compound, with thousands of derivatives synthesised.101 Simplied structures were found to retain activity: morphinans such as butorphanol (12), benzomorphans such as ketazocine (13) and remarkably even phenylpiperidines such as pethidine (meperidine, 14). Some of these compounds (such as 14) closely mimic morphines actions. Some, however, proved to be antagonists, which have proven immensely valuable in reversing opioid overdose. Other derivatives caused quite distinct behavioural eects, but were not antagonists and did not exhibit cross-tolerance. The study of these dierences led to the discovery of opioid receptor subtypes. The and subtypes were named after the archetypes morphine (11) and ketazocine (13), while comes from the vas deferens, in which that subtype was discovered.101, 102 These receptors are commonly abbreviated as MOR, KOR and DOR. Neither of these terminologies is endorsed by the International Union of Pharmacology, but the ocially sanctioned names, OP13 ,102, 78 have not gained wide acceptance. The study of opioid receptors led to the discovery of the endogenous ligands, the endorphins.101, 102 Remarkably, recent work has proven that morphine is endogenous in humans.103 Opioids are compounds which act at opioid receptors, and whose eects are reversed by the antagonist naloxone.101 This terminology lends itself readily to
56
selective compounds; thus, 1a will be referred to below as a opioid. This is synonymous with the common but unwieldy tautology opioid receptor agonist.
1.4.4.2
Development.
N N N N O Cl Cl 15 16 R O O
Cl Cl 17 18 O
Figure 1.14: Arylacetamide opioids. U50,488 (15) was devised as a structurally simplied morphine derivative.104, 105 In animal tests, U50,488 was an eective analgesic, whose eects were reversed by naloxone. Remarkably, however, the compound did not produce physical dependence like morphine, and was not self-administered. Initially referred to as a nonopioid, 15 was soon found to be the rst selective agonist at the opioid receptor. These remarkable ndings inspired extensive research; numerous derivatives were synthesised, some of which proved to be even more potent and selective.105 Examples include U69,593 (16), spiradoline (U62,066, 17) and enadoline (CI977, 18). Testing of these compounds conrmed the ndings with 15: at last it appeared that the long-awaited nonaddictive opioid analgesics had been found.
1.4.4.3
CNS Eects.
Addiction can be studied using animal models. The opioids are positively reinforcing. This means test animals trained to self-administer them will do
1.4. PHARMACOLOGY.
57
so compulsively, at the expense of social activity, food and sleep. By contrast, opioids (including 1a)78 are negatively reinforcing, or aversive. Test animals will not self-administer them; if administered when an animal behaves in a certain way, the animal will avoid that behaviour. In other words, opioids act as a punishment, where opioids act as a reward. This is a vast eld of research in its own right, and a thorough review is available.106 The mechanism of these eects, believed to be mediated by dopamine, has also been thoroughly studied.78, 107 Since the objective had been to eliminate the euphoria associated with opioids, the aversive nature of opioids was initially regarded as desirable. Human tests soon revealed a problem: rather than euphoria, opioids produced dysphoria. This lowering of mood was accompanied by psychotomimetic eects: confusion, hallucinations and depersonalisation. Enadoline (18), for instance, caused side eects such as somnolence, hallucinations, anxiety, depersonalisation, confusion and abnormal thinking; the severity of these eects led to termination of clinical development.108 Subsequent tests with higher doses reported visual, auditory and tactile hallucinations, along with impaired coordination and recall. One subject felt waves moving through the oor, and felt his body was blending into them109 (cf. 1a).18 Spiradoline (17) caused altered perceptions, impaired coordination and slurred speech; subjects reported being more irritable, anxious and sad.110 Responses to other opioids range from personality disorders and mild confusion111 to depersonalisation, dreamlike experiences and episodes of unmotivated and uncontrolled laughter112 (again cf. 1a).18 In one study, heroin users were given various substances and asked do you like the drug? and does the drug have any good eects? They consistently answered yes after administration of a opioid, but no after enadoline (18).109 The opposite was true when asked about bad eects. Similar results were reported in tests of nonselective opioids such as 13113 and 12.114 Subjects with extensive histories of illicit drug use reported dysphoria, hallucinations,
58
paranoia and confusion after the opioids, but euphoria after morphine. Thus, both drug-nave subjects and experienced illicit drug users overwhelmingly nd opioids dysphoric and aversive.
1.4.4.4
Therapeutic Potential.
Clinical development of these compounds as analgesics was eventually abandoned due to these psychological eects.115, 105 Some opioids which do not cross the blood-brain barrier, and are therefore not psychoactive, remain in development for arthritis116 and abdominal pain.117 This is interesting in light of Mazatec use of S. divinorum infusion against abdominal pain and swelling.11, 15 Besides analgesia, many other therapeutic uses have been proposed for opioids.115, 118, 119, 120 None of these therapies has reached clinical use. There has been a report of an antidepressant eect of U50,488 (15) in an animal model.121 However, there was at the time no validated protocol for that model (learned helplessness in mice),122 and compounds of known activity were not used as controls. In another model, the forced swim test in rats, one study reported that a opioid had no eect.123 However, other studies reported a prodepressant eect,124 one of which used 1a.107 This conclusion, which is consistent with the aversive and dysphoric eects discussed above, is also more consistent with prior results and recent theoretical insights.107, 125 Thus, it seems that opioids exacerbate depression in animal models. Surprisingly, however, there have been reports of antidepressant eects from S. divinorum use.126, 127 However, the plant was used only three times weekly. It may be that, just as the acute euphoria caused by opioids is followed by prolonged dysphoria, the converse may be the case with opioids.
1.4.4.5
Claimed Subtypes.
There have been claims that certain opioids lack dysphoric eects. TRK 820 (19, Figure 1.15 on the facing page) allegedly produces moderate behavioural/psychological side eects, but not psychotomimetic activity, but
1.4. PHARMACOLOGY.
59
this is based on unpublished data.128 It is also often claimed that 19 is not aversive in animals,128 which is untrue.129 This is reminiscent of earlier claims that enadoline (18) appears to be devoid of psychotomimetic activity,130 which was subsequently and spectacularly discredited as mentioned above. One proposed mechanism for dierences between opioids is that there are subtypes of opioid receptor.131 Dierences between opioids might reect dierent selectivities for these subtypes. The evidence for subtypes is purely pharmacological131 the gene encoding the receptor has now been cloned from several species, with no subtypes detected. For this reason, and because of the lack of selective ligands for each proposed subtype, the claim has not won general acceptance.132 The apparent subtypes may in fact represent dimer formation between dierent receptors.131, 132
1.4.4.6
Structure, Potency and Selectivity.

N OH O O N
HO
O H 19
N O
HO 13
phenylalkylamine moiety N N O O F O 18 N HN N 20 O S
Figure 1.15: Structurally diverse opioids. The major structural classes of opioid are peptides such as dynorphin A, morphine derivatives such as 19, benzomorphans such as 13 and arylacetamides such as 18 (Figure 1.15). The rst major departure from these categories was the benzodiazepine tiuadom (20).105 The discovery of such compounds, and
60
similarly diverse ligands at other subtypes, eventually rendered generalisations about structure-activity relationships impossible.101 Nonetheless, as diverse as these compounds may appear, they are all alkaloids. 1a was the rst non-nitrogenous opioid reported. Indeed, it was the rst opioid lacking the phenylethylamine moiety or its propyl homologue (Figure 1.15). These moieties are near-ubiquitous in hallucinogens,133 and indeed in psychoactive compounds generally. In a random sample of compounds from the Merck Index, 82% of psychoactive compounds contained a phenylalkylamine moiety, versus 8% of non-psychoactive compounds.134 Conversely, 58% of phenylalkylamines were psychoactive, versus 3% of other compounds. Being non-nitrogenous and lacking a benzene ring, 1a is thus not merely unique among opioids, but extremely unusual among psychoactive compounds in general. 1a 1 7 4.5 45 3.1 4.6 EC50 (nM) 15 16 19 1.2 24 13 4.5 207 3.9 2.2 0.025 1.1 0.0048 16 0.15 Ref 67 135 95 136 137 81 138 139
Table 1.1: Relative potency of 1a at cloned opioid receptors. The binding anity and potency of 1a in vitro are close to those of U50,488 (15) and U69,593 (16). Potencies are shown in Table 1.1. The most potent opioid to date appears to be TRK 820 (19). However, few groups have studied this compound, and their binding anity data are wildly discordant (K i = 75 pM81 vs 3.5 nM).139 The compound also has lower / selectivity than U50,488138 and other arylacetamides,104 which are in turn less selective than 1a.67 Indeed, there has been no report of 1a showing any anity at the opioid receptor (K i > 10 M). In summary, opioids have not lived up to initial expectations as analgesics.
1.5. TOXICOLOGY.
61
Ultimately, the enormous eort to improve upon morphine has failed. Two hundred years after its discovery, morphine remains the opioid of rst choice for severe pain under World Health Organization guidelines.140 There is no superior analgesic: Morphine remains the most widely used opioid for the management of pain and the standard against which other opioids are compared.141 Opioids remain the topic of extensive research, however, for other therapeutic purposes, as mentioned above.
1.5
Toxicology.
There has been little toxicological research on S. divinorum and 1a. Valds made several incidental observations on the topic. One of his rats died during oral administration of an extract, but this was due to choking on the large volume of liquid rather than toxicity.142 The crude aqueous extract proved highly toxic by injection in cats. Subcutaneous injection of 0.7 g/kg, approximately equivalent to a human dose of the infusion, caused a sterile abscess at the injection site.58 A higher dose (1.3 g/kg) caused kidney failure in two cats, one of which died. The other recovered with a sterile abscess. Based on the known toxicity of polyphenols in crude extracts,143 and veterinary examinations suggesting this was the cause,58 Valds prepared a polyphenol-free extract by partitioning between CH2 Cl2 /MeOH and water. The resulting organic fraction was not toxic, even in much higher doses, estimated to be equivalent to 635 the human dose.58 Further toxicity testing was performed in mice. The ether extract of the leaves was dissolved in aq. MeOH and washed with hexanes. The resulting methanolic fraction, which was presumably free of polyphenols, was nevertheless still toxic enough to kill the mice within a week (LD50 = 340 mg/kg i.p.)69 This is
62
approximately equal to the toxicity of vitamin B3.144 Pure 1a caused no fatalities at 1 g/kg,13 the largest dose which could be administered as an emulsion, and more than 105 the psychoactive dose in humans. As discussed above, however, the emulsion was very poorly absorbed. This would reduce the immediate systemic dose of the extracts, giving a misleadingly low impression of acute toxicity. On the other hand, the undissolved material at the injection site may have had toxic eects through another mechanism, such as shock. It would also presumably slowly dissolve, causing chronic rather than acute toxicity, especially to the liver. Surfactants such as Tween 80 can also have adverse eects.145 These data therefore cannot be used to make reliable comparisons with other substances. Mowry et al later published a study specically devoted to the toxicology of 1a.146 Injection of 6.4 mg/kg/day for two weeks caused no fatalities; postmortem examination of organs and tissues showed no apparent changes. Unfortunately, since the drug was administered intraperitoneally as a ne suspension, and behavioural eects were not described, the same issue of bioavailability arises as with Valdss results. It is also regrettable that larger doses were not administered to determine the LD50 . Human toxicological data are not available. There have been no published reports of toxic eects or hospitalisation related to S. divinorum or salvinorin A.147
1.6
1.6.1
Social impact.
Recreational Use.
S. divinorum came to the attention of recreational drug users even before chemical investigations began. Books published in 1973 described its use148 and cultivation,149 listing a mail-order source for live cuttings. By 1975, Ott reports seeing some young Mexicans smoking the dried leaves recreationally.7
1.6. SOCIAL IMPACT.
63
Although he claims Daz also reported this at the time, the latter was in fact referring to a dierent plant.41 The plant nonetheless remained extremely obscure until Sieberts ndings were popularised in 1996 by Turner.85 This immediately overturned the plants reputation as a weak substitute for illicit drugs:
Salvinorin A is the most potent naturally occurring psychedelic known ... it frequently induces experiences of an intensity [...] beyond those experienced with any other psychedelic ...85
The plants notoriety then began to grow. Online companies began selling live cuttings, dried leaves, extracts and tinctures for recreational use. Interestingly, however, after being widely available for a decade, the drug has not had any noticeable social impact (as for instance LSD and MDMA rapidly did). As Turner noted:
Many who have used salvinorin A nd the experience extremely unnerving, frightening and overly intense. Most have no desire to repeat the experience.85
This is consistent with the dysphoric and aversive qualities discussed above.
1.6.2
1.6.2.1
Legal Status.
Australia.
In 2002, Australia became the rst country to prohibit S. divinorum and salvinorin A. The committee responsible, the National Drugs and Poisons Schedule Committee (NDPSC), justied the decision on the basis of high potential for abuse and risk to public health and safety.150 Salvinorin A was prohibited under a purportedly systematic name:
64
8-METHOXYCARBONYL-4A,8A-DIMETHYL-6-ACETOXY5-KETO-3,4,4B,7,9,10,10A-SEPTAHYDRO-3-(4-FURANYL)2,1-NAPHTHO[4,3-E]PYRONE.150
This name is in fact meaningless. Compare the one used by Chemical Abstracts Service:
2H -Naphtho[2,1-c]pyran-7-carboxylic acid, 9-(acetyloxy)-2-(3-furanyl)dodecahydro-6a,10b-dimethyl-4,10-dioxo-, methyl ester, (2S,4aR,6aR,7R,9S,10aS,10bR)-48
NDPSC CAS 2S,4aR,6aR,7R,9S,10aS,10bR 7-carboxylic acid, methyl ester 8-METHOXYCARBONYL6a,10b-dimethyl4A,8A-DIMETHYL6-ACETOXY 9-(acetyloxy)4,10-dioxo5-KETO 3-(4-FURANYL) 2-(3-furanyl) dodecahydro3,4,4B,7,9,10,10A-SEPTAHYDRO 2H -Naphtho[2,1-c]pyran 2,1-NAPHTHO[4,3-E]PYRONE Table 1.2: Alternate systematic names for 1a. Enumerating all the errors in the NDPSC name would be beyond the scope of this work, but a brief selection will be given here. The names are contrasted in order of functional group in Table 1.2. Note especially the nonexistent terms 4-furanyl and septahydro. The latter was presumably intended to mean heptahydro, which is incorrect.
3 4 5 6 7 4a 4b 8 8a 9 10a 10 2 1 10 9 8 7 10a 1 10b 2
O
4a 5
6a
phenanthrene
2H-naphtho[2,1-c]pyran
Figure 1.16: Relevant IUPAC fused-ring numbering schemes. The name given by the NDPSC originated in a post to the internet newsgroup alt.drugs by William E. White:
1.6. SOCIAL IMPACT. I tried to get an IUPAC name out of this; unfortunately, I gagged when trying to decide whether it was a naphthopyrone, or a modied phenanthrene, or whether it should start from cyclohexane instead of aromatic rings, or whatever. The fact that phenanthrene is numbered funny didnt help. The best I could do is 8-methoxycarbonyl-4a,8a-dimethyl-6-acetoxy-5-keto-3,4,4b,7,9,10, 10a-septahydro-3-(4-furanyl)-2,1-naphtho[4,3-e]pyrone. (whew!) Sorry, I havent done o-chem since the Reagan administration.151
65
This explains the incorrect numbering of 1a: White named it as a naphthopyran derivative, but numbered it as a phenanthrene derivative (see Figure 1.16).152 The resulting name was then copied without acknowledgment by the NDPSC, with the sole change being capitalisation. After these criticisms153 of their decision attracted publicity,154 the NDPSC revisited the issue in late 2002. Conceding that the name used was incorrect, they decided to adopt the CAS name.155 However, they rearranged the substituents in nonalphabetical order, in conformance with their own unspecied naming conventions. The new name is therefore also incorrect. The NDPSC is Australias peak body for the regulation of chemicals, whose chief task is to list controlled chemicals by name. It is surprising and disturbing, therefore, that no member of this body understands basic chemical nomenclature. Whites posting also raises another serious issue. The newsgroup alt.drugs156 is a forum for users, dealers and manufacturers of illicit drugs to share ideas and advice. Publications which promote, incite or instruct in matters of crime cannot be legally imported into Australia, as for instance by downloading from the internet.157 How, then, did the committee legally obtain this material? Committee members repeatedly refused to reveal the source to a federal shadow minister.158, 159 Similarly, documents released under a freedom of information request160 contained nothing on this topic.
66
A disturbing question thus remains unanswered: how did obvious errors in a prohibited publication nd their way into Australian law?
1.6.2.2
Other Countries.
In 2003, Denmark became the second country to prohibit S. divinorum and salvinorin A.161 William Whites erroneous name was used again, with the locants capitalised. This new and peculiar error suggests that the authors copied the Australian version, changing the substituents to lower case, but leaving the locants in capitals. Ironically, the NDPSC had by this time already admitted the name was incorrect and adopted a new one. S. divinorum and/or salvinorin A have since been prohibited in other countries, including Belgium, Italy, South Korea and Spain.162 There have been similar moves in several US states, but a 2002 bill163 to prohibit the plant nationally met with well-organised opposition,147 and was not enacted.
1.7
Summary.
Previous work had thus established that 1a was a psychoactive opioid. However, virtually nothing was known of its structure-activity relationships, and only three other compounds had been isolated from S. divinorum. The suspicion that other psychoactive compounds were present persisted.
Chapter 2 Isolation.
2.1
2.1.1
Isolation Procedure.
Extraction Conditions.
The isolation methods reviewed above suer from a number of drawbacks. Both Valds et al 23 and Ortega et al 22 employed reuxing solvents for the initial extraction of the dried leaves. Subsequently, however, Gruber164 reported that extraction at room temperature was superior, based on HPLC analysis. He found that steeping for several days at room temperature gave markedly better recoveries of 1a than reuxing for two hours, in either CHCl3 or MeOH (see Table 2.1 on the next page). In optimising the room-temperature extraction, he found that highest recoveries of 1a were achieved after less than 30 minutes; levels then declined steadily (by nearly 50% over two days).164 Gruber speculated that this decline might be due to 1a precipitating out of solution as other compounds accumulated. This explanation is implausible since 1a is freely soluble in CHCl3 (evaporation gives an amorphous resin rather than a crystalline precipitate), and the solutions were extremely dilute. Furthermore, the decline is faster at reux, when solubility is greater. A more plausible explanation of the decline is decomposition, with the rate proportional to temperature. This conjecture was conrmed in the course of 67
68
CHAPTER 2. ISOLATION.
this work. When chromatographically pure 1a was recrystallised from EtOH, and the resulting mother liquor recrystallised in turn, new higher spots became apparent by TLC (70% Et2 O/petrol). After another recrystallisation, the mother liquor (18% of the original fraction) consisted almost entirely of the higher spots, with only traces of 1a. The decomposition products were not characterised. In contrast, however, Valds reports that he recovered 1a unchanged after reuxing in MeOH or MeOH/H2 O for 2 weeks.165 It may be that decomposition only occurs in the presence of oxygen. recovery CHCl3 1.75 1.41 (mg/g) MeOH 1.56 0.44
T rt reux
time 4d 100 min
Table 2.1: Eects of solvent and temperature on recovery of 1a. After his optimisation experiments, Gruber adopted a standard procedure of steeping the dried, powdered leaves in CHCl3 for 30 minutes. For scale-up, however, this presents the problem of handling and evaporating large volumes of this carcinogenic solvent. Therefore, for this work a less toxic substitute was sought, of lower polarity than MeOH (given the inferior recoveries achieved in that solvent). Acetone was selected as an aordable solvent meeting both of these criteria. Dried, powdered leaves were steeped for 3060 minutes in room-temperature acetone ( 3). TLC analysis (5% MeOH/CH2 Cl2 ) showed that the second extract was much lower in 1a than the rst; the third extract showed no detectable 1a. A further steep in CHCl3 also showed no 1a. This extraction procedure was therefore adopted for all of the work described below. Further improvements are possible. Powdering the leaves, for instance, appears to be unnecessary. Pseudonymous reports166 on the internet that 1a can be extracted in high yield from intact leaves have been conrmed: Siebert found that dipping fresh leaves in CHCl3 (30 seconds 3) gave excellent recovery of terpenoids with minimal contamination by pigments.40 Drying and powdering the leaves, followed by a further extraction, gave a pigment-rich extract with negligible terpenoids. Thus powdering the leaves before extraction is unneces-
2.1. ISOLATION PROCEDURE.
69
sary, and indeed counterproductive. While this analytical procedure has not yet been tested on a preparative scale or on dried leaves, these results strongly suggest that extraction procedures based on brief steeping of intact leaves will prove superior to the protocol given above. Claims166 that extraction in chilled solvents gives a cleaner extract also deserve investigation.
2.1.2
2.1.2.1
Problems Caused by Pigments.

Interference with Previous Isolations.
1a 1b youngest oldest
Developed in 50% EtOAc/petrol, visualised in vanillin/H2 SO4 .
Figure 2.1: Sieberts TLC analysis of crude CHCl3 extracts.40 The extract from the above procedure was a blackish-green tar. The numerous pigments present in this extract rendered TLC analysis dicult. Relative to the colourless terpenoids of interest, the pigments were present in larger quantities and stained more vividly, obscuring the terpenoid spots. The problem is illustrated in Figure 2.1. Daniel Siebert analyzed chloroform extracts of leaves of varying age, which had been dried and powdered.40 In all cases, the intensities of pigment spots were equal to or greater than 1a, and much greater than 1b. The pigments thus present a major obstacle to the isolation of other terpenoids, which are present in lower levels. For instance, 1c is not even detectable in the gure; in this system it appears immediately above 1a, which is itself nearly obscured in the youngest leaf extract. Valds et al had addressed this problem by partitioning the crude extract between aqueous MeOH and hexanes.23 Using his partitioning procedure proved dicult: both phases were black, and could not be distinguished even by close
70
inspection in direct sunlight. The interface was ultimately located by holding a torch against the separatory funnel and looking directly into the beam, which was dimly visible through the MeOH phase. The hexane phase was totally opaque. Another drawback to the procedure was the tendency of the MeOH phase to spatter during evaporation, but the water content (10%) was far too high to use a drying agent such as MgSO4 . After partitioning, while there was some reduction in mass, the extract remained intensely coloured. Gruber also encountered diculties with Valds et als procedure, obtaining greenish crystals of 1a even after partitioning, repeated chromatography and recrystallisation.167 HPLC analysis indicated purity of approximately 85% based on UV detection at 208 nm;168 however this cannot be translated to percentage by mass, since the coloured impurities would be expected to have higher molar absorptivities. Gruber found that the trace pigments could be removed by rinsing with cold MeOH.167 It was presumably to remove these trace pigments that Valds recrystallised 1a a second time, for which no explanation was given.23 This repeated recrystallisation is clearly a major drawback to the procedure, adding complexity and reducing recovery. Others have simplied Valds et als procedure, omitting the chromatography and simply washing the crude extract with less polar solvents to remove the pigments.86, 166 These procedures, while ingeniously simple, have not yet been replicated in the peer-reviewed literature. A recent paper took the opposite approach, omitting partitioning and recrystallisation in favour of chromatography. Tidgewell et al 169 isolated 1a in very high yield as a green powder, after repeated chromatography of the crude extract on silica gel. Despite the colouration, the material exhibited a reasonably sharp melting point range of 3 C; other evidence of purity was not reported. In summary, the pigments present in the extract caused considerable problems in the isolation of 1a. They therefore posed even greater obstacles to the isolation of the other diterpenoids present in S. divinorum, since these are present in much smaller quantities than 1a. Indeed, as noted at the start
71
of this section, the pigments made the mere detection of other compounds by TLC dicult. It was therefore critical to this work that the crude extract itself be decolourised, rather than individual compounds after isolation. Valds et als partitioning procedure was unsatisfactory, as was chromatography on silica gel. Gruber found that reverse-phase chromatography (using a C-18 solid-phase extraction cartridge), while removing some pigmentation, was likewise inadequate.167 Both of these techniques rely upon dierences in polarity to achieve separation. Since the pigments exhibited a very wide range of polarities, spreading from baseline to solvent front on TLC (Figure 2.1), another basis for separation was clearly necessary.
2.1.2.2
Chemistry of Plant Pigments.

O N Mg N N O N O O O 21
OH HO O O OH O O OH OH OH 23 OH
22
Figure 2.2: Representative major plant pigments. The overwhelming majority of plant pigments fall into three categories: chlorophylls (eg. chlorophyll-a, 21, Figure 2.2), carotenoids (eg. -carotene, 22) and avonoids (usually present as glycosides eg. quercitrin, 23).170, 171 In addition
72
to these ubiquitous classes, there are many minor pigments conned to particular taxa, which have been reviewed elsewhere.170 The common factor among these pigments is an extended system, often but not always aromatic, which results in intense absorption of visible light. As a result, pigments are generally large molecules. They vary widely in polarity, from extremely hydrophilic (eg. 23) to extremely hydrophobic (eg. 22). The chlorophylls themselves vary: one of the earliest procedures for separating chlorophylls a and b was partitioning between aq. MeOH and petrol.171 Thus, it is unsurprising that Valdss partitioning procedure, and polarity-based fractionations in general, are of limited eectiveness in decolourising plant extracts.
2.1.3
Use of Activated Carbon.
The standard method for decolourisation of a solution is adsorption, and the most common adsorbent is activated carbon, often referred to as decolourising carbon.172 Activated carbon has been in widespread use, both scientically and industrially, for centuries. The characteristics of compounds which aect adsorption are well-established empirically.173 Within a homologous series of compounds, adsorption increases with length. In addition, aromatic compounds are strongly adsorbed, especially polycyclic aromatic compounds. Generally, increased polarity reduces adsorption, but this eect is smaller, and easily overcome by size. Polyphenols, for instance, are strongly adsorbed despite being freely soluble in water. Many other variables aect adsorption to lesser extents.173 While these general trends are well-established, the underlying mechanisms remain in dispute.
2.1.3.1
Mechanism of Action.
The surface chemistry of dierent types of activated carbon varies, from highly oxidised forms with numerous polar functional groups distributed across the surface, to graphitised forms with very little functionality.173 The most important common characteristic of all forms is their enormous surface area,
73
estimated at up to 1500 m2 /g in some cases,174 resulting from the materials extremely porous structure. Recent work suggests that this is the key factor controlling adsorption. In solvophobic theories,175, 176 adsorption is treated as being driven by two key factors: the (unfavourable) formation of a solvent cavity around the solute, and (favourable) van der Waals attraction between the solute and the carbon surface. As a result, the theory predicts the adsorbability to depend linearly on the nonpolar surface area of the adsorbate.176 This simple model predicts very accurately the adsorption data for a wide range of compounds.175, 176 Another recent study, based on a much larger data set, used a more complex model (the linear solvation energy relationship model).177 Interestingly, however, despite being based on dierent parameters, the study reached very similar conclusions. Planar molecules were found to be more strongly adsorbed than nonplanar molecules. This was rationalised as follows:
Since the dispersion force of attraction is very sensitive to the distance of separation between the surface of the activated carbon and the center of the solute molecule, it is reasonable that a planar molecule would be adsorbed more strongly than an otherwise similar globular molecule.177
That is, consistent with the solvophobic model, surface area available for contact is the key factor in determining adsorbability. Interestingly, this eect was found not only for aromatic compounds; planar olens were just as strongly adsorbed. A similar eect had been observed in earlier work,178 based on another model, and used to predict that square planar organometallic complexes would adsorb more strongly than octahedral complexes. This prediction was conrmed experimentally.178 This also helps explain the strong adsorbability of carotenoids such as 22, which although aliphatic can assume planar conformations which present a much larger surface area to the carbon surface than less conjugated hydrocarbons. Indeed, in this work non-conjugated terpenoids
74
comparable in size to 22 were easily desorbed from carbon (Section 2.2.4.2 on page 97), while 22 and other carotenoids were not. These works establish a strong theoretical and empirical relationship between surface area and adsorbability. Nonetheless, other factors aect adsorption, and the underlying mechanisms remain controversial, even for particular classes of compound.179
2.1.3.2
Use during Recrystallisation.
The most common application of activated carbon by chemists is in removing coloured impurities from crude compounds. Typically, a small amount of powdered activated carbon (12% w/w relative to the crude compound) is added during recrystallisation; the decolourised solution is then ltered before cooling.172 In many cases, however, this is not eective. Indeed, during the preliminary work described above, ash column chromatography of the crude extract gave an orange solid which was rich in 1a by 1 H NMR. During recrystallisation from MeOH, the bright yellow colour of the solution was not diminished by activated carbon, even at 10% w/w. Upon cooling, the resulting crystalline 1a remained faintly yellow, and the mother liquor vividly so. While this might be dealt with by using larger amounts of activated carbon, this would also result in increased adsorption of 1a itself, and therefore reduced recovery. Even if the procedure had been eective, moreover, it would still only be useful for removal of trace pigments from individual compounds. As noted above, it was essential to this work that the crude extract itself be decolourised. This would therefore require the use of a large amount of activated carbon. The compounds of interest would then have to be desorbed, in what would be eectively a chromatographic procedure.172
2.1. ISOLATION PROCEDURE. 2.1.3.3 Use in Chromatography.
75
Activated carbon is rarely employed as a stationary phase in chromatography. It is not suited to general use; for most separations, other adsorbents oer superior selectivity and resolution.180 Nonetheless, it has found use in certain niche applications. Perhaps the most common use has been the isolation of antibiotics181 and enzymes182 from microbial fermentation broths. Activated carbon remains a standard part of such procedures, although alternatives have been proposed.183, 184 Another application where activated carbon has proven superior to conventional adsorbents is in the separation of fullerenes,185, 186 giving better resolution and recovery than alumina, along with higher speed and lower cost. The larger ellipsoidal fullerenes, above C70 , are very dicult to desorb.186 Activated carbon has also proven valuable in separation of polychlorinated biphenyls (PCBs).187, 188, 189 Commercial PCB mixtures are very complex, containing hundreds of PCBs diering in the number and position of chlorine substituents.189 Many of these compounds dier very little in polarity and volatility, resulting in incomplete resolution by either HPLC or GC. This problem can be resolved by multi-dimensional GC, but activated carbon permits separation on a dierent basis: extent of ortho-substitution.189 Non-o-substituted PCBs (such as 24, Figure 2.3) can assume a coplanar conformation, and are therefore very strongly adsorbed (as expected). Each o-substituent presents a steric barrier to this conformation, and thus reduces adsorption, reaching a minimum with fully o-substituted PCBs (such as 25). PCBs can therefore be eluted from a carbon column in decreasing degree of o-substitution.190, 188 These fractions are greatly simplied relative to the initial mixture, and can be further puried using conventional techniques.187, 188 This method also has the attraction that degree of o-substitution is of great biological signicance. In addition to their toxicity, non-o-substituted PCBs such as 24 are extremely teratogenic; this eect is reduced by one o-substituent, and abolished by more than one.187
76
Cl Cl Cl Cl Cl Cl 24
Cl
Cl
Cl Cl 25
Figure 2.3: Ortho- and non-ortho-substituted PCBs. A surprisingly uncommon use of chromatography on activated carbon is for decolourisation of crude extracts during isolation191, 192 or analysis193 of natural products. In most cases this is used, as here, to permit analysis of the crude mixture rather than purication of individual compounds. As with any adsorbent, eluent selection is critical to achieving eective separation. The eluotropic series on activated carbon given by Gordon180 is: H2 O < MeOH < EtOH < acetone < i PrOH < Et2 O < EtOAc < hexane benzene. In general, then, the larger and less polar the solvent, the greater its eluting strength. That benzene is the strongest eluent in the series is unsurprising given the strong adsorption of aromatic compounds. In recent decades, toluene has been used as a less toxic and equally eective substitute for benzene.189, 185 Indeed, toluene appears on the balance of the evidence to adsorb more strongly than benzene,177 which is unsurprising given its larger surface area. Consistent with this, recent work has shown that o-dichlorobenzene is a much stronger eluent again.186 Based on their measured adsorption constants,177 p-xylene would probably oer comparable eluting strength, with the advantage of a lower boiling point (138 vs 178 C). Another approach to elution of very strongly adsorbed compounds is exhaustive extraction with hot solvent, using Soxhlet apparatus or similar.187 A major obstacle to practical chromatographic use of activated carbon powder is that it packs very tightly; it is almost impermeable to solvent even under pressure.180 One solution to this is to uidise the activated carbon, using a stream of nitrogen to remove the nest particles.187 Similarly, industrial-scale water treatment usually employs granulated activated carbon.174 Using larger
77
particles in this way, however, has the undesired eect of reducing available surface area. A more common solution is therefore to use mixed adsorbents.180 The most common choice is diatomite lter aid (eg. Celite),187 but other possibilities include cellulose powder193 and silica gel.189, 185
2.1.3.4
Application to S. divinorum Extract.
For this work, the mixed adsorbent approach was taken. Diatomite lter aid was selected due to its low cost, safety and ready availability. The procedure was performed as standard ash column chromatography, monitored by TLC.194 Initial testing on a small scale indicated that MeOH did not elute 1a, while acetone did so rapidly. Additionally, 1a is far more soluble in acetone. Some tailing was evident, so a stepwise gradient was employed based on Gordons eluotropic series,180 through Et2 O, EtOAc and EtOAc/petrol. Petrol was not used alone due to the insolubility of 1a. Gradient elution helped to minimise tailing, but did not eliminate it 1a was detectable from near the solvent front into the EtOAc/hexane fractions. No pigments were eluted when sucient carbon was used; a carbon:solute mass ratio of 20:1 was found adequate. At a ratio of 3:1, the adsorbent was overloaded, and orange pigments were eluted even by acetone. After the subsequent isolation of pure 1a (detailed in Section 2.1.4 on page 82), the recovery of this compound from carbon was tested. A microscale test column on 15 mg of 1a at 100:1 carbon ratio gave approximately 80% recovery. The loss was more likely due to the small scale than irreversible adsorption, since stripping the column in 50% toluene/petrol gave no additional 1a. Nonetheless, use of minimal carbon is clearly advisable to minimise any losses. Interestingly, the elution with toluene/petrol gave an orange fraction, presumably carotenoids. Chlorophylls, which are more strongly adsorbed than carotenoids,195 were not eluted. A number of practical points should be noted: Activated carbon powder can contain much ner particles than ash
78
CHAPTER 2. ISOLATION. chromatography grade silica; these can contaminate fractions and block glass frits. A layer of lter aid below the carbon prevents this. As noted by others,185 the carbon/lter aid mixture shrinks and cracks when dry, which may contaminate the decolourised fractions with crude extract. To prevent this, it is critical that the solvent level never be allowed to touch the surface of the carbon. Since it is dicult to locate the interface between the very dark crude extract and the black carbon, it is helpful to add a thick layer of sand on top. This layer should never be allowed to run dry. In later work196 ltration under vacuum proved more convenient and practical than ash chromatography. This preferred approach is shown in Figure 2.4. Fractions are collected by closing the separatory funnel tap and applying vacuum. When the desired amount has been collected, vacuum is released and the fraction dispensed. A gradient from 50% to 20% EtOAc/petrol gave faster elution and less tailing, without eluting pigments. All fractions containing terpenoids by TLC were pooled.
Sand Activated carbon/filter aid Filter aid Glass frit Vacuum adaptor
Separatory funnel
Figure 2.4: Apparatus for Filtration through Activated Carbon. While some separation between compounds in the decolourised fractions was
79
apparent (see Section 2.1.4.2 on page 85), separation of individual compounds proved unworkable due to tailing. Thus, since the pigments were not eluted and the eluted compounds were not separated, in practice the preferred procedure resembled ltration more than chromatography.
2.1.3.5
General Applicability of the Method.
This procedure has some attractive features for use in isolating biologically active compounds from plants. Activated carbon ltration provides a much greater reduction in mass and complexity of the extract than more common partitioning procedures. Thus, as seen with PCBs, analysis is simplied by utilising an extra dimension for separation (surface area/planarity), compared to the typical reliance on polarity dierences. Using the procedure described here, as discussed below, diterpenoids and triterpenoids were rapidly eluted, while the tetraterpene carotenoids were completely removed. Moreover, this added dimension (size) is of biological relevance. It is well established pharmacologically that large molecules show reduced absorption and permeation. This is embodied in Lipinskis famous rule of 5,197 which states among other things that compounds with a relative molecular mass over 500 will be less readily absorbed in vivo, and therefore less likely to show pharmacological activity. Thus, compounds eluted rst from activated carbon are more likely to be responsible for biological activity. Nonetheless, as indicated previously, size is not the sole determinant of adsorption. Compounds of interest may not be recovered using the procedure described here if they are highly planar, or strongly adsorbed for some other reason. Of particular relevance here are avonoids, some of which are known to be psychoactive. Apigenin (26a), isolated from S. ocinalis, binds weakly to the benzodiazepine site of the GABAA receptor (IC50 = 30 M).198 Other avones isolated from plants have shown much greater potency at this site; 27 binds with anity comparable to diazepam (K i = 6 nM).199 Interestingly, some avonoids appear to share the anxiolytic properties of benzodiazepines,
80
without the amnesic200 and sedating201 qualities. There has been no published report of avonoids from S. divinorum. However, in unpublished work, Valds isolated the known apigenin derivative 5-hydroxy7,4-dimethoxyavone (26b)202 from S. divinorum in 20 mg/kg yield.203 This compound has been isolated from several Salvia species; identity was conrmed by comparison of 1 H and 13 C NMR, IR, MS(CI) and mp with literature values.202 In other recent unpublished work, Claudio Medana detected a compound whose ESIMS data were consistent with quercitrin (23) or a stereoisomer in a crude extract of S. divinorum.204 The aglycone of 23, quercetin, activates several CNS receptors.205 Quercetin, but not 23 itself, has been proposed to have antidiarrhoeal eects;206 if so, this might conceivably contribute to the known eects of S. divinorum on gastric motility.207 However, quercetins gastrointestinal eects remain in dispute.208
OR RO O HO O OH O 26a 26b R H Me OH O 27 23 OH OH O O O HO O
2'
OH HO O O OH OH OH
Figure 2.5: Flavonoids. Flavonoids are not recovered using the isolation procedure described above. However, they are unlikely to play a detectable role in the plants eects. The binding anity of 26a is more than three orders of magnitude below that of 1a.23 Moreover, Valds isolated 1a in two orders of magnitude greater yield than 26b. In other plants, however, where avonoids or other strongly adsorbed compounds are responsible for activity, the above procedure would clearly be inadequate. Nonetheless, this obstacle is readily overcome. If bioassay-guided fractionation showed loss of activity in the decolourised extract, further elution with stronger eluents as outlined in Section 2.1.3.3 could be used. The
81
active fraction would then be greatly simplied relative to the initial extract, providing another dimension of separation, as with PCBs.
2.1.3.6
Other Decolourisation Adsorbents.
Although extremely cheap for laboratory use, use of activated carbon can be expensive on an industrial scale. For this reason, cheaper alternative adsorbents have been investigated. These are generally waste products: agricultural (leaves, straw, husks, pulp) or industrial (ash, slag).174 One adsorbent which has been used extensively is acid-activated clay.174 Activated clays (or activated earths) are widely used, for instance, to decolourise cooking oils.209 Although not as eective by weight as activated carbon,195 the lower cost makes industrial use attractive.
The adsorbent Tonsil used by Ortega et al in the original isolation22 of 1a is a type of activated clay used for decolourisation.209, 210 Unfortunately, this was not explained; the material was described (in a footnote) merely as bentonitic earth composed of silica (72.5%), alumina and other minerals.22 The body of the paper and the experimental procedure make no mention of pigments or colouration. Thus, readers may assume that Tonsil is simply a silica substitute. It is perhaps for this reason that no subsequent worker has tested this adsorbent. Nonetheless, Dr Ortega has conrmed that Tonsil was selected in order to decolourise the crude extract.211 He reports that the procedure was highly eective: the collected fractions were almost colourless. Tonsil was used for other isolations,212 but proved reactive towards certain functional groups, especially epoxides,213 and therefore unsuited to general use. Indeed, this substance has since found wide synthetic use as an acid catalyst.214 Activated carbon is milder and more generally applicable.
82
100 m
Figure 2.6: SEM image of salvinorin A crystals (blade morphology).39
2.1.4
2.1.4.1
Separation of Terpenoids.
Crystallisation of 1a.
TLC indicated that the major component of the decolourised extract (Section 2.1.3.4 on page 77) was 1a, which was obtained as ne colourless needles by crystallisation from MeOH. 1 H NMR showed other trace components 1%
each. Another crystallisation of the mother liquor gave additional 1a. The resulting mother liquor was no longer predominantly 1a by TLC. In other tests, EtOH was also an eective recrystallisation solvent (Lee et al subsequently used acetone).137 Serendipitously, trituration in Et2 O also proved eective for isolation of 1a from the decolourised extract. In early tests, a chromatographic fraction rich in 1a was dissolved in Et2 O for transfer; an amorphous, aky precipitate formed instantly. The ltrate proved, on further chromatography, to be devoid of 1a. The precipitate consisted of 1a which, despite its inferior appearance, was highly pure by 1 H NMR. Thus, trituration in cold Et2 O represents a promising alternative to recrystallisation from boiling alcohols easier, safer and (more signicantly) much faster. Moreover, the risk of decomposition in heated solutions (Section 2.1.1 on page 67) is avoided. This procedure, which was unfortu-
83
nately given insucient emphasis in our initial publication,215 clearly deserves further exploration. The ne needles of 1a were examined by scanning electron microscopy (SEM) (Figure 2.6). Shorter prisms were also observed, as well as plates or akes (Figure 2.7), and close examination of crystals showed small defective, semicrystalline particles adhering.
84
200 m
200 m
20 m
20 m
20 m
20 m
Figure 2.7: Other salvinorin A crystal morphologies (stereoview).39
2.1. ISOLATION PROCEDURE. 2.1.4.2 Chromatography.

O O O
85
O RO
O O R2
OR1 H
O O
R1
H R2
O R 1a 1b Ac H
O R1 1c 1d 1e 1f Ac Ac H H R2 OAc OH OAc H
OR3 R1 28a 28b 28c 29a OH OH H H R2 H OH OAc H R3 H Me H H
OH 30 H HO H H
H OH O 31
OH
32
OH
33
Figure 2.8: Terpenoids isolated from S. divinorum. The mother liquor from crystallisation of 1a was an unexpectedly complex mixture. Although at the time only four compounds had been reported from S. divinorum, TLC showed many more. The spots were poorly resolved, making a precise count impossible. Valds et al had reported27 that 1a and 1c were resolved in EtOAc/hexanes, but not in MeOH/CHCl3 ; this information proved invaluable. 2D TLC using these systems gave much greater resolution than either system alone. Many compounds not resolved by one system were
86
hRf (10% acetone/CH2Cl2) 80
33 70 31 60 1a 50 1c 30
1e 1f 1b 32 28b 1d 28c
29a
40
30
20 28a 10 10 20 30 40 50 60 70 80 hRf (70% Et2O/petrol)
Figure 2.9: TLC data of isolated compounds. resolved by the other. These dierences in selectivity proved to be general for petrol-based versus chlorinated solvent-based systems. After experimentation with various petrol-based systems, Et2 O gave similar resolution of 1a and 1c to EtOAc, but better separation from other compounds. Similarly, in CH2 Cl2 -based systems, acetone resolved 1a and 1c from other compounds better than MeOH. Also, these new systems spread the mixture over a wider range, from near the baseline to near the solvent-front. These systems were therefore adopted for general use. Extensive column chromatography yielded six new diterpenoids in addition to 1a1c: salvinorins DF (1d1f ) and divinatorins AC (28a28c, Figure 2.8 on the previous page). Five known terpenoids were also isolated: ()-hardwickiic acid (29a), (E)-phytol (30), oleanolic acid (31), presqualene alcohol (32) and peplusol (33).

Dried leaves (860 g) Powder, steep in acetone ( 3) Crude extract (30.5 g) Column on activated carbon acetone to petrol gradient Decolourised extract (5.7 g) 1a (2.6 g) Combine Recrystallise (MeOH, EtOH)
87
Column on silica 5-50% acetone/CH2Cl2
Series A 656 mg
Series B 150 mg Column 70-90% Et2O/petrol
Series C 359 mg
Series D 77 mg Columns
Column 50-80% Et2O/petrol 1c (219 mg) 1a (2.9 g)
1b (13 mg)
Triturate (Et2O) 1d (75 mg)
28a (36 mg)
29a (7 mg)
Combine Column 60-100% Et2O/petrol
Fraction C1 55 mg
Fraction C2 119 mg
Fraction C3 57 mg Columns
1d (114 mg)
Column 60-100% Et2O/petrol Columns 28c (23 mg) 28b (41 mg) Combine 31 (3 mg) 1e (3 mg)
HPLC 60% EtOAc/petrol 1f (1 mg)
Figure 2.10: Isolation of terpenoids from commercial S. divinorum.
Figure 2.9 on the facing page shows the TLC data for all compounds isolated in this work. The most dramatic dierences between the two systems are
88
Dried leaves (224 g)
Powder, steep in acetone ( 3) Crude extract (7 g) Vacuum filtration on activated carbon 50-20% EtOAc/petrol
Series E 97 mg Columns on silica 33 (1 mg) 32 (23 mg)
Series F 279 mg
Recrystallise (MeOH 2)
30 (12 mg)
1a (126 mg)
Figure 2.11: Isolation of terpenoids from Australian S. divinorum. emphasised by dashed lines. For instance, separation of 1c from 28b or of 1a from 1f is not possible in Et2 O/petrol, but trivial in acetone/CH2 Cl2 . The converse is true for separating 1f from 28c or 1a from 30. With these solvent systems in place, separation of the components of the mother liquor was straightforward, albeit time-consuming. Repeated column chromatography, alternating between Et2 O/petrol and acetone/CH2 Cl2 , resolved most compounds. In deciding which fractions to pool, TLC analysis in each of the main solvent systems proved invaluable. For particular separations, other solvent mixtures were sometimes used when TLC indicated better resolution. The overall procedure, as applied to the extraction of commercial dried S. divinorum leaves, is summarised in Figure 2.10 on the previous page. For more detail, see Experimental Section 5.2.1 on page 185. Yields expressed in mg/kg are given in Table 5.1 on page 187. A similar procedure was used on a sample of Australian-grown S. divinorum leaves (Figure 2.11), resulting in the isolation of several additional compounds. Partial separation of these compounds was achieved in the initial ltration through activated carbon, but the severe tailing exhibited by all compounds
89
prevented full separation. Since chromatography on silica was ultimately required, attempts to separate compounds on carbon are not recommended.
2.1.4.3
Crystallisation of new compounds.
As with 1a, trituration in Et2 O proved eective for purication of 1d. Attempted dissolution of a fraction containing 33% 1d in boiling Et2 O gave white crystals of the pure compound. Also, as reported by Tidgewell et al,169 trituration in cold MeOH is eective for purication of 1b. The ease of crystallising these compounds varies widely. Evaporation of a solution of 1b gave crystals irrespective of the solvent. 1a and 1d gave crystals from some solvents, but an amorphous resin from others (especially CH2 Cl2 and CHCl3 ). Although Valds reported the crystallisation of 1c,27 attempts to replicate this were unsuccessful. Attempts to recrystallise the other new compounds, under a variety of conditions, were also unsuccessful. On one occasion, 28b gave crystals from cold CH2 Cl2 on addition of petrol, but repeated attempts to replicate this were unsuccessful.
2.1.4.4
Losses during isolation.
The recoveries of these compounds do not accurately reect their actual proportions in the crude extract. The main objective in this work was to obtain the maximum number of pure compounds; yield was a much lower priority except in the case of 1a. Traces of one compound were often discarded during the isolation of another. In the most extreme case, all of the 29a isolated was derived from a small portion (13%) of the crude extract. The actual content of this compound in the crude extract, therefore, was much higher than the reported yield. There were also losses due to other causes. In one case (separation of 1e and 1f ), despite good separation by TLC (Figure 2.9 on page 86), repeated column chromatography gave incomplete resolution. Ultimately, HPLC gave
90
full resolution, but unfortunately the recovery after this lengthy process was very poor. From a 34 mg fraction containing only 1e and 1f by TLC and 1 H NMR, the nal combined recovery of these two compounds was only 4 mg. Similarly, yield of 1b (15 mg/kg) was very low compared to earlier work (75 mg/kg).23 This was probably largely due to losses in chromatography. During subsequent synthetic work, it became apparent that column chromatography of 1b in petrol-based systems gave very poor recoveries. Stripping the column in acetone/CH2 Cl2 gave an improvement, but full recovery was only achieved using MeOH/CH2 Cl2 . Additionally, recrystallisation of 1b-rich fractions from MeOH gave poor recoveries, and TLC showed extensive decomposition. As noted in Section 2.1.1 on page 67, this decomposition also occurred with 1a.
2.2
Structure Elucidation.
2.2.1
Revised NMR Assignments for Salvinorin A (1a).
The published NMR assignments23 of 1a (based on decoupling experiments) were veried on the basis of 800 MHz 1 H NMR (Figure 2.12 on the next page), along with HSQC, HMBC, DEPT and nOe experiments. These new data supported the original 1 H and
13
C assignments in all cases except H-6 and -7.
The HSQC spectrum showed cross peaks from C-6 ( 38.1 ppm) to multiplets at 1.57 & 1.78 ppm, and from C-7 ( 18.1 ppm) to 1.63 & 2.15 ppm (Figure 2.13 on the facing page). The 800 MHz 1 H NMR spectrum also gave well-resolved rst-order multiplets for H-6, 7 and 11, which overlap to form a single multiplet ( 1.54 1.68 ppm) at 400 MHz. Resolution enhancement revealed all coupling constants in these multiplets, allowing unambiguous assignment of all peaks (Figure 2.14 on page 92).
2.2. STRUCTURE ELUCIDATION.

7.402 7.384 6.369 5.521 5.138 3.724 2.745 2.504 2.304 2.178 2.163 2.074 1.792 1.642 1.574 1.447
91
1.113
1.56 ppm
O O
O O
1a
2.35
2.33
2.31
2.29
1.66
1.64
1.62
1.60
1.58
2.76 6.37 6.36
2.74
ppm
2.18
2.16
2.14 1.80 1.78 ppm
7.40 7.39
ppm 5.53 5.52 5.51
5.16 5.15 5.14 5.13 ppm
2.52 2.51 2.50 2.49 2.06
7.5
1.63 1.59
7.0
6.5
0.82
6.0
5.5
0.85 1.00
5.0
4.5
4.0
3.25
3.5
3.0
1.02
2.5
0.88 2.27
2.0
5.30 1.54
1.5
1.51 1.28 1.94 2.96
ppm
2.70
Figure 2.12: 1 H NMR spectrum of 1a (800 MHz, CDCl3 ).
C-7
80 90 100 110 120 130 140 7.4 7.2 7.0 6.8 6.6 6.4 6.2 6.0 5.8 5.6 5.4 5.2
20
30
C-6
40
50
60
3.6
3.4
3.2
3.0
2.8
2.6
2.4
2.2
2.0
1.8
1.6
1.4
1.2
ppm
Figure 2.13: HSQC spectrum of 1a (800 MHz, CDCl3 ).
92
NOESY NOESY (REVISED ASSIGNMENT) 143.7 139.4 7.41 dt 125.2 1.57 ddd 20.5 72.0 2.50 dd 43.3 5.52 ddd 169.9 5.14 ddt 202.0 2.74 dd 30.9 53.5 2.31-2.28 m 16.4 171.5 1.11 s 1.79 dt 2.18 br s 35.4 75.0 64.0 15.2 1.58 td 42.0 38.1 51.3 1.45 s 18.1 2.17-2.14 m 1.64 tdd 2.31-2.28 m 2.07 dd 169.9 171.1 20.5 2.16 s 7.38 t
H C
6.37 dd 2.16 s
H C
(REVISED ASSIGNMENT) 143.7
7.38 t
108.3
108.3 6.37 dd 125.2 1.57 ddd 72.0 2.50 dd 43.3 5.52 ddd 5.14 ddt 202.0 2.74 dd 30.9 53.5 16.4 171.5 1.11 s 2.18 br s 35.4 75.0 64.0 15.2 1.58 td 42.0 38.1 51.3 1.45 s 18.1 2.07 dd
139.4 7.41 dt
171.1
2.17-2.14 m 1.64 tdd
1.79 dt
51.9 3.72 s
51.9 3.72 s
Figure 2.14: Revised NMR assignments for 1a (stereoview).
2.2.2
Revised NMR Assignments for Salvinorin C (1c).
The published structure and NMR assignments27 of 1c were veried on the basis of 1 H NMR (Figure 2.15 on the facing page), decoupling, HMQC, HMBC and DEPT experiments. These data were consistent with the published structure, but the assignments of H-19, H-20, C-17 and the acetates were revised as shown in Figure 2.16 on the next page. Contrary to the original assignments, the HMQC spectrum showed strong cross peaks between C-19 (21.8 ppm) and H-19 (1.71 s), as well as C-20 (15.7 ppm) and H-20 (1.21 s) - see Figure 2.17a. This was conrmed by HMBC cross peaks from H-19 to C-4 (142.6 ppm), and H-20 to C-11 (44.2 ppm) - Figure 2.17b. HMBC data also allowed the assignment of the acetate signals. The C-2O-acetate carbonyl (169.8 ppm), originally assigned to C-17,27 showed cross peaks to H-2 (5.54 ppm) and 2.04 ppm; its C-1-O- counterpart (170.5 ppm) showed cross peaks to H-1 (5.75 ppm) and 2.12 ppm - Figure 2.17c,d.
2.2.3
Other Known Diterpenoids.
In addition to 1a-1c, ve known compounds not previously reported from S. divinorum were isolated. The data and sources used to identify them are given
93
2.037
1.716
3.740
7.4
7.260
2.2
2.1
O
1.8
1.7
1.3
1.2
1.1
O O O
2
O
1
11
H
20
O
17
O
19
1c
6.5
6.4 5.8 5.7 5.6 5.5 2.6 2.4
7
0.67 0.70 0.78 0.88
6
1.97 1.00
4
3.00
3
1.03 1.00
2.122
ppm
5.06 1.14 1.17 4.54 2.95 4.43
Figure 2.15: 1 H NMR spectrum of 1c (400 MHz, CDCl3 ).
7.42 t HMBC 143.9 HMBC
7.42 t 143.9
C C/H
2.04 s
H
6.41 dd 108.4 139.4 7.45 m 125.4 1.68 dd 2.48 dd 5.54 dd 44.2 1.49 br s 64.1 132.4 170.5 69.2 37.2 51.8 1.23 td15.7 38.1 142.6 165.7 21.8 1.71 s 51.8 3.74 s 37.0 2.59 dt 1.85-1.75 m 52.6 1.21 s 18.4 2.17-2.10 m 6.46 dd 71.8 5.53 dd 2.13 dd 171.4 2.04 s
C C/H
H
6.41 dd 108.4 139.4 7.45 m 125.4 1.68 dd 2.48 dd 5.54 dd 44.2 71.8 5.53 dd 171.4
(REVISED ASSIGNMENT)
(REVISED ASSIGNMENT)
20.7 169.8
5.75 br d
20.7 169.8
5.75 br d 1.49 br s 69.2
2.13 dd 37.2 52.6
64.1 132.4 170.5 21.1 2.12 s
6.46 dd 21.1 2.12 s
51.8 1.23 td 15.7 38.1 37.0 2.59 dt 21.8 1.71 s
1.21 s 18.4 2.17-2.10 m 1.85-1.75 m
142.6 165.7
51.8 3.74 s
Figure 2.16: Revised NMR assignments for 1c (stereoview).
in Table 2.2 on the following page.
1.216
94
H (ppm) 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 23 22 21 20 19 18 17 16 15
a) HMQC
11 .5 12 .0 12 .5 13 .0 13 .5 14 .0 14 .5 15 .0 15 .5 16 .0 16 .5 17 .0 17 .5 10 4 10 3 10 2 10 1 0 10 9 0 8 0 7 0 6 0 5 0 4 0
b) HMBC
1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12 2.14 2.16 2.18 2.20
c) HMBC
55 .0 55 .5 56 .0 56 .5 57 .0 57 .5 58 .0 7. 125
d) HMBC
171.5
170.5
169.5
168.5
120 7.
115 7.
110 7.
105 7.
100 7.
195 6.
190 6.
185 6.
C(ppm)
Figure 2.17: HMQC and HMBC spectra of 1c (400 MHz, CDCl3 ). Compound 29a 29b* 30 31 32 33
* c cs d sa sp
H NMR sa, d216, 217 sa, d218 sp,219 d220 d221 d,222 sp223 d224
13
C NMR
FTIR d218 d219, 220 d222 d224
[]D d218, 217
HRESIMS
EIMS d217
d218 d219, 220 d221 d223 d224
TLC cs cs
c d223 d224
29a was methylated with CH2 N2 for full characterisation. consistent with calculated value. cospotted with authentic material in petrol- and CH2 Cl2 -based systems. consistent with published data. superimposable with spectrum of authentic material. superimposable with published spectrum.
Table 2.2: Data and sources used to identify known compounds. 2.2.3.1 ()-Hardwickiic acid (29a).
()-Hardwickiic acid (29a) was rst isolated from Hardwickia pinnata by Misra et al,225 and the structure (including absolute stereochemistry) elucidated by

O
95
H R 29a 29b H Me
OR
Figure 2.18: Single-crystal X-ray structure of 29a (stereoview).
spectroscopic methods and extensive degradation studies.226, 227 Recently, Xray crystallography has conrmed this structure (Figure 2.18).228 29a has since been isolated from many other plant species, including Salvia regla.229 29a has been reported to exhibit insecticidal230 and antimicrobial217 activity. Interestingly, its enantiomer (+)-hardwickiic acid (ent-29a) has been isolated from Copaifera ocinalis 231 and related species.232 For denitive identication, authentic ent-29a was isolated as the methyl ester ent-29b from commercial Copaiba balsam.218 The material from S. divinorum, after methylation to give 29b, cospotted by TLC and gave a superimposable 1 H NMR spectrum, but the opposite sign of optical rotation.
96
30
OH
Figure 2.19: (E)-Phytol. 2.2.3.2 (E)-Phytol (30).
(E)-Phytol (30) is ubiquitous in photosynthetic organisms, as the sidechain of the chlorophylls (eg. 21 on page 71). 30 displays anticancer activity,233 and may play a role in the health benets of green-yellow vegetables.234 30 also exhibits antiplasmodial,235 antimycobacterial236, 237 and antiteratogenic238 activities.
2.2.4
2.2.4.1
Known Triterpenoids.
Oleanolic Acid (31).
H OH H HO H O 31
Figure 2.20: Oleanolic acid. Oleanolic acid (31) occurs commonly in plants, including many Salvia species.239 It exhibits an extremely wide range of potential therapeutic activities, which have been reviewed elsewhere:240, 241, 239 anti-inammatory, hepatoprotective, gastroprotective, cardiovascular, hypolipidemic and immunoregulatory eects. Of more signicance is anti-tumour activity, which appears to occur via several independent mechanisms (inhibition of tumourigenesis, inhibition of tumour promotion, and induction of tumour cell dierentiation).240 Similarly, 31 is active against HIV-1 by two mechanisms: inhibition of HIV protease242 and reverse transcriptase243 enzymes. The compound also exhibits activity against
97
many other pathogens: herpes simplex virus,244 a range of bacteria245, 246 (including mycobacteria)247 and Leishmania spp,248 as well as being an insect antifeedant.249
2.2.4.2
Presqualene Alcohol (32) and Peplusol (33).

H
32 OH
OH
33
Figure 2.21: Presqualene alcohol and peplusol. The diphosphate of presqualene alcohol (32) is a precursor in the biosynthesis of sterols (including cholesterol), and has therefore been the subject of intense synthetic223 and biosynthetic250 interest. The isolation of 32 from a plant is unusual. The closely related compound peplusol (33) was isolated from Euphorbia peplus by Giner et al.224 In a paper submitted a year later, Connolly et al reported the same compound from E. lateriora, naming it anhydrobisfarnesol.251 The earlier paper by Giner et al was mentioned, but only in a footnote; priority was claimed252 based on a preliminary communication.253 That communication, however, contained no experimental or characterisation data, and the absolute conguration was not determined even in the full paper. Additionally, the ambiguous semisystematic name anhydrobisfarnesol leaves regio- and stereochemistry unspecied. Thus, Giner et al clearly have priority, and their name is used here. Connolly and co-workers conrmed the proposed structure by synthesis, obtaining the racemic compound in 6% yield over six steps.251 However, they failed to cite a clearly superior synthesis (50% over two steps)254 published fteen years earlier. Connolly has proposed 33 as a biosynthetic precursor to 32 itself.252
98
It is interesting to note that 33 and 32 are approximately equal in size (22carbon chain) to the carotenoid 22 (p.71, 24 carbons), but were easily eluted from activated carbon in the above chromatography procedure. Since, unlike 22, these compounds are not conjugated, this again conrms the strong inuence of planarity on adsorption.
2.2.5
2.2.5.1
Salvinorins D-F (1d-1f).

Salvinorins D and E (1d and 1e).
Inspection of the 1 H NMR spectra of 1d-1f suggested that they were derivatives of 1c. Compared to 1c, the 1 H NMR spectra of 1d (Figure 2.22) and 1e (Figure 2.26 on page 102) showed only one acetyl peak, and gained one new peak ( 2.01 in 1d; 1.94 in 1e) which exchanged with D2 O. The presence of a hydroxyl group was conrmed by IR spectroscopy (3475 cm1 in 1d; 3510 cm1 in 1e). HRESIMS established the molecular formula of both compounds as C23 H28 O8 , consistent with the loss of one acetyl group. These data indicated that the compounds were monoacetates of 1c. Relative to 1c, the H-2 signal of 1d was shifted upeld from 5.55 to 4.44, and coupled to the hydroxylic proton (2.01 ppm) as well as to H-1 (5.70 ppm) and H-3 (6.54 ppm). The HMBC spectrum (Figure 2.25 on page 101) showed correlations between H-2 and C-4 (141.2 ppm), as well as H-1 and C-5 (37.6 ppm). These quaternary carbons, not apparent in the HMQC (Figure 2.24 on page 100) and DEPT spectra, were in turn located by HMBC correlations. This established the structure of 1d as 2-deacetylsalvinorin C. Relative to 1c, the H-1 signal of 1e was shifted from 5.76 to 4.46. Irradiation sharpened the H-10 singlet. The HMBC spectrum (Figure 2.29) again showed a correlation between H-2 (5.40 ppm) and C-4 (143.4 ppm), not apparent in the HMQC spectrum (Figure 2.28). Thus 1e is 1-deacetylsalvinorin C, as shown. The quaternary carbon at 169.8 ppm was incorrectly assigned to C-17 in our original publication;215 this has been reassigned on the basis of its

O O O HO
2 3 1 5 4
99
O O
1d
2.150
3.738
7.5
7.4
2.2
2.2
2.0
1.8
1.6
7.260
5.7
5.6
5.5
1.3 1.2 1.1
6.6
6.5
6.4
4.5
4.4
2.6 2.5
7
0.83 0.85 0.87 0.92
6
0.99 0.98
5
1.00
4
3.00
3
2.00
1.686
ppm
4.43 4.98 1.05 0.99 0.97 4.07
Figure 2.22: 1 H NMR spectrum of 1d (400 MHz, CDCl3 ). HMBC crosspeak with the acetate protons (see Figure 2.27 on page 102).
1.225
100
7.42 t J (Hz) 143.8 J (Hz)
7.42 t 143.8
H
108.4 HMBC 6.40 dd 7.44 s 125.4 1.64 ddd 2.54 dd 4.44 ddd 5.70 dt 43.9 1.42 br s 68.6 1.3 Hz 135.7 171.6 66.5 37.0 51.8 52.5 1.20 td15.6 37.6 141.2 37.0 2.56 dt 166.2 21.6 1.69 s 1.78 dtd 18.3 1.22 s 2.17-2.09 m 2.15 s 139.4
H
108.4 HMBC 6.40 dd 7.44 s 125.4 1.64 ddd 2.54 dd 4.44 ddd 5.70 dt 43.9 139.4
71.6 5.53 dd
71.6 5.53 dd 171.6
2.01 br d 2.4 Hz
2.13 dd
171.6 2.4 Hz
2.01 br d 68.6 1.3 Hz 6.54 dd 135.7 171.6 21.2 166.2 66.5
1.42 br s 37.0
2.13 dd 51.8
6.54 dd 2.15 s
52.5 1.20 td 15.6 37.6 141.2 37.0 2.56 dt 21.6 1.69 s 18.3
1.22 s 2.17-2.09 m
21.2
1.78 dtd
51.7 3.74 s
51.7 3.74 s
Figure 2.23: NMR assignments and key 2D correlations for 1d (stereoview).
H (ppm) 4.5 5.0 5.5 6.0 6.5 7.0 7.5 140 130 120 110 100 C(ppm) 90 80 70
H (ppm)
1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 50 45 40 35 30 25 20 15
C(ppm)
Figure 2.24: HMQC spectrum of 1d (400 MHz, CDCl3 ).
101
H (ppm)
1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 160 140 120 100 80 C(ppm) 60 40 20
4.36 4.40 4.44 4.48 4.52 4.56 143 142 141 140 139 138 137 136 135 134 5.45 5.50 5.55 5.60 5.65 5.70 5.75 45 44 43 42 41 40 39 38 37 36
Doublets resulting from direct coupling (1 JCH ) have been removed digitally.
Figure 2.25: HMBC spectrum of 1d.
102
O
O O
OH H
O O
1e
1.726
2.2 7.4
2.1
1.9 1.8 1.7 1.6
1.3
1.2
1.1
5.6
5.5
5.4
6.4
4.5
4.4 2.6 2.5 2.4
7.0
0.84 0.91
6.5
0.92 0.93
6.0
5.5
1.05 1.03
5.0
4.5
1.06
4.0
3.5
3.02
3.0
2.5
2.08
2.0
1.553
7.260
3.729
2.170
1.467
ppm
5.19 0.97 3.16 1.53 0.87 1.18 1.20 3.12
Figure 2.26: 1 H NMR spectrum of 1e (400 MHz, CDCl3 ).

7.42 t J (Hz) 143.9 J (Hz) 143.9 7.42 t
H
HMBC
H
(REVISED ASSIGNMENT) 6.41 dd 125.8 1.62 dd 2.46 dd 5.40 dd 71.7 2.17 s 108.4 139.3 7.44 br s HMBC
H C C
H
(REVISED ASSIGNMENT) 6.41 dd 125.8 1.62 dd 2.46 dd 5.40 dd 71.7 108.4 139.3 7.44 br s
C C
2.17 s
21.0 169.8 2.4 Hz 6.43 dd 72.3 1.6 Hz 131.5
4.46 ddd 64.3 1.30 br s 37.5 54.0 1.19 td16.2 37.8
44.4 5.60 ddd 2.18-2.07 m 51.7 1.47 s 18.4 2.18-2.07 m 1.84 dtd 171.8
21.0 169.8 2.4 Hz 1.6 Hz 6.43 dd 131.5 72.3
4.46 ddd 64.3 1.30 br s 37.5
44.4 5.60 ddd 2.18-2.07 m 51.7 1.47 s 18.4 2.18-2.07 m 1.84 dtd 171.8
1.94 dd 143.4 166.0
54.0 1.19 td 16.2 1.94 dd 37.8 143.4 37.0 2.52 ddd 21.9 1.72 s
37.0 166.0 2.52 ddd 21.9 1.72 s
51.8 3.73 s
51.8 3.73 s
Figure 2.27: NMR assignments and key 2D correlations for 1e (stereoview).
103
H (ppm) 4.5 5.0 5.5 6.0 6.5 7.0 7.5 140 130 120 110 100 C(ppm) 90 80 70
H (ppm)
1.4 1.5 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 55 50 45 40 35 30 25 20 15
C(ppm)
Figure 2.28: HMQC spectrum of 1e.
H (ppm) 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 170 165 160 155 150 145 140 135 130 125 165 160 155 150 145 140 135 130 125 120 115 110 5.4 5.6 5.8 6.0 6.2 6.4 6.6 6.8 7.0 7.2 7.4 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 70 65 60 55 50 45 40 35 30 25 20 15 C(ppm)
Figure 2.29: HMBC spectrum of 1e.
104 2.2.5.2
CHAPTER 2. ISOLATION. Interconversion of 1c-1e via Diol 1h.

O
R 2O
OR1 H
O O
a a
R2 R1 1c 1d
a
Ac Ac H H Ac
b
1e 1h
Ac H H
a) Ac2 O, pyridine, DMAP b) Na2 CO3 , CH2 Cl2 /MeOH (1:1), rt.
Scheme 2.1: Interconversion of 1c-1h.

3.731
1.603
7.260
(+D2O)
2.50 2.45 2.40 2.35 2.30 2.25 2.20 2.15 2.10 2.05
7.5
7.4 (+D2O)
5.7
5.6
1.90 1.85 1.80 1.75 1.70 1.65 1.60 1.55
(+D2O)
1.20 1.15 1.10
6.5
6.4
4.4
4.3
4.2
7
0.78 0.78 0.82 0.83
6
0.95
5
1.01 1.03
4
3.00
1.702
1.476
ppm
1.95 1.15 3.18 0.99 2.06 2.14 3.16 1.31
Figure 2.30: 1 H NMR spectrum of 1h (400 MHz, CDCl3 ). To conrm the structures of 1c-1e, diol 1h was prepared by deacetylation of 1c. Tidgewell et als standard conditions for deacetylation of 1a (Na2 CO3 in minimal MeOH)169 caused extensive epimerisation (see Section 3.2.4). However, epimerisation could be suppressed by addition of CH2 Cl2 , aording 1h

7.42 t J (Hz) 143.9 J (Hz) 143.9 7.42 t
105
H
HMBC 6.40 dd 108.4 139.3 7.43 br s 125.8 1.60 ddd 2.49 dd 4.28 dd 71.8
H
HMBC 6.40 dd 108.4 139.3 7.43 br s 125.8 1.60 ddd 2.49 dd 4.28 dd 71.8
NOESY
NOESY
2.40-2.30 m 69.6 2.4 Hz 6.48 dd 135.3
4.32 br d 1.22 d 65.6 2.40 2.30 m 142.2 37.5 54.1
44.4 5.60 dd 2.14-2.10 m 172.0 51.8 1.47 s 18.4 2.09 dq
2.40-2.30 m 69.6 2.4 Hz 6.48 dd 135.3
4.32 br d 1.22 d 65.6 37.5
44.4 5.60 dd 2.14-2.10 m 172.0 51.8 1.47 s 18.4 2.09 dq
16.3 1.16 tdd 37.5 37.0
166.6 22.1 51.7 3.73 s 1.70 s
1.82 dtd 2.50-2.45 m
2.40 - 54.1 2.30 m 16.3 1.16 tdd 37.5 37.0 142.2
1.82 dtd
166.6 22.1 51.7 3.73 s 1.70 s
2.50-2.45 m
Figure 2.31: NMR assignments and key 2D correlations for 1h (stereoview).

H (ppm) 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 130 110 90 70 50 30 10 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 170 165 160 155 150 145 140 C(ppm)
a) HMQC
b) HMBC
Figure 2.32: HMQC and HMBC spectra of 1h.

4.3 2.1 4.4 4.5 4.6 2.2 4.7 4.8 2.3
nOe effect direct coupling
4.9 5.0 5.1 5.2 5.3 5.4 5.5 5.6
2.4
2.5 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1
2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 ppm
Figure 2.33: NOESY spectrum of 1h. from 1c or 1d in up to 86% yield. K2 CO3 was equally eective. Analysis of the NMR data (Figures 2.30 on the facing page and 2.32) was straightforward. Carbons C-17 and -18 were distinguished on the basis of
106
HMBC correlations (Figure 2.32), and H-1 and 2 on the basis of coupling constants to H-3 (Figure 2.31 on the preceding page). These couplings were only resolved after D2 O exchange. The relative stereochemistry was conrmed using NOESY data (Figure 2.33 on the previous page). H-1 showed cross peaks to H-2, -10 and -11, while H-8 correlated to H-10 and -11, and H-12 correlated to H-20 (Figure 2.31 on the preceding page). This conrms the congurations shown for 1h, and therefore of 1c-1e, through the interconversions summarised in Scheme 2.1 on page 104. The proposed structure of 1d was veried by acetylation, which proceeded smoothly to give 1c, identical in all respects with the isolated material - TLC cospot, 1 H and
13
C NMR, IR, and optical rotation (Scheme 2.1 on page 104).
Insucient 1e was available for acetylation. In our initial publication,215 we assumed that 1e would be too hindered for direct acetylation under standard conditions. This proved incorrect; the 1-hydroxy group of 1h is readily acetylated. However, the crude product after three hours consisted of 1d and 1e in approximately 1:2 ratio, with only traces of 1c; thus, the 1-position of 1h is (as expected) less reactive than the 2-position, and each of the monoacetates is less reactive than the diol. At 45 C, the starting material was consumed in approximately 1 hour, giving a mixture of 1c and 1e, conrming that 1e is the less reactive of the monoacetates. It also appears to be less stable; after acetylation under these standard conditions, it was contaminated by an inseparable impurity, apparently an isomer other than the 8-epimer. This did not occur with 1d. 1e was also much more prone to decomposition in storage than 1d. During the isolation of 1c, Valds detected traces of what appeared to be 1d, 1e, and 1h in S. divinorum;255 while we subsequently conrmed the presence of 1d and 1e,215 no-one has yet isolated 1h. It may be that the compound occurs in the plant in very low levels; this would be expected if its formation were the rate-limiting step on the path to 1c1e. Alternatively, current isolation procedures may give poor recoveries. Indeed, like 1b, 1h precipitates when loaded on silica gel in most solvent systems. Stripping the column with
2.2. STRUCTURE ELUCIDATION. MeOH/CH2 Cl2 was required to achieve satisfactory recovery.
107
A sample of 1h was supplied to Claudio Medana for LCMS comparison with the crude extract. This proved dicult, since 1b and 1h were not resolved by reverse-phase HPLC, either C-18 functionalised silica or polymeric C-18. The compounds were ultimately resolved on a porous graphitic carbon column, eluting with a MeOH to CH2 Cl2 gradient. Under these conditions, a crude CH2 Cl2 extract showed no detectable 1h.256
2.2.5.3
Salvinorin F (1f).
The molecular formula of compound 1f, C21 H26 O6 , was established by HRESIMS. In contrast to 1d and 1e, 1 H NMR (Figure 2.34 on the next page) showed no acetyl peak, only one oxymethine signal, and a diastereotopic methylene ( 2.35 and 2.60) coupling to H-1 and H-3 (Figure 2.35 on page 109). This implied the 2-deoxy structure shown. Protons H-2 and H-2 were distinguished by their coupling constants: molecular modelling257 predicted J 1,2 = 5.4 Hz (observed: 5.5 Hz) and J 1,2 = 1.5 Hz (observed: 1.1 Hz). In our initial publication,215 this latter coupling was not observed, but subsequent resolution enhancement allowed direct measurement. H-2 also showed the expected HMBC cross peaks to H-4 and -10 (Figure 2.37 on page 109).
108
OH H
2 3 4 1 10
O O
1f
1.556
7.260
2.6 7.5 7.4
2.5
2.4
2.3 1.8 1.7 1.6
2.2
2.1
6.7
6.6
6.5
6.4
3.720
1.3
1.2
1.1
4.6 5.7 5.6 5.5
4.5
4.4
7
0.82 0.85 0.90 0.97
6
1.01
5
1.06
4
3.02
1.707
1.476
ppm
2.08 2.16 3.07 1.04 1.01 1.05 1.02 3.19 1.28 1.11 1.31
Figure 2.34: 1 H NMR spectrum of 1f (400 MHz, CDCl3 ).

7.42 t J (Hz) 143.8 J (Hz) 7.42 t 143.8
109
H
HMBC 6.41 dd 108.4 139.3
H
7.44 br s 125.9 1.62 ddd 5.5 Hz 2.46 dd 4.51 br dd 1.25 br s 37.7 63.9 44.4 5.60 ddd 2.17-2.08 m 172.1 52.2 1.48 s 18.6 2.17-2.08 m 2.35 ddt 71.7
H
HMBC 6.41 dd 108.4 139.3 7.44 br s 125.9 1.62 ddd 5.5 Hz 2.60 ddd 1.1 Hz 38.0 133.4 2.46 dd 44.4 5.60 ddd 4.51 br dd 1.25 br s 37.7 63.9 2.17-2.08 m 52.2 1.48 s 18.6 2.17-2.08 m 1.82 dtd 172.1 71.7
2.60 ddd 1.1 Hz 2.35 ddt 38.0 133.4 6.67 ddd
54.8 16.4 1.18 tdd 1.29 dd 140.6 36.6 37.3 2.53 dt 21.6 1.71 s
54.8 1.18 tdd 16.4 1.29 dd 140.6 36.6 37.3 2.53 dt 21.6 1.71 s
6.67 ddd
166.9
1.82 dtd
166.9
51.5 3.72 s
51.5 3.72 s
Figure 2.35: NMR assignments and key 2D correlations for 1f (stereoview).
H (ppm)
4.5 5.0 5.5 6.0 6.5 7.0 7.5 140 130 120 110 100 90 80 70 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8
C(ppm)
55
50
45
40
35
30
25
20
15
Figure 2.36: HMQC spectrum of 1f.
H (ppm)
1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 160 150 140 130 120 110 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5
C(ppm)
70
65
60
55
50
45
40
35
30
25
20
15
Figure 2.37: HMBC spectrum of 1f.
110
2.2.6
2.2.6.1
Divinatorins A-C (28a-28c).

Structure Elucidation.
O
R1
H R2 28a 28b 28b 29a
R1 OH OH H H
R2 H OH OAc H
R3 H Me H H
18
OR3
Figure 2.38: Divinatorins AC (28a-28c) and hardwickiic acid (29a).

O
OH H
OH
28a
7.3
7.2
2.6
2.5
2.4
2.1 2.3
2.0
1.9
1.8 1.72 1.70 1.68 1.66
6.92 6.90 6.88

4.51 4.50 4.49 4.48 4.47
1.6
1.5
1.4
1.3
1.2
6.26 6.25
0.87 0.88 0.99 0.92 1 1 3.1 1.1 4.2 2.1 1.1 2.1 3.2 1.1 3
1 ppm
Figure 2.39: 1 H NMR spectrum of 28a (400 MHz, CDCl3 ).

7.36 t J (Hz) 142.8 J (Hz) 7.36 t 142.8
111
H
110.9 HMBC 6.25 dt
138.4 7.20 td 125.2
H
110.9 HMBC 6.25 dt
138.4 7.20 td 125.2
NOESY
NOESY
C
1.67 ddd 4.9 Hz 15.7 37.1 1.15 s 27.4 0.84 d 2.56 ddd
18.2 2.05 ddd 4.49 br d
18.2
2.34 td
1.85 ddd 39.1
2.05 ddd 4.49 br d
2.34 td
1.85 ddd 39.1 1.67 ddd
4.9 Hz 2.56 ddd
64.7 2.43-2.33 m 4.8 Hz 6.90 ddd 171.8 38.1 2.7 Hz 136.2
1.45 br s 39.7 49.0 19.8
1.60-1.54 m
64.7 2.43-2.33 m 4.8 Hz 1.47-1.42 m 6.90 ddd 171.8 38.1 2.7 Hz 136.2
1.45 br s 39.7 49.0
1.60-1.54 m 15.7 37.1 1.15 s 0.84 d
19.8 1.23-1.17 m 37.4 140.8 21.4 38.6
1.23-1.17 m 37.4 140.8 21.4 38.6
27.4 1.47-1.42 m 1.60-1.54 m
1.60-1.54 m
2.43-2.33 dt 1.64 s
2.43-2.33 dt 1.64 s
Figure 2.40: NMR assignments and key 2D correlations for 28a (stereoview).
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5 140 130 120 110 100 90 80 70 60 50 40 30 20 ppm
Figure 2.41: HMBC spectrum of 28a. The structures of 28a-28c were elucidated mainly by NMR spectroscopy (1 H,
13
C, DEPT, HMQC, HMBC, COSY, and NOESY). The 1 H NMR spectra
suggested that they were derivatives of hardwickiic acid (29a). The molecular formula of 28a (C20 H28 O4 by HRESIMS) diered from that of 29a by one oxygen atom, suggesting the presence of a hydroxy group, which was conrmed
112
0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2.0 nOe effect direct coupling 2.1 2.2 2.3 2.4 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1.0 0.8 ppm 4.47 4.48 4.49 4.50 4.51 4.52 2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 ppm
Figure 2.42: NOESY spectrum of 28a. by a strong IR absorption at 3392 cm1 . This was located at C-1 based on 1 H NMR: the oxymethine at 4.49 showed couplings to H-2, -3 & -10 (COSY), and C-3, -9 and -10 (HMBC) - see Figure 2.41 on the preceding page. The 1 conguration was conrmed by an H-1 to H-11 crosspeak in the NOESY spectrum (Figure 2.42). NOESY cross peaks also permitted stereospecic assignment of the rotatable H-11 and -12 methylenes, as shown in Figure 2.40 on the preceding page. In the favoured conformation thus established, C-12 is approximately anti-periplanar to the C-20 methyl group, as found in the crystal structure of 29a228 (Figure 2.18 on page 95). Along with the cross peaks from H-20 to H-17 and -19, these NOESY data conrm the relative conguration of all stereogenic centres.
113
7.3
7.2
OH H
H OH
2.1
2.0
1.9
1.8
28b
1.60
1.55
6.68
6.66
6.64
3.4
2.5
2.4
2.3
6.26 6.24 0.76 0.77 7 0.92 6 5 0.81
4.45
3.85 1 0.93 3 4 3 0.93 3 1 2 3.1 4.6 4.1
0.89
2.9 0.92 ppm 1
Figure 2.43: 1 H NMR spectrum of 28b (400 MHz, CDCl3 ).

7.35 t 142.8 J (Hz) 138.4 J (Hz) 7.35 t 142.8 138.4 7.20 dd 124.9 HMBC 18.2 2.08 dddd 4.46 dq 1.90 ddd 1.45 br s 64.3 2.31 ddt 3.38 dd 4.9 Hz 1.85 dq 6.65 ddd 167.3 38.0 2.8 Hz 1.49 br s 133.2 37.1 141.4 21.4 38.0 2.36 dt 1.66 s 51.3 2.42 dddd 1.77 ddd 38.8 1.64-1.49 m 39.1 44.8 1.18 s 21.9 1.85 dq 1.64-1.49 m 63.9 3.38 dd 5.1 Hz 1.49 br s 63.9 44.8 1.18 s 21.9 2.53 ddd 3.84 dd 1.49 br s
H C
110.8 6.25 dd 124.9 HMBC 18.2 2.08 dddd 4.46 dq 1.90 ddd 1.45 br s 39.1 48.7 20.9 1.19-1.13 m 37.1 141.4 21.4 167.3 38.0 2.36 dt 1.66 s 51.3
7.20 dd
H C
110.8 6.25 dd
NOESY
NOESY
2.42 dddd 1.77 ddd 38.8 1.64-1.49 m
5.1 Hz 2.53 ddd
3.84 dd
64.3 2.31 ddt 4.9 Hz 6.65 ddd 38.0 2.8 Hz 1.49 br s 133.2
48.7 20.9 1.19-1.13 m
1.64-1.49 m
3.71 s
3.71 s
Figure 2.44: NMR assignments and key 2D correlations for 28b (stereoview). The 1 H NMR spectrum of 28b suggested a methyl ester ( 3.71) with two hydroxy groups ( 1.49, 2H, D2 O-exchangeable). IR showed a strong absorption at 3434 cm1 . The molecular formula, C21 H30 O5 (HRESIMS), was consistent with this. One of the hydroxy groups was again located at C-1, showing the same couplings as in 28a. The second was located at C-17, based on the cou-
114
3.4 3.5 3.6 3.7 3.8 3.9 4.0 4.1 4.2 4.3 4.4
130
120
110
100
90
80
70
60
50
40
30
20 ppm
Figure 2.45: HMBC spectrum of 28b.
3.4 1.2 1.3 1.4 1.5 1.6 1.7 3.5 3.6 3.7 3.8 3.9 4.0 4.1 4.2 4.3 4.4 4.5 2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 ppm 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5
Figure 2.46: NOESY spectrum of 28b. plings of the oxymethylene signals ( 3.38 and 3.84) to H-8 (COSY) and to C-7, -8 and -9 (HMBC) see Figure 2.45. The NOESY spectrum showed H-1 to H-11 and H-17 to H-20 cross peaks, conrming the conguration at these centres (Figure 2.46). Again, NOESY data permitted full assignment of the diastereotopic H-11, -12 and -17 methylene protons. The 1 H NMR spectrum of 28c showed an acetyl methyl signal ( 2.03). Com-

O
115
H O
OH
28c
2.03 0.83 2.9 1.27 3.3 5.4 1.2 3.8
2.03 s 21.0 171.2 66.1 40.9 0.83 s 3.79 dd
2.5 7.3 7.2
2.4
2.3
2.2
1.8
1.7
1.6
1.5
1.4
1.2 6.91 6.90 6.89 6.88 6.87
1.1
4.25
3.80
3.75
0.81 0.8 1
0.87
6.29
6.28 1
2.1 0.99 2
1 ppm
Figure 2.47: 1 H NMR spectrum of 28c (400 MHz, CDCl3 ).

7.35 t 7.35 t
H
HMBC
142.8 138.5 7.22 dd 125.2 2.03 s 21.0 171.2
H
HMBC
142.8 138.5 7.22 dd 125.2
NOESY
110.9 6.28 dd
NOESY
110.9 6.28 dd
18.3 2.40 td 1.64 ddd 38.9 4.26 dd 1.82-1.67 m 66.1 40.9 0.83 s 3.79 dd
18.3 2.40 td 1.64 ddd 38.9 4.26 dd 1.82-1.67 m
2.24-2.16 m 1.82-1.67 m 1.42 br d38.4 46.8
2.24-2.16 m
1.82-1.67 m 1.42 br d 38.4 46.8
17.0 2.35 dt
17.0 2.35 dt
27.4 1.54-1.44 m 140.3 37.4 141.2 171.9
19.0 1.15 td 35.2 20.5 2.53 dt 1.27 s
27.4 1.54-1.44 m 140.3 37.4 141.2 171.9
19.0 1.15 td 35.2 20.5 2.53 dt 1.27 s
22.3 1.82-1.67 m 1.54-1.44 m 6.89 dd
22.3 1.82-1.67 m 1.54-1.44 m
6.89 dd
Figure 2.48: NMR assignments and key 2D correlations for 28c (stereoview).
116
1.0
1.5
2.0
2.5
3.0
3.5
4.0
70
60
50
40
30
20
ppm
Figure 2.49: HMBC spectrum of 28c.
3.8 3.9 4.0 4.1 4.2 4.3 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 1.0 0.9 0.8 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3
0.9 1.0 1.1 1.2
1.3 1.4 1.5 1.6 1.7 1.8 ppm
Figure 2.50: NOESY spectrum of 28c. pared to 28b, the C-1 oxymethine was absent, while the oxymethylene signals were shifted downeld to 3.79 and 4.26, with the same COSY and HMBC couplings (Figure 2.49), establishing the 17-acetoxy structure shown. This was consistent with the molecular formula, C22 H30 O5 (HRESIMS). As with 28b, NOESY data (Figure 2.50) conrmed the conguration of all stereogenic centres and permitted full assignment of the H-11, -12 and -17 methylene protons. The neoclerodane absolute stereochemistry shown for the salvinorins and divinatorins is common to all clerodanes isolated from the Lamiaceae,26 including
2.2. STRUCTURE ELUCIDATION. 1a24, 25 and 29a.226
117
2.2.6.2
Derivation of the Name Divinatorin.
The inspiration for the name divinatorin was a comment by Albert Hofmann, who obtained the type specimen of S. divinorum for botanical identication: It was determined at the Botanical Department at Harvard that it was a new species of Salvia and it got the name Salvia divinorum. It is a wrong name, bad Latin; it should be actually Salvia divinatorum. They do not know very good Latin, these botanists. I was not very happy with the name because Salvia divinorum means Salvia of the ghosts whereas Salvia divinatorum, the correct name, means Salvia of the priests ...258 The original name was coined by Carl Epling, a botanist uent in Latin, and was intended to mean of the seers.4 According to classicist Dr Parshia LeeStecum, the original name is in fact correct, being taken from the adjective divinus:259 divinus ... can be (and was as early as classical antiquity) used as a substantive, meaning soothsayer/prophet/seer. Hence divinorum: of the prophets/soothsayers/seers. Hofmanns alternate rendering, derived from the medieval Latin noun divinator, would have had the same meaning in the medieval period. However, this word was not used in antiquity, and divinatorum would have been interpreted as a form of the verb divino (I foresee): To someone of ... Virgils, Ovids or even Tacitus time, Salvia divinatorum would thus mean salvia of those who have been foretold ... Nonetheless, in honour of Hofmanns key role in the scientic identication of the plant, compounds 28a-28c were named divinatorins.
118
2.2.7
Subsequent isolations.
O O O H O O
4
R O O HO R HO H O O H H O
OH H
O H R O
1g
O 28d 28e
R CH2OAc CHO
O 34a 34b
R OMe OMe
Figure 2.51: Subsequently isolated compounds. Five new diterpenoids have been isolated from S. divinorum since the initial publication of this work. Lee et al isolated salvinorin G (1g) and divinatorins D and E (28d & 28e)137 after decolourisation of the crude extract as described above. The structures of the three compounds, elucidated using 2D NMR, were incorrectly drawn with tetrahedral geometry at C-4 in the original publication; this has been corrected in Figure 2.51. Harding et al isolated salvinicins A and B (34a & 34b)25 without decolourisation. The structure of 34a was elucidated using 2D NMR, and denitively conrmed by X-ray crystallography.
Chapter 3 Synthesis.
In order to study the structure-activity relationships of salvinorin A (1a) at opioid receptors, various synthetic modications were made to the compound.
3.1
Known derivatives.
O O O
O O
O OH
O O
O O R O
2
OR1 H
O O
O R2 R1 H H Ac H H Ac Ac Ac
35
1a
36h 36e 36d c 36c
c d
Conditions: a) i Bu2 AlH, THF, 78 C, 25 min, 65% (81% borsm); b) NaBH4 , EtOH, 40 C, 56%; c) Ac2 O, pyridine, DMAP, rt; d) i) trimethyl orthoacetate, AcOH, 105 C; ii) AcOH, cat. HCl, THF/H2 O, rt, 74% over 2 steps.
Scheme 3.1: Preparation of known derivatives. Several known derivatives of 1a were prepared by published procedures with minor modications. Lactol 35 was prepared using i Bu2 AlH following Browns 119
120
CHAPTER 3. SYNTHESIS.
procedure,260 albeit in lower yield. Brown reported a yield of 89% after 6 minutes at 78 C. In my hands, the reaction was not complete even after 25 minutes; yield plateaued at 65% (81% based on recovered starting material). Although warming to 40 C forced the reaction to completion, the yield was not improved due to side reactions. Diol 36h was originally prepared by Valds from 1a using 0.6 equivalents of NaBH4 in i PrOH;23, 15 as found by Brown,261 EtOH proved equally satisfactory; also, the precise stoichiometry of NaBH4 proved unimportant, although a large excess (10 eq) reduced the yield. Acetylation of 36h under standard conditions gave dihydrosalvinorin E (36e).23 Several trace byproducts could not be completely removed, even by HPLC; this problem was solved by HPLC purication of the starting diol 36e itself. Using Ac2 O/pyridine/DMAP, direct formation of dihydrosalvinorin C (36c) from 36h occurs in negligible yield, even after prolonged reux.262 Valds et als indirect route via the orthoacetate and 36d27 was therefore used. Satisfactory purication of 36c required HPLC (40% EtOAc/petrol). Harding et al have since reported that 36h can be directly diacetylated using NEt3 instead of pyridine.82
3.2
Epimerisation at C-8 under Basic Conditions.
3.2.1
Previous Reports.
In the NaBH4 reduction of 1a described above, Valds isolated equal amounts of 36h and a byproduct which appears to be stereoisomeric at C-8 and/or C-9.23 Acetylation and oxidation of this compound gave a thus far undetermined stereoisomer of 1a. Subsequently, Brown identied this compound as 8-epi-salvinorin A (37a, Scheme 3.2 on the facing page), and also reported that deacetylation of 1a under basic conditions gave 8-epi-salvinorin B (37b).263
3.2. EPIMERISATION AT C-8 UNDER BASIC CONDITIONS.
121
No characterisation data was reported for any of these compounds. Further references to these epimerisations appeared over the following decade,24, 14 but no data was published until 2001, when Valds et al characterised diol 38h.27 No basis was given for this proposed structure, however.
O
O HO O BROH O O O H H O O
H
8
O O Ac2O / pyr O
O 37b
O O
H
8
O O
O NaHCO3, DMF O O 1a
37a
NaBH4 OH H
PCC
HO
H
8
O O
O O
OH H
H
8
O O
O 38h
Ac2O / pyr
O 38e
Scheme 3.2: Formation of 8-epi-salvinorins A and B.
3.2.2
8-epi-Salvinorins A and B (37a and 37b).
Deacetylation of 1a under mildly basic conditions (saturated NaHCO3 in MeOH) gave a mixture of 1b and another compound. The new compound proved, as Brown had claimed, to be 8-epi-salvinorin B (37b). Acetylation with Ac2 O/pyridine gave 8-epi-salvinorin A (37a). The structures of 37a and 37b were elucidated using 2D NMR (HMQC,
122 1a 1b 11.6, 3.1 37a 11.7, 3.0 37b
CHAPTER 3. SYNTHESIS. 5.0, 2.2 4.8, 2.2
Table 3.1: Coupling constants (Hz) at H-8 for 1a, 1b and 8-epimers. HMBC, COSY and NOESY). The conguration of H-8 was apparent from the loss of the trans-diaxial coupling constant, establishing an equatorial conguration (Table 3.1). Also, irradiation of H-12 gave a strong nOe enhancement of H-8 (Figure 3.1 on the next page). The corresponding experiment on 1a caused an enhancement of H-20 rather than H-8 (Figure 2.14 on page 92). The coupling constants and NOESY correlations shown in Figure 3.1 also establish an approximately trans-diaxial relationship between H-11 and H-12. This suggests the preferred conformation is approximately as shown, with the furan equatorial and the C ring in a pseudoboat conformation. Consistent with this, X-ray crystallography of stereoisomeric furanolactones has shown that in the solid state, the furan is invariably equatorial264, 265, 266 even where the C ring is forced into a pseudoboat conformation.267 NMR analysis, based on coupling constants and NOESY spectra, suggests that this also holds in solution.266 It should be noted that both 37a and 37b lack the distinctive infrared carbonyl absorption reported for pseudoboat lactones ( 1760 cm1 ).268 However, those values were obtained in solution, and are subject to solvent eects; recording the spectrum as a mull gives typical carbonyl absorptions ( 1730 cm1 ).269 The spectra of 37a and 37b were recorded as neat lms. Our initial publication270 contained characterisation data for 37a but not for 37b. Harding et al 82 and Lee et al 271 subsequently published 1 H NMR data for 37b which is inconsistent with our own.
3.2.3
Control of Epimerisation and Separation of Epimers.
Brown reported that treating 1a with KCN in reuxing MeOH/THF gave 1b in quantitative yield,272 but in my hands 37b was the major product, as had been the case with NaHCO3 . Valds noted the same result.273 Additionally, the epimers were not resolved on silica using Browns solvent system (3%
123
NOESY
NOESY
1.50 dd 2.36 dd (15.0, 12.0 Hz)
(15.0, 2.2 Hz) 5.25 dd (12.0, 2.2 Hz)
1.50 dd 2.36 dd (15.0, 12.0 Hz)
(15.0, 2.2 Hz) 5.25 dd (12.0, 2.2 Hz)
2.24 br s
2.45 dd (5.0, 2.2 Hz)
2.24 br s
2.45 dd (5.0, 2.2 Hz)
Figure 3.1: Key NOESY correlations for 37a (stereoview). MeOH/CH2 Cl2 ). Petrol-based systems resolved the epimers, but gave poor recoveries: some precipitation occurred when loading in these systems. Thus, eective separation required elution with petrol-based systems for separation, followed by stripping with MeOH/CH2 Cl2 to achieve full recovery of 1b. Yield was generally below 50%, making this an unsatisfactory route to 1b. Both of these problems were elegantly overcome by Tidgewell et al,169 using a suspension of 1a and Na2 CO3 in minimal MeOH at room temperature. In my hands, very little 37b was evident by TLC under these conditions. This was removed by trituration in MeOH, giving 1b in 76% yield without chromatography. The supernatant from trituration was a complex mixture containing very little 37b. These results were published without discussion, but demand explanation. It is puzzling that Na2 CO3 appears to cause less epimerisation than the much weaker base NaHCO3 (the conjugate acid of Na2 CO3 ). When Tidgewell et als procedure was repeated in oxygen-free MeOH, the supernatant contained almost pure 37b, which was obtained as an amber resin in 23% yield by evaporation. The precipitate consisted again of crystalline 1b (74%). Thus, 1b is nearly insoluble in room-temperature MeOH, while 37b is freely soluble. Additionally, these results conrm that as expected, epimerisation does occur
124
under Tidgewell et als conditions, but is usually not apparent due to decomposition in oxygenated solution. Nonetheless, the yield of 1b is not aected, so the reaction can be performed without precautions against oxygen or moisture. Apparently, then, the improved epimeric ratio results from performing the reaction as a suspension in minimal MeOH. The reaction solution must rapidly become saturated with 1b; as additional 1b is generated, it precipitates, unlike 37b. Thus, although an equilibrium mixture of the epimers is present in solution, this represents only a small proportion of the total reaction mixture. When the deacetylation is performed entirely in solution, for example by the use of cosolvents such as THF or CH2 Cl2 , the epimeric mixture ultimately comprises the entire reaction mixture. Similarly, as will be discussed below, the procedure causes extensive epimerisation when applied to 1c, where there are no dramatic dierences in MeOH solubility among starting material and products. Substituting NaHCO3 for Na2 CO3 under Tidgewell et als conditions gave no reaction after one day. Conversely, Na2 CO3 in MeOH/CH2 Cl2 gave a complex mixture; the distinctive H-12 signals of 8-epimers were apparent by 1 H NMR. The above deacetylations proceed via base-catalysed solvolysis; the use of polar aprotic solvents (strictly, non-hydrogen bond donor solvents) permits epimerisation without deacetylation. Thus, treatment of 1a with NaHCO3 in dry DMF or DMPU at 150 C gives approximately 50% epimerisation to 37a (addition of water caused deacetylation). In commercial DMPU (98% purity), epimerisation occurred in the absence of NaHCO3 , presumably due to a basic impurity; heating in distilled DMPU or DMF alone gave no reaction. Of course, this reaction can also be used to regenerate the natural compounds from the 8-epimers.
3.2.4
8-epi-Salvinorin C (37c) and Related Compounds.
As discussed above (Section 2.2.5.2), application of Tidgewell et als deacetylation conditions to 1c (Na2 CO3 in minimal MeOH)169 gave predominantly
125
the 8-epi-diol 37h (in a 4:1 ratio with 1h) see Scheme 3.3. Again, KCN in MeOH gave the same result. Both epimers were too soluble in MeOH to permit trituration. The 8-epimer 37h was freely soluble in Me2 SO, but poorly soluble in acetone and chloroform, with a tendency to precipitate during preparation of NMR samples.
O O O O O H H O O R 2O OR1 H H O O R2 R1 37c Ac Ac
c
37d H 37h H
Ac H
O 1c
O a
O
b
37e Ac H
Conditions: a) Na2 CO3 , MeOH, rt 4 h. b) Ac2 O, pyridine, ( DMAP) c) CDCl3 , rt overnight
Scheme 3.3: Synthesis of 37c-37h. Acetylation of 37h under mild conditions gave the 2-acetate 37e in high yield. Surprisingly, standing 37e in CDCl3 overnight gave 50% conversion to the 1acetate 37d, presumably due to traces of DCl. Filtration of either compound through a plug of basic Al2 O3 also caused this unexpected acetate migration. Alumina-catalyzed acyl migrations have been reported previously274 in 1,2dihydroxy terpenoids. No such migration was ever observed in CDCl3 solutions of the natural epimers 1d and 1e. Acetylation of 37d or 37e under forcing conditions (with catalytic DMAP at 50 C) gave 37c.
3.2.5
Chromatographic Identication of Epimers.
The epimers gave dierent colours when visualised with vanillin/H2 SO4 in EtOH. After brief heating, the natural H-8 compounds slowly developed a pink/purple colour, while the 8 compounds turned blue. These colours were observed for salvinorins AE (1a, 1b and 1c1e) and the diol 1h. Additionally, in each case, the 8-epimer gave a higher Rf in Et2 O or EtOAc/petrol.
126
1a 37a eluent: Et2O
1b 37b 1h 37h 50% EtOAc/petrol Et2O
Figure 3.2: TLC comparison of epimers using vanillin/H2 SO4 . These relationships also held for many derivatives to be discussed below; thus, conguration at C-8 can be condently inferred from TLC alone. Given the tendency for salvinorins to epimerise at C-8 under basic conditions, this information should prove useful for future synthetic work.
3.2.6
Mechanism.
Koreeda and co-workers have on several occasions23, 24, 263 proposed a complex mechanism for the epimerisation of 1b, involving cleavage of the C-8/9 bond (see Scheme 3.4 on the next page). No explanation was given for rejecting the obvious mechanism of lactone enolate formation, followed by protonation from the opposite face. Presumably Koreeda and co-workers rejected this simpler mechanism because the -protons of ketones are more acidic than those of esters, by several orders of magnitude.275 Why, therefore, should epimerisation occur at C-8 rather than C-2 or 10? The more complex mechanism reconciles this apparent contradiction: abstraction of H-10 inverts the conguration of
127
H-8 indirectly. This mechanism has been tentatively endorsed by Lee et al.271
O BO RO H H O O RO
10
O H+ O
9 8
O ORO
O O
Scheme 3.4: Koreeda et als proposed mechanism of epimerisation. This mechanism is not consistent with the above results. Since epimerisation of 1c occurs under identical conditions, the ketone is not essential to the process. This is also true of many derivatives of 1a lacking the ketone, which will be discussed below. Other furanolactone terpenoids undergo epimerisation at C-8 under basic conditions,268, 276, 277 despite lacking the ketone (Scheme 3.5). Several of the structures have been denitively established by X-ray crystallography: columbin (39),265 isocolumbin (40),267 palmarin (42)266 and jateorin (43).264
O O R R O O OH 39 40 H a 44
8
O O R H O O O OH
R'
8
O O 41 42 43
R H H H H
R' H H H H a a
Conditions: a) ROH, OH , .
Scheme 3.5: Epimerisation of related natural products with base. Conclusive evidence against the mechanism can be found in Browns own results, performed under Koreedas guidance. Reuxing 1a with KCN in CD3 OD gave 45,278 deuterated at C-2, -8 and -10 (Scheme 3.6 on the following page). This establishes that exchange occurs at H-8 in a methanolic solution of CN
128
(a weaker base than CO2 ).275 Thus their proposed mechanism, in which this 3 deprotonation does not occur, is incorrect. Both the ketone and lactone are enolised; evidently, however, inversions at C-2 or -10 are thermodynamically unfavourable relative to C-8.
O O
O O
O O KCN, CD3OD HO
D
2
D
10
D
8
O O
1a
45
Scheme 3.6: Browns deuteration of 1b.
3.2.7
Attempted Deacetylation under Acidic Conditions.
Given the instability of 1a under basic conditions, deacetylation under acidic conditions was investigated. In collaboration with us, Ken G. Holden treated 1a with 5% methanesulfonic acid in CH2 Cl2 /MeOH at room temperature. Monitoring by TLC showed initial formation of 1b and byproducts; after 2 days, these compounds were consumed, giving two barely resolved spots. NMR analysis showed that each spot consisted of two compounds; deacetylation appeared to have been accompanied by lactone methanolysis (additional methoxy peaks and an upeld shift of H-12). Since this route is clearly inferior, the products were not characterised further.
3.3
3.3.1
Simple Derivatives.
Esters (46 and 47).
Treatment of 1b with HCO2 H/Ac2 O279 in pyridine280 gave formate 46. Treatment in neat HCO2 H281 proved too vigorous, giving inseparable byproducts. NMR assignments of 46 were inferred from the near-identical spectra of 1a.
3.3. SIMPLE DERIVATIVES.

O O O
129
O O
O O Br
O O
2
OH H
1
O O O
O O
46
47
48
Figure 3.3: Ester and ether derivatives.
In the hope of obtaining crystallographic conrmation of the structure and absolute stereochemistry of 1h, the para-bromobenzoate 47 was prepared using the acyl chloride, DMAP and NEt3 . As found in benzoylation of 3,4-dihydro analogues,282, 24 no dibenzoylated compound was detected. Unfortunately, 47 was obtained as a powder. Recrystallisation from MeOH, and slow evaporation of a CH2 Cl2 /Et2 O solution, failed to give crystals suitable for X-ray analysis.
3.3.2
Attempted Benzyl Ether Formation (48).
Inspired by a report that salvinorin B benzoate is a opioid agonist,82 formation of the benzyl ether 48 was attempted. Benzyl trichloroacetimidate and catalytic Me3 SiOTf in CH2 Cl2 283, 284 gave a complex, inseparable mixture which did not appear to contain 48 by 1 H NMR. An alternative route, benzyl bromide with freshly prepared285 Ag2 O in CH2 Cl2 ,286 gave negligible reaction after 15 hours, despite both reagents being in excess. Addition of Bu4 NI287 gave completion in 4 hours. Again a complex mixture of benzylated compounds resulted, which appeared to contain traces of 48. Subsequently, Bguin et al reported successful preparation of 48 using BnBr and Ag2 O in MeCN.95 This is surprising, since in previous work MeCN was the worst solvent tested, giving hydrolysis of BnBr and decreased alkylation rates.286
130
3.3.3
17-Deoxy Compounds (49 and 50).
17-Deoxysalvinorin A (49) was synthesised by deoxygenation of lactol 35 using Et3 SiH and BF3 OEt2 .288 The enol ether (50) was also formed as a byproduct (Scheme 3.7). To improve the yield of 49, other routes289 were explored. Use of Et3 SiH with Amberlyst 15 sulfonic acid resin290 instead gave 50 exclusively in 76% yield. Lactol 35 proved extremely prone to acid-catalysed elimination; storage overnight in CDCl3 at -20 C gave 52% yield of 50. Such dehydrations typically require much harsher conditions;291, 292 however, low-temperature dehydration of a hemiacetal with BF3 OEt2 has been reported.293 In rare cases, lactols may be so prone to dehydration that they cannot be isolated.294 In this case, as with epimerisation at C-8, the elimination may be driven by relief of steric interactions or strain in the natural H-8 conguration. The furan substituent may also stabilise the oxonium intermediate through electron donation.
1H 13C
HMBC O O
O O
O OH
O O
H
8 7 6
O
17
O Et3SiH, BF3OEt2, CH2Cl2, 0 C Et3SiH, Amberlyst 15 CH2Cl2
O 49
O 50
35
48%
23%
76%
Scheme 3.7: Deoxygenation of lactol 35. By 1 H NMR, the H-17 oxymethylene of 49 appeared as a doublet, 3.58, coupling to H-8 (COSY crosspeak). As with 1a (Figure 2.14 on page 92), irradiation of H-12 gave a strong nOe enhancement of H-20 rather than H-8,
3.3. SIMPLE DERIVATIVES.
131
conrming the conguration at C-8. The quaternary C-8 peak of 50 showed HMBC correlations to H-6, -7 and -17 (Scheme 3.7). A long range coupling (1.8 Hz) was evident between one of the H-7 protons and H-17.
3.3.4
Tetrahydrosalvinorin A (51).
O
13
O H O O H H H O O
O O
H O H H OH O
O 52
94% H2, Pd/C MeOH 1a
O 51
59%
H2 (4 atm), Rh/C MeOH/CH2Cl2
O H2
O + R R' OH O B R
O + O R' O C R
O O A
O O
cat.
OH R' O D
Scheme 3.8: Hydrogenation of 1a and other furanolactones.

O O O O H O H H O O Limonin (53) O O O OAc OAc O O OH H H O O H O H O O
Montanin C (54)
Teucrin A (55)
Figure 3.4: Furanolactones 53, 54 and 55. To explore the role of the furan ring in the eects of 1a, tetrahydrosalvinorin
132
A (51) was prepared. Valds had reported15, 23 that catalytic hydrogenation of 1a over palladium on carbon gave hexahydrosalvinorin A (52) in nearquantitative yield after 24 hours (Scheme 3.8). Saturation of the furan ring was accompanied by hydrogenolysis of the lactone. This is typical of the pseudobenzylic bond of furanolactones (A): although high yields of tetrahydro compounds C have been reported with Pd/C, hexahydro compounds D generally predominate (see Table 3.2 on the facing page). In some cases, use of Pd/BaSO4 also gave B. The highest reported yields of C have been achieved with PtO2 in acetic acid; however, results are highly variable. This is true of other catalysts: dierent groups report very dierent results for the same substrate under similar conditions (compare the divergent results for 53 and 54). This may be due to dierences in the catalyst; for instance, some batches of Pd/C are acidic.295 Generally, acidic conditions favour hydrogenolysis, while bases (especially nitrogenous bases)296 suppress it.297 Consistent with this, hydrogenation of 1a at rtp over Pd/C in 0.1% H2 SO4 /MeOH gave complete conversion to 52 in 10 minutes. The substrate also aects the outcome: dierent compounds sometimes give dramatically dierent results under identical conditions (compare 55 and 54 in Ref. 298). Rhodium on carbon is reputed172, 319, 297 to cause less hydrogenolysis than palladium catalysts. There have been reports of selective reduction of furanolactones using this catalyst, giving C without D, albeit in low yields.317, 318 When applied to 1a, Rh/C gave negligible progress at atmospheric pressure. At 4 atm, the reaction proceeded smoothly to give 51 in unusually high yield (59%). For characterisation, the less polar 13-epimer was separated by repeated HPLC (baseline resolution was not achieved). By 1 H NMR, H-12 showed a new coupling to H-13, but its coupling constants to H-11 were scarcely aected, suggesting little change of conformation in the lactone. Determination of the conguration at the tetrahydrofuran C-13 position is challenging. This has recently been achieved for the tetrahydro derivative C of limonin (53) by comparison of nOe (ROESY) data for each epimer with predictions from molecular modelling.307 Since only one epimer of 51 was
3.3. SIMPLE DERIVATIVES. Catalyst 30% Pd/C 10% Pd/C Solvent EtOAc MeOH Time 22 24 h* 24 h* 8h 35 min 1h 1h 30 min 45 min 5h 24 h B % C % 0 53 14 10 6 10 7 17 20 29 50 10 12 8 48 4 74 30 20 2 24 20 D % 0 65 20 78 76 89 84 85 56 80 67 59-69 41 44 59 Ref.
133
EtOH
AcOH
5% Pd/C 10% Pd/BaSO4
CH2 Cl2 dioxane EtOAc H2 O AcOH EtOH MeOH EtOH AcOH AcOH
24 h* 36 h 17 h 6h 22 18 n.s. 5h 20 h 7h 30 h 48 h 96 h 10 h 48 h * = 2 Atm.
PtO2
34 42 30 88 86 42 36
77 51 39 35 n.s. 55 10 <55 53
AcOH/dioxane 5% Rh/C AcOH EtOAc
299 298(55) 298(54) 300(54) 301 268 268 268 302 303 304 305 306(53) 276 276 307(53) 308 309 310 311(53) 312(53) 313 313 314(55) 314 315 316 305 299 311(53) 311(53) 317(55) 318
n.s. = not stated
Table 3.2: Some previously reported furanolactone hydrogenations.
obtained pure, this was not attempted in this case. Since this work, Bguin et al have synthesised 51 using Pt/C catalyst, albeit in lower yield (36%).125
134
O
H R
OR
ent-29a H ent-29b Me
Figure 3.5: (+)-Hardwickiic acid (ent-29a).
3.3.5
(+)-Hardwickiic Acid (ent-29a).
(+)-Hardwickiic acid (ent-29a) occurs in commercially-available copaiba balsam. The compound is dicult to separate from other acids in the crude mixture, and was therefore isolated as the methyl ester ent-29b following a published procedure.218 Cleavage of the methyl ester proved challenging. Heating in KOH/MeOH172 under reux gave only slow decomposition, but microwave irradiation on KF/Al2 O3 320 provided ent-29a in low yield. The reaction did not proceed in the absence of KF.
3.4
3.4.1
Modication of the Methyl Ester.

Relevant Results from Previous Work.
To explore the pharmacophore of 1a, a selective modication of the C-18 methyl ester was desirable. It was apparent from previous work that hydride reduction was unsuitable. As discussed above, Valds and Brown had achieved selective reductions of the ketone and lactone respectively (Scheme 3.1 on page 119). Varying the solvent in the i Bu2 AlH reduction was of no benet.260 Use of LiAlH4 at 78 C gave the triol 56 (Scheme 3.9). Reduction of the methyl ester was not observed in any of these reactions, demonstrating its very low reactivity. Bguin et al have recently reported full reduction of all carbonyl groups with LiAlH4 at room temperature, giving pentaol 57.125
3.4. MODIFICATION OF THE METHYL ESTER.

LiAlH4 O -78 C 1a LiAlH4 rt O Li/NH3, -100 C O
135
HO
OH H
O OH HO
OH H
OH OH
OH H
OH O
56
OH
57
OH
58
Scheme 3.9: LiAlH4 and Li/NH3 reductions of 1a. Similarly, Brown reduced the methyl ester using lithium in ammonia,321 but this was accompanied by deoxygenation at C-2, epimerisation at C-1, and cleavage of the lactone, giving 58. Brown gives two conicting structures for 58; the hybrid structure shown here incorporates those features consistent with the 1 H NMR data. The furan peaks downeld of 6 ppm rule out saturation of the furan ring. The C-18 oxymethylene and J values are nearly identical to those of 18-hydroxy derivative 77 (see Scheme 3.17 on page 152), ruling out epimerisation at C-4. Low reactivity at the methyl ester was also evident under other conditions. As noted earlier, KCN/CD3 OD gave deuteration at C-2, -8 and -10, but not C-4, to the methyl ester (see Scheme 3.6 on page 128). Clearly none of these approaches oered the possibility of a selective reaction at C-18. As an alternative approach, cleavage of the C-18 methyl ester to the acid would permit selective borane reduction to the hydroxyl, as in 57 and 58, in the presence of other carbonyls.
3.4.2
Treatment of Salvinorin A with KOH in MeOH.
The rst route attempted for methyl ester cleavage was basic hydrolysis. Since the methyl ester was unaected by Na2 CO3 in MeOH, 1a was treated with 1M KOH in MeOH . The solution turned a deep orange, and the starting material was rapidly consumed. The major product, found in the neutral fraction,
136
was enedione 59 (Scheme 3.10). The base-soluble fraction was dicult to analyze, smearing on TLC and giving a poorly-resolved 1 H NMR spectrum. Surprisingly, at least eight peaks were apparent in the methoxy region. After methylation with Me3 SiCHN2 ,322 TLC showed only a single spot.
1
H NMR
analysis, however, revealed three major compounds (60a, 60b and 60c), which were separated with diculty by HPLC. Baseline resolution was not achieved, necessitating repeated repurication and poor recoveries.
3.4.2.1
Structure Elucidation of the Products.

O O O R O O O 60b O 1a O 47% 59 O O O 53% 60c R
10
O O
O O
1. KOH / MeOH 2. Me3SiCHN2
OH O
O O
R'
8
O O 60a
R' H H H
H H H
Scheme 3.10: Autoxidation of 1a. The 1 H spectrum of 59 showed two new singlets at 6.91 and 6.99 ppm. The peak at 6.91 showed no COSY or HMQC crosspeak, and exchanged with D2 O. Such strongly deshielded exchangeable peaks are typical of cyclic -diones, whose enol tautomers are stabilised by internal H-bonds.323 Consistent with this, the compound exhibited strong IR absorptions at 3373 and 1651 cm1 (OH and enol C=C). The structure was further elucidated by analysis of the HMBC spectrum. The quaternary C-10 peak, located unambiguously by its correlations to the H-19 and 20 methyls, showed a correlation to the enolic proton, placing the enol at C-1 and the ketone at C-2. The vinylic H-3 peak showed the expected correlations to C-1, 4, 5 and 18 (see Figure 3.6). UV absorptions at 215, 249 and 324 nm conrmed an extended system, as in methyl 4-oxopentenoate (61, Figure 3.7 222 and 324 nm).324 Compare the spectra of 1a and 1c. Note that the longest-wavelength peak approaches visible wavelengths. The absorptions of -diones exhibit a bathochromic shift
137
towards visible wavelengths in basic solution:325, 326, 327 the orange colour observed during formation of 59 in basic MeOH is consistent with this.
O
1
13C
H O
2
O
1
10
20 5
O O O 59 O O
O O
3
4 19 18
H O
60a
Figure 3.6: Key HMBC correlations of 59 and 60a. HRESIMS conrmed the molecular formula. The remainder of the structure, unchanged from the salvinorins, was fully elucidated and assigned by NMR experiments (DEPT, COSY, HMQC, HMBC, and nOe). The H-12 coupling constants were closer to those of 37a than of 1a, suggesting that epimerisation at C-8 might have occurred. However, the conguration of H-8 was evidenced by a trans-diaxial coupling constant (9.7 Hz); in addition, irradiation of H-12 gave an nOe enhancement of H-20. The structure of 59 thus established is remarkably similar to salvinorin G (1g, Figure 2.51 on page 118).137 The major product from the base-soluble fraction was identied as 1,2-secotriester 60a (Figure 3.6) based on extensive NMR experiments. The H-10 singlet showed an HMBC correlation to the new C-1 ester carbonyl (Figure 3.6). H-4 showed correlations to C-3, 5 and 10, and formed an isolated spin system with the two deshielded H-3 peaks. This conrmed the location of the new methyl esters at C-1 and 2, although the three esters were not suciently resolved in the 2D spectra to allow individual assignment. The remaining NMR data was very similar to 1a. The chemical shift and coupling constants of H-12 were near-identical to those of 1a, conrming the conguration at C-8. HRESIMS conrmed the molecular formula. The second major product was identied as the 8-epimer 60b based on the shifts and coupling constants of H-8 and 12 (near-identical to those of 37a).270 Assignment of the remaining data was straightforward. 2D NMR showed the same correlations as 60a; HRESIMS
138
20000 18000
O O
O
16000 14000 12000 10000 8000 6000 4000 2000 0 200
59
O O
O O O O O O H H
61
O O OH H O O O
1c
1a
O O
1c 59 1a
O O
220
240
260
280
300
320
340
360
380
400
(nM)
Figure 3.7: UV/Visible spectra of 1a, 1c and 59 in MeCN.
again conrmed the molecular formula. Interestingly, although 60a and 60b cospotted by TLC, they gave the expected colours with vanillin/H2 SO4 in EtOH (Figure 3.2 on page 126). The third (minor) compound decomposed in CDCl3 before characterisation was completed, but was tentatively assigned as 60c (Scheme 3.10 on page 136). HRESIMS established that the compound was also an isomer of 60a, and the appearance of the same couplings in the COSY spectrum suggested another stereoisomer. The coupling constants of H-8 and 12 established that the Cring congurations matched those of 60a. Indeed, all of the coupling constants determined were close to those in 60a, whereas many chemical shifts showed large changes. This implied a change in the electronic environment of the coupling protons, without a change in their conguration. The most plausible candidate structure was therefore the 10-epimer 60c, since H-10 is not coupled. Placing a large substituent in an axial conguration would be expected to aect the conformation of both remaining rings, and hence the chemical shifts around those rings. By contrast, inversion at C-4 would not be expected
139
to dramatically alter the conformations of the rings, but would be expected to alter the coupling constants with the H-3 protons. These couplings were scarcely changed, while the chemical shifts of H-4, 7, 8, 10, 11, and 12 (but not H-3) were dramatically altered. Thus 60c is the more plausible structure. The particularly large change at H-4, shifted downeld by 0.83 ppm, might be due to falling within the deshielding region of the C-1 carbonyl. Also, as mentioned earlier, H-10 is much more readily exchangeable than H-4. Brown reported278 that when 1a was reuxed with KCN in CD3 OD, deuterium exchange occurred at H-2, 8 and 10 but not H-4 (Scheme 3.6 on page 128). In the absence of nOe data, however, the proposed structure 60c must remain tentative.
3.4.2.2
Comparison with Previous Reports.
These results conict with those published previously. Tidgewell et al 169 reported that heating 1a with NaOH in MeOH caused cleavage of the methyl ester and opening of the lactone, without giving further detail. Since lactone hydrolysis would be reversed upon neutralisation, this presumably refers to methanolysis. Ester cleavage may have been inferred from the formation of a base-soluble fraction, and lactone methanolysis from the methoxy peaks in the 1 H NMR spectrum. As shown above, however, the acidic fraction and its methoxy peaks result from cleavage of the -hydroxy ketone. When the 1,2-diol 36h23 (Figure 3.8) was reuxed in KOH/MeOH for 30 min, an epimeric mixture at C-8 was recovered in near-quantitative yield, conrming that neither methyl ester hydrolysis nor lactone methanolysis occur under these conditions.
O O
HO
OH H
O O O
O O
36h
62
Figure 3.8: Diol 36h and proposed autoxidation product 62.
140
More recently, Lee et al treated 1a with Ba(OH)2 in MeOH, reportedly obtaining 62 in 75% yield.271 The 1 H and 13 C NMR data quoted for 62 are identical to those of 59, apart from the omission of the H-6 multiplet at 1.77-1.67 ppm; they are evidently the same compound. Their proposed structure 62 is not consistent with the additional data presented here. Specically, HRMS established the molecular formula as C21 H22 O7 (59) rather than C21 H22 O6 (62). Further, the singlet at 6.91 cannot be attached to C-1, since it exchanges with D2 O, has no HMQC crosspeak, and lacks the expected HMBC correlations to C-3, 5, and 9. A corresponding methine peak is also absent from the DEPT spectra. The red colour of the reaction mixture125 again suggests the distinctive red shift of -diones. The presence of the C-1 ketone was conrmed by reduction (Section 3.4.2.5 below). Finally, deoxygenation of a ketone would not be expected under these conditions.
3.4.2.3
Proposed Mechanism.
Autoxidation of -hydroxy ketones (acyloins or -ketols) to -diones (diosphenols) under basic conditions is well-established.325, 328, 329 The reaction consumes one equivalent of O2 , generating H2 O2 .329 While the autoxidation of unsubstituted ketones requires stronger bases such as t BuOK, -hydroxy ketones are more readily enolised, and the reaction proceeds with KOH.328 A proposed mechanism, via saturated dione 63, is shown in Scheme 3.11. The alternate pathway, to seco-diester 64, also has numerous precedents.328, 330 We based our proposed mechanism on the generally-accepted formation of hydroperoxide intermediates,328 although this mechanism has been disputed.330 Tautomerisation of the enolate or radical will give the regioisomeric diester, which along with epimerisation at C-8 and 10 explains the numerous methoxy peaks in the 1 H NMR spectrum of the crude product. Dehydrogenation of 63 to form 59 is unusual. While there have been several reports of dehydrogenation of 1,4-diones in alcoholic KOH,331 literature searches48 revealed no such reaction involving a 4-ketoester. However, -diones

O O O2O O2 HO O
141
O O
O O
-OH
O HO
O O
1a
O MeOH
O MeOH/MeOO
O O
O O
-
HOO HO OH O
O O
HO O HO
O O H
O O
63
1) -OH
2) O2 O MeOH O O 59 O
O O O HOO
O O HO O
O O
O2 O O
64
-OH
Scheme 3.11: Proposed mechanism of the autoxidation.

O O O
O HO
O O CrO3 pyridine O
O O -H2O HO O
O O
1b
65
66
Scheme 3.12: Unexpected oxidation product 65.
are much more readily enolised than unsubstituted ketones; in the case of 59, no trace of the 1-keto tautomer was detectable by NMR. It is therefore plausible that 63 should be extremely reactive, and unsurprising that this compound
142
was not isolated. Consistent with this, Browns attempts to prepare 63 via PCC oxidation of 1b gave no isolable product.332 More recently, Harding et al oxidised 1b using similar conditions (CrO3 in pyridine), unexpectedly obtaining the lactone 65136 (Scheme 3.12) with the loss of C-2, rather than the expected 63. No mechanism was proposed. It is interesting to note that lactols analogous to 66 have been reported from autoxidation328, 333 or ozonolysis with oxidative workup334 of -diones, and that dehydration of 66 would give 65. However, if 66 were formed in the presence of excess CrO3 , oxidation rather than elimination would be expected.
3.4.2.4
Variation of Reaction Conditions.
The yield and selectivity of autoxidations of this type can be subject to strong solvent eects.328 The reaction was therefore repeated in EtOH, i PrOH and
t
BuOH. The resulting neutral and acidic fractions were noticeably more com-
plex. In each case, 59 was contaminated by inseparable impurities (presumably including the 8-epimer); the original selection of MeOH thus proved fortuitous. The reaction proceeded with only traces of oxygen, for instance when performed under N2 in MeOH pre-saturated with N2 . This is again typical.328, 330 Nonetheless, the reaction was faster and more consistent when the solution was saturated with O2 . Anther useful renement was the use of dilute KOH rather than NaHCO3 to extract the extremely hydrophobic acid diesters during workup.
3.4.2.5
Attempted Reductions of 59.
The unexpected installation of the 3,4-double bond in 59 suggested the compound might provide a route to diol 1h. However, reduction with NaBH4 in EtOH/CH2 Cl2 was accompanied by conjugate addition and epimerisation, giving 8-epi-diol 38h27 in low yield (Scheme 3.13). Attempted Luche reduction with NaBH4 CeCl3 in MeOH (with335 or without336 sonication) was also unsuccessful, giving a complex mixture whose unstable major components re-

O O
143
HO
OH H
O O HO
OH H
O O
O 38h NaBH4 59
O NaBH4 CeCl3
O 1h
Scheme 3.13: Attempted reductions of 59.
tained the characteristic enedione peaks at 6.8 and 7.0 ppm. These products cospotted with the starting material in petrol-based systems, making the reaction dicult to follow, but were resolved by acetone/CH2 Cl2 and gave a darker purple with vanillin. There are precedents for ketones which are smoothly reduced by NaBH4 alone, but not in the presence of CeCl3 .335 Note that use of forcing conditions337 (reux for 12 hours) was futile. NaBH4 decomposes to unreactive tetramethoxyborate within 5 minutes under the standard conditions (with CeCl3 in MeOH at room temperature).336 Prolonging or heating the reaction will therefore achieve nothing. Continued reaction would require frequent additions of fresh reagent.
Commonly used drying procedures for CeCl3 can cause partial decomposition.338 The quality of the CeCl3 used was veried by reduction of 2-cyclohexenone,336 which gave 2-cyclohexenol rapidly and quantitatively. The enols of -diones form complexes with a wide variety of metal salts such as FeCl3 ;339 it appears likely that such an enolic complex forms with 59 in preference to the desired ketonesolventCe3+ complex,336 and the reaction therefore does not follow the desired course. The only successful precedent located for this reduction involved non-enolisable enediones.340 It might be possible to prevent this problem by protection of the enol, but an attempt at Et3 Si protection was unsuccessful.
144
3.4.3
O-Demethylsalvinorin A (67a).
To summarise the above results, previous work showed the C-18 methyl ester to be the least reactive carboxyl of 1a under a variety of conditions. The few reactions vigorous enough to attack this position gave undesired side reactions at other positions; no selective transformation had been reported. The preliminary results above were consistent with this. The selective cleavage of methyl esters has been the topic of extensive research; thorough reviews are available.341, 342 The most eective procedures involve nucleophilic substitution at the alkoxy (or carbinol) carbon, via an SN 2 mechanism, with the carboxylate as leaving group. This mechanism (BAl 2 ester cleavage) contrasts with the BAc 2 mechanism typical of basic hydrolysis, involving attack at the carboxyl carbon (Scheme 3.14).343 The tendencies of dierent nucleophiles to act via these mechanisms can be understood in terms of hard/soft acid-base theory.342 Soft nucleophiles such as I attack the soft electrophilic carbinol via the BAl 2 mechanism, while hard nucleophiles such as F favour attack at the hard electrophilic carboxyl carbon, via the BAc 2 mechanism. Thus soft nucleophiles are less aected by hindrance around the carboxyl carbon, but more aected by hindered alkyl chains; hence the selectivity for methyl esters. Given the severe hindrance exhibited by the C-18 carboxyl carbon of 1a, the desirability of methyl ester selectivity, and the disastrous eects of hydroxide, the use of soft nucleophiles was clearly preferable.
NuO H H R O H BAl2 R
O O+ CH3Nu
O R O
-OH
BAc2 R
-O
O OH R
O O+ CH3OH
Scheme 3.14: BAl 2 and BAc 2 ester cleavage mechanisms.
3.4. MODIFICATION OF THE METHYL ESTER. 3.4.3.1 Cleavage with Iodide and Cyanide Reagents.
145
Perhaps the most widely used source for a soft nucleophile is LiI. Early reports used pyridine and 2,6-lutidine344 as solvent, but subsequent work found more polar non-hydrogen bond donor solvents such as DMF345 and especially HMPA346 (Figure 3.10) to be superior, as would be expected for an SN 2 reaction.347 Anhydrous LiI has been reported to give greater selectivity for methyl esters, but lower yield, than the hydrate.348
O O O
O O
O O
O O
O O HO
O O
1a
OH
67a
OH
67b
Figure 3.9: O-Demethyl salvinorins A and B. An initial trial of anhydrous LiI in dry DMF, after reuxing for 24 h, gave a 29% yield of the epimerised acid 67a. The dark brown colour and shy odour of the reaction mixture suggested decomposition. The neutral fraction was a complex mixture containing little starting material. Adding sodium acetate has been reported to lower the required temperature and reaction time for this procedure.349 However, in this case the reaction showed no progress after 6 hours at 130 C, and the yield after reuxing overnight was not improved.
O N N N O N O P N N N HMPA
2,6-lutidine
DMF
DMPU
Figure 3.10: Useful non-hydrogen bond donor solvents. The use of HMPA was avoided due to its carcinogenicity,347 but work with other nucleophiles has found that the safe substitute DMPU350 (Figure 3.10)
146
gives comparable rate enhancements.351, 352 Substituting this solvent for DMF (150 C for 25 h) gave no improvement in absolute yield (27%), but a much cleaner neutral fraction of epimerised 1a, giving a 79% yield based on recovered starting material. The reaction mixture again showed a dark colour and a shy smell suggestive of decomposition. A run at a higher temperature (190 C) gave a dierent base-soluble product which was not identied. The NMR spectrum showed an epimeric mixture lacking the acetyl peaks of 67a. It did not match 67b, however, and was not acetylated by Ac2 O/DMAP in pyridine. The use of NaI in DMPU at 150 C for 23 h gave a light-coloured reaction mixture and a greatly improved yield of 67a (73%). This was surprising, since lithium has been found to be superior to sodium as a counterion with iodide345 and other353 nucleophiles. Unfortunately, two careful attempts to replicate the reaction gave the same yield as LiI. Sodium cyanide has been reported to give superior results to lithium halides, including LiI, especially in HMPA.346, 343 Treatment of 1a with KCN in DMPU at 90 C for 27 h, however, gave a complex mixture with only traces of 67a. With NaCN the reaction proceeded more slowly, but again gave a complex mixture after consumption of starting material (70 h).
3.4.3.2
Cleavage with Thiolates.
Thiolates (RS ) are another group of soft nucleophiles commonly used for methyl ester cleavage. A recent study reported excellent results using arylthiols and catalytic base at 190 C, generating thiolate in situ.351 Attempts to apply this procedure were unsuccessful: treatment of 1a with 4-methylbenzenethiol with K2 CO3 in DMPU gave no reaction after 10 minutes. Prolonged reaction (16 h) gave decomposition, with no indication of acid formation. Aliphatic thiolates have found greater use. Lithium methanethiolate (LiSMe),354 as well as the ethyl355 and propyl353 homologues all cleave methyl esters at room temperature, whereas halide and cyanide reagents are typically used above 100
147
C. Again lithium353 appears to be a superior counterion to sodium,356 and
HMPA353 and DMPU352 superior solvents to DMF. Alkylthiolate solutions are unstable.353, 355 However, LiSMe is easily prepared and air-stable;354 ethanethiolates are now commercially available. For this work, LiSEt was chosen since ethanethiol (bp 35 C) is much less volatile than methanethiol (bp 6 C). The sickening thiol odour is therefore greatly reduced. The salt was prepared from ethanethiol and n BuLi using a simplied version of published procedures for LiSMe354 and LiSPr.357 Again, DMPU was substituted for the carcinogen HMPA.
O O
O HO
O O Ac2O, DMAP DMPU
O O
O O
OH
67b
O O
68
Scheme 3.15: Formation of mixed anhydride 68. The reaction between 2a and LiSEt showed little progress after 6 hours at room temperature. After heating to 55 C for 23 hours, the reaction went to completion, with little change in colour. This was in marked contrast to the high-temperature procedures. Ester cleavage was accompanied by deacetylation, giving 67b; standard acetylation conditions gave the epimeric acids 67a in good yield (typically 73% over two steps). On one occasion the acetylation was performed in one pot, after quenching the thiolate with acetic acid (the orange colour faded to faint yellow). Addition of Ac2 O/DMAP gave 67a in four hours, unfortunately accompanied by the mixed anhydride 68 (crude 1 H NMR of the neutral fraction showed additional characteristic peaks at 2.24 and 2.25 ppm; Scheme 3.15). The anhydrides proved remarkably stable. Some starting material remained after reux in THF/H2 O for 1 hour; addition of NaHCO3 was necessary for complete hydrol-
148
ysis to 67a. This process was more laborious than a separate acetylation step, and was therefore abandoned. The results of the various cleavage attempts are summarised in Table 3.3. T C LiI DMF 150 LiI NaOAc 130-150 LiI DMPU 150 LiI 190 NaI 150 NaCN 90 KCN 90 200 p-MePhSH cat. K2 CO3 LiSEt 55 * = based on recovered starting material Nu Base Solvent time h 24 21 25 24 14 70 27 16 23 Yield of 67a Isolated borsm* % % 29 26 27 79 0 23 0 0 0 73 (after acetylation)
Table 3.3: Summary of results - nucleophilic cleavage of 1a methyl ester.
3.4.3.3
Conrmation of Structure of 67a.
The mixed acids 67a streaked on TLC, and reduction by BH3 was also expected to cause epimerisation, so chromatographic separation was not attempted. The structure of 67a was denitively conrmed by methylation with CH2 N2 , giving a mixture of 1a and 37a by 1 H NMR. The products cospotted with authentic material in both petrol and CH2 Cl2 based TLC systems and gave identical colours with vanillin dip. The 1 H NMR spectrum of 67a lacked the methoxy peak, but was otherwise near-identical to a mixture of 1a and 37a. One interesting dierence was the rst-order H-4 multiplet at 2.80 (dd, J = 5.2, 3.5). The H-4 multiplet of 1a is slightly non-rst order due to the almost coinciding H-3 peaks, but at 800 MHz can be approximated as 2.74 (dd, J = 11.3, 5.6). Apparently the formation of a carboxylic acid dimer (as with hardwickiic acid, Figure 2.18) pushes the C-4/18 bond away from the equatorial position, altering the conformation of the A ring. Thus the trans-diaxial coupling constant for H-4 is lowered, and the H-3 multiplets are no longer coincident.
3.4. MODIFICATION OF THE METHYL ESTER. 3.4.3.4 Subsequent Developments and Discussion.
149
Since this work, Me3 SnOH has been reported to cleave methyl esters selectively under mild, neutral conditions in complex substrates too sensitive for previous methods.358 While the method appears very promising, it has several drawbacks. Firstly, the compound is extremely neurotoxic.359 Secondly, like other oxytin reagents, Me3 SnOH is a hard nucleophile, attacking the carboxyl carbon.360 The reactivity of other oxytin reagents is greatly reduced against hindered carboxyls, including terpenoids resembling 1a.360 Also since the publication of the above work,270 Lee et al have reported an improved yield of 1a via the LiI method (56%125 or 72%),361 by reuxing in pyridine for 36 hours. The LiI was presumably a hydrate since no drying is mentioned. If the higher yield proves reproducible, this method oers clear advantages over the thiolate route above: the acetate remains intact, and the reagents are common, stable and odourless. This result is surprising in light of the early results on solvent eects discussed in Section 3.4.3.1 on page 145. By the same token, the alkylthiolate route has not been optimised. This route oers the inherent advantage of proceeding at or near room temperature. Given the accumulation of the basic carboxylate during BAl 2 ester cleavage, and the base-sensitivity of 1a, this should permit higher maximum yields. The deacetylation observed with LiSEt is puzzling, since alkylthiolates have been found previously to cause less deacetylation than LiI.353 One potential explanation is contamination, since the thiolate was not puried before use. On several occasions after the cleavage and acetylation, 2-hydroxyethyl acetate362 was unexpectedly isolated. If one of the reagents used (presumably the thiol) was contaminated with ethanediol, the resulting thiolate would be contaminated with the dihydroxide. Also, any LiOH present in the n BuLi used (from exposure to moisture) would also have been present in the thiolate. The
n
BuLi solution used was labelled 2.5 M, but titration363 gave a value of 2.1 M.
Finally, as discussed in Section 3.2.3 on page 124, the commercial DMPU used subsequently proved to contain an impurity capable of causing epimerisation.
150
However, this was not accompanied by deacetylation, even at 150 C. One of these possible contaminants may have been responsible for the deacetylation observed, and may also have lowered the yield. The use of demonstrably pure solvents and reagents may improve the outcome of this reaction.
3.4.4
3.4.4.1
O-Demethyl-18-deoxysalvinorin A (77).
Borane Reduction of 67a.
With the acid 67a in hand, the planned borane reduction could be attempted. BH3 THF reduces carboxylic acids rapidly at low temperatures (0 C or below in many cases), even when hindered (69, Scheme 3.16).364 Esters and lactones are generally reduced at a much lower rate,365 allowing mild, selective reductions of polyfunctional acids such as 70366 and 71.367 However, some acids are much less reactive. For example, many aromatic acids such as 72368 require excess BH3 , longer reaction times and higher temperatures.369 In some extreme cases, reuxing for several hours with excess BH3 is required (73370 and 74).371 Conversely, some lactones (75 and 76) are rapidly reduced to lactols under mild conditions, and even to cyclic ethers and acyclic diols with prolonged reaction.372 Thus, the selectivity achieved depends on the substrate and precise conditions. An excellent review is available.365 Slightly dierent protocols have been used, but generally borane is added dropwise at low temperature, then the mixture is warmed gradually to the desired reaction temperature.364, 373 During the initial addition, deprotonation of the acid gives intermediate acyloxyboranes374 (with visible evolution of H2 ), which are reduced much faster than other functionalities, accounting for the selectivity of the reagent. The low temperature and slow addition of BH3 minimises side reactions until these intermediates are formed. Unfortunately, attempts to apply the mildest possible protocol to 67a were unsuccessful. Addition of BH3 THF at -25 C gave no visible evolution of H2 .

X OH O Br 69 1.3 eq, 0C to rt, 1h, 95% 70 X O O O 71 O X OH OH O
151
1 eq, 0 C, 3h, 86%
1.3 eq, 0C, 30 min, 66%
X HO 73
X OH
X
O H2 BH3THF HO
O O
O O X OH
72
4 eq, reflux, 4h, 94% X HO OR 74 (R= Me/Et) H H H X O 75 X O H H H OH H OR OR OR
3.5 eq, rt, 48h, 71% OH X 4 eq, reflux, 3h, 77%
76
X
O HO H2 1 eq, rt, 30 min, 80% 3 eq, rt, 3 days, 55%
Scheme 3.16: Some previously reported BH3 THF reductions.
Gradual warming to 0 C and addition of excess borane had no eect. Warming to room temperature, then 45 C, over several days gave no reaction. Starting material was recovered almost quantitatively. In later trials, borane was added at room temperature, giving the rst visible signs of H2 evolution. The actual
152
reduction, however, required a large excess of borane. Small amounts of the 8-epi-alcohol 78 were detected (Scheme 3.17), but not the desired alcohol 77. Reuxing with excess borane gave 78 accompanied by unidentied byproducts. Therefore a smaller excess was tried at intermediate temperature. The most successful conditions are shown in Scheme 3.17, giving the epimeric products in 48% total yield.
O O
1H 1H
COSY O O O H H O O BH3THF (1.3 eq) O OH 67a dropwise, rt x 5 min, then 55 C x 90 min OH 23% 77 25% 78 H O O O H R O O R H
Scheme 3.17: Borane reduction of 67a. These conditions are not optimised. None of the target 77 was isolated until the last few experiments, when it was found to cospot with the starting material in the TLC system used previously (1% AcOH in 20% acetone/CH2 Cl2 ). Thus, in earlier experiments, clean formation of the desired product may have gone undetected, while decomposition of that product may have been mistaken for consumption of starting material. An alternate solvent system, 1% NEt3 /EtOAc, resolved all compounds. Earlier trials may have therefore been more successful than they appeared, and milder conditions warrant reinvestigation. Nonetheless, it is apparent from the lack of hydrogen evolution at 0 C that 67a is an exceptionally unreactive substrate.374 This is unsurprising given the near-inertness of the C-18 position to all other procedures tried - hydride reagents, deuteration, strong base and soft nucleophiles. On the other hand, given the side reactions apparent at reux with BH3 , 67a is also a sensitive substrate. It is evident that the optimal conditions will be intermediate between the mild, original protocols364, 373 and the harsh conditions required for stubborn, resilient substrates such as 74.
3.5. MODIFICATION OF THE KETONE. 3.4.4.2 Structure Elucidation of 77.
153
Although the H-4 multiplet of 77 ( 1.89 ppm) was too complex for determination of coupling constants, the COSY spectrum showed cross peaks with the diastereotopic H-18 oxymethylene ( 3.94 & 3.49 ppm). The chemical shifts and coupling constants for H-8 and H-12 were near-identical to those of 1a, and irradiation of H-12 gave a strong nOe enhancement of H-20 rather than H-8 (compare Figure 2.14 on page 92), conrming the conguration at C-8. A strong infrared absorption at 3468 cm1 conrmed the presence of a hydroxyl. HRMS conrmed the molecular formula.
3.5
Modication of the Ketone.
Selective reduction of the C-1 ketone in 1a had already been reported (36e).15, 23 Since the 1-hydroxy of this compound was axial, however, any change in biological activity would be ambiguous; attributable either to the loss of the ketone, or the presence of a new protruding H-bond donor. Other modications of the ketone were therefore investigated. One obvious alternative was deoxygenation. Methylenation was also attempted, since this would preserve the sp2 hybridisation of C-1 and thus have less eect on the conformation of ring A, albeit with a much larger substituent.
3.5.1
Attempted Methylenation.
Initially, Wittig olenation was attempted. Treatment of 1a with the ylide formed from Ph3 PMeBr375 and n BuLi (1.6 eq, THF, 35 C, 20 hours) gave only epimerisation and partial deacetylation. No trace of the desired methylenated compound 79 was detected. The Wittig reaction often fails with hindered substrates, and epimerisation is also common due to the basicity of the ylide.376 The Tebbe reagent and related titanium compounds377 allow the methylenation of some carbonyl compounds
154
O
O O H H
O O
79
Figure 3.11: Methylenated target compound 79. for which Wittig conditions fail. Treatment of 1a with Zn-CH2 Br2 -TiCl4 reagent378 (room temperature, CH2 Cl2 , 1 hour) gave an extremely viscous, black reaction mixture. After standard workup, however, only starting material was detectable (72% recovery). The reagent was not tested for activity against a known substrate, but was freshly prepared379 and had the reported grey colour and thick consistency. On addition to water, the reagent blackened and eervesced vigorously.
3.5.2
Attempted Direct Deoxygenation.
O S O O R R TsNHNH2 R N R NH NaBH4 R R
Scheme 3.18: Ketone deoxygenation via a tosylhydrazone. The next modication of the ketone to be attempted was deoxygenation. The Clemmensen reduction (Zn/HCl) was not suitable, since -acetoxy groups are eliminated,380 even under the mildest conditions.381 An alternative mild approach is via a tosylhydrazone, which can be reduced to the hydrocarbon by hydride reagents (Scheme 3.18).382, 383 In attempts to form the tosylhydrazone, 1a was treated with tosylhydrazide384 under a variety of conditions (Table 3.4). Since no reaction occurred in solution (even with sonication), microwave irradi-
3.5. MODIFICATION OF THE KETONE.
155
ation of the neat reagents was attempted. Under normal conditions, this is not eective for ketones.385 However, some ketones react readily when the ask is supported in an alumina bath, generating much higher temperatures.386 Since these conditions proved ineective, a procedure for oxime formation on silica gel387 was adapted. Basic and acidic alumina were also tried as substitutes. Unreacted starting material was recovered in all cases. It is interesting to note that 1a is stable at room temperature in acetic acid (compare formic acid, Section 3.3.1 on page 128). Solvent Catalyst THF AcOH AcOH EtOH basic Al2 O3 silica gel basic Al2 O3 acidic Al2 O3 Ultrasound: 40 kHz, 50 Conditions Time Ref. reux 6h 388 rt 18 h 389 ultrasound, rt 2h 390 reux 3h 391 386 microwave, rt to 115 C 5 min microwave, rt to 115 C 5 min 387 microwave, rt to 115 C 5 min 387 microwave, rt to 115 C 5 min 387 W transmitted. Microwave: 1400 W (700 W output).
Table 3.4: Unsuccessful treatment of 1a with excess tosylhydrazide. The tosylhydrazide used was prepared by a published procedure,384 omitting the optional recrystallisation, which caused decomposition. Identity and purity (especially the absence of the typical384 impurity ditosylhydrazide)392 were conrmed by TLC and 1 H NMR.393 No other promising methods of tosylhydrazone formation were located. ElSayeds recent review of sulfonohydrazides,394 although published in 2004, appears to have been written decades earlier. The only post-1971 references cited are the authors own, and these are cited only in the nal sentence.
3.5.3
3.5.3.1
Indirect Deoxygenation.
Formation of Cyclic Thionocarbonate (80).
Given the failure of the direct reductions, deoxygenation of 36h was attempted. The usual approaches to hydroxyl deoxygenation,395 involving either
156
O S
O S HO OH H H O O DMF, 90 C O O 36h O O 80 N N N N O O H H O O
67% from 1a
Scheme 3.19: Formation of cyclic thionocarbonate 80.
hydride reduction of sulfonate derivatives or radical reduction of thiocarbonyl derivatives,396 would require derivatisation of the extremely hindered and unreactive 1-hydroxyl group. While direct acetylation has been achieved using NEt3 ,82 this is ineective for benzoylation (as seen with the less-hindered 1h, Section 3.3.1 on page 128). 1b has been mesylated at C-2 under mild conditions;82 however, the yield was very low (32%), and C-1 is much more hindered.
Thus, direct installation of hindered functionalities at C-1 is likely to be difcult. Given that borohydride reduction of 1a proceeds exclusively from one face, and sluggishly, the reduction step is also likely to be challenging. Of the two approaches, ionic versus radical reduction, radical reductions are less susceptible to hindrance.396 Inspired by the indirect diacetylation of 36h via the 1,2-orthoacetate (Section 3.1),27 radical deoxygenation of a cyclic thionocarbonate was attempted.
Treatment of 36h with 1,1-thiocarbonyldiimidazole (Scheme 3.19) in DMF gave 80 in high yield. Since this gave inseparable 8-epimers, the reaction was subsequently performed in two steps from 1a without separation of intermediate 36h and 38h. By 1 H NMR, the H-1 and -2 signals of 80 were shifted downeld, and the characteristic ates was present.
13
C peak (191 ppm) of cyclic thionocarbon-
3.5. MODIFICATION OF THE KETONE. 3.5.3.2 Unsuccessful Radical Deoxygenation Attempts.
157
Although radical reductions have traditionally been performed with organostannanes, these compounds are highly toxic, and purication of the products tends to be dicult.397 This has led to extensive and fruitful research into alternatives, such as silanes and phosphites.396 Perhaps the most promising396 of these alternatives is hypophosphorous acid, H3 PO2 . Typically, this is buered with NEt3 , and AIBN serves as radical chain initiator.398 However, other initiators and bases have been successfully substituted.399 The standard396 conditions of excess H3 PO2 and NEt3 in reuxing 1,4-dioxane were used; however, benzoyl peroxide was substituted for AIBN because the latter was not at hand. (BzO)2 gives superior results to AIBN with alkyl phosphites.396 Despite adding seven portions of (BzO)2 over ve hours at reux, most starting material was recovered (70%), along with 15% deprotected diols 36h/38h. No deoxygenated products were detected. One report used intriguingly simple conditions: magnesium in methanol.400 Although the method had only been proven for cyclic thionocarbonates of 2,3-dihydroxy esters, its extreme simplicity and nontoxicity were attractive. Treatment of 80 with excess Mg turnings in MeOH at reux for 40 minutes gave no reaction. Addition of activated401 Mg turnings had no eect after a further two hours reux. Starting material was recovered (94%).
3.5.3.3
Radical Deoxygenation using n Bu3 SnH.
The traditional n Bu3 SnH/AIBN route was then attempted. This is the best-established route (especially for cyclic thionocarbonates),402, 403, 404 but often gives side reactions,405, 406 sometimes to the exclusion of the desired deoxy products.407 An indispensible paper by Kanemitsu et al thoroughly describes the theory and practice of the procedure, including control of side reactions.405 The n Bu3 SnH used was prepared by a published procedure408 with reduced reaction time (10 minutes),409 distilled and stored in darkness under argon at
158
O
O O H O O + O O 81b 82b R H H 22% 25% O O 36f detected by LCMS + O O 83 4% O O H H O O O
HO 80 1) nBu3SnH AIBN toluene 80 C, 6 h 2) silica gel
O O
OH H
Scheme 3.20: Radical deoxygenation of 80. -20 C. A mixture410 of this and AIBN was added, in small portions, to 80 in deoxygenated405 toluene at 80 C over 6 hours. Flash chromatography gave the desired 81b (25%) and its 8-epimer 82b (22% see Scheme 3.20). Further chromatography gave a small amount of the cyclic carbonate 83, a common405 byproduct of this procedure. The expected 2-deoxy regioisomers 36f were not isolated; however, the less polar fractions eluted rst contained a complex mixture of products, heavily contaminated with organotin compounds. Analysis of this mixture by reverse-phase liquid chromatography/mass spectrometry (RP-LC/MS) was performed by Claudio Medana at Turin University. Comparison of the early fractions with 81b and 82b conrmed the presence of two less polar compounds isomeric with 81b (Figure 3.12). Surprisingly, the presumed 36f coeluted with the tin contaminants on C-18 reverse-phase, as on silica gel. Also, total ion count detection showed only a single broad organotin peak. Mass-selective detection, however, gave well-resolved peaks for the four products, illustrating the power of LCMS; none of the peaks were apparent by UV detection. The presence of 81b/82b in the early fractions shows that these were the major products, at least 2:1 relative to 36f.
3.5.3.4
Potential Improvements.
Given the incomplete recovery of 81b/82b, the yield could clearly be increased. One simple and eective way of removing organostannane reagents
3.5. MODIFICATION OF THE KETONE.

100 80 60 40 20 0 18.60 total ion count 50 m/z 700
159
(early fractions)
relative abundance
100 80 60 40 20 0 100 80 60 15.24 40 20 0 0 2 4 6 8 10 12 14 81b 15.95 15.21
19.03 m/z = 377 19.39 (early fractions)
82b 15.95 m/z = 377 (pure 81b/82b)
16 18 20
22 24 26 28 Time (min)
30 32 34 36
38 40 42
44
Figure 3.12: RP-LCMS traces of early fractions versus 81b/82b.
and byproducts is by washing an acetonitrile solution with hexanes.411, 410 However, in this case, the early fractions remained complex after washing, despite a dramatic reduction in organostannanes (mass reduced by a third). An alternative method is C-18 reverse phase chromatography.412 In this case, C-18 RP-TLC in 80% MeCN/H2 O conrmed the eectiveness of the wash: the petrol layer consisted of baseline material which was absent from the MeCN layer. However, the desired products were not resolved by RP-TLC, smearing badly. It would not be advisable to apply either of these methods to the crude reaction mixture: the initial reaction products are themselves alkyltinsubstituted thiocarbonates, which are cleaved on silica to give the desired alcohols.405 Two more complex methods of organostannane removal, involving treatment with NaBH3 CN413 or I2 and KF,397 are harsher and have the potential for side reactions.
160
Thus, the organostannane route used above has serious disadvantages: the requirement for freshly prepared reagents, the diculty of purication, the low recovery of products and most importantly toxicity. While n Bu3 SnH
is less toxic than other organostannanes,414 it is prepared from the highly toxic (n Bu3 Sn)2 O,415 and the toxicity of the organostannane byproducts is unknown. This route is therefore not recommended. There are now many alternatives,398, 416, 417 some of which have been successfully applied to cyclic thionocarbonates.418, 419
3.5.3.5
1-Deoxysalvinorin A (81a).
O
O O
2
H
1 10
O O
81a
Figure 3.13: 1-Deoxysalvinorin A (81a). Acetylation of 81b (Ac2 O/pyridine/DMAP, room temperature) gave 81a. Both H-2 ( 4.74, tt) and H-10 ( 1.10, dd) showed new trans-diaxial couplings to H-1 ( 1.50, td), and smaller couplings to H-1 ( 1.95-1.89, m). The HMQC spectrum correlated these peaks with a new 13 C peak at 26.7 ppm. The DEPT spectrum conrmed C-1 as a methylene, and HRMS conrmed the molecular formula. The remaining data was extremely similar to 1a.
Chapter 4 Bioassays.
4.1 Insect Antifeedant Activity.
A number of neoclerodane diterpenoids, very similar in structure to the salvinorins, display a broad range of activities against insects.2 Some act as insect antifeedants, or appetite suppressants, rather than insecticides. Hundreds of clerodane diterpenoids have been tested; a thorough review is available.420 Salvinorins C-F (1c-1e and 1f ) t an empirical pharmacophore for antifeedant activity against Tenebrio molitor:421 an ,-unsaturated carbonyl approximately 10 from the oxygen of a conformationally constrained furan. By contrast, a saturated C-18 carbonyl, as found in salvinorins A and B (1a and 1b), seems to confer activity against Spodoptera littoralis.422 A selection of these compounds were therefore screened for antifeedant activity. This work was done in collaboration with Dr Charles Robin and Dr David Heckel in the Department of Genetics at the University of Melbourne. The species selected was Helicoverpa armigera, which has been widely used in antifeedant tests.420 A standard choice assay employing sweetened breglass discs422, 423 was used, with one modication. In early tests, the larvae refused to eat dry discs; the discs were therefore moistened, which gave satisfactory results. Larvae were presented with two discs: both were sweetened with sucrose, and one was treated with test compound. Tests were terminated when 161
162
CHAPTER 4. BIOASSAYS.
more than half of one disc had been eaten. For full detail, see the Experimental Section. The change in masses of the control (C) and treated (T) discs allow the calculation of the antifeedant index, AI: AI = 100 (C T ) C + T
Antifeedant indices fall between +100 (maximal antifeedant eect) and -100 (maximal phagostimulant eect); a value of zero indicates no eect. The results are shown in Table 4.1.
Compound salvinorin A 1a salvinorin C 1c divinatorin A 28a divinatorin B 28b divinatorin C 28c
AI 0 10 6 2 16
SEM 30 26 11 23 33
Table 4.1: Antifeedant test results.
None of these compounds had a statistically signicant eect. While the fact that four of the ve compounds had a slightly positive AI suggests a weak (nonsignicant) eect, the standard error of the mean (SEM) is in each case larger than the mean itself. By contrast, potent antifeedants often have AI values of close to 100, with SEM < 5.420 The low means and high variability observed here suggest that these compounds are not potent antifeedants against this species. Nonetheless, the value of this experiment was limited by the unavailability of a proven antifeedant as a positive control compound. Also, there are signicant dierences between species; compounds inactive against one species are often active against others.422, 420 The most common larvae observed feeding on S. divinorum in Oaxaca were of the genus Chryocerinae.41
4.2. EUKARYOTIC PROTEIN SYNTHESIS INHIBITION.
163
4.2
Eukaryotic Protein Synthesis Inhibition.
Dr Jerry Pelletier at McGill University (Montreal) tested salvinorins A-D (1a1d) and divinatorins A-C (28a-28c) for inhibition of eukaryotic protein synthesis. The assay used has been described in detail elsewhere.424 Briey, luciferase, the protein responsible for chemiluminescence in reies, is produced in vitro in a cell culture. Addition of a protein synthesis inhibitor (such as anisomycin) causes a reduction in light output relative to the control. The light output thus provides a simple measure of protein synthesis. The results are shown in Figure 4.1. Luciferase from two species (rey and Renilla) was used. None of the compounds tested caused signicant inhibition at 50 M (signicant inhibition is dened as a relative light output of 30% or less). The positive control, anisomycin, caused complete inhibition at 10 M.
anisomycin (+ve control) 1a
Firefly Renilla
1b 1c
1d
28a
28b
28c 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Relative light output (%)
Figure 4.1: Luciferase assay results for salvinorins and divinatorins (50 M).
164
4.3
4.3.1
Antimicrobial Activity.
Bacteria and Fungi.
O
R1
H R2 28a 28b 28b 29a
R1 OH OH H H
R2 H OH OAc H
R3 H Me H H
18
OR3
Figure 4.2: (-)-Hardwickiic acid and divinatorins A-C. (-)-Hardwickiic acid (29a) has been reported to display potent, broad-spectrum activity against bacteria and fungi.217 Divinatorins A-C (28a-28c) were therefore screened against standard antibiotic susceptible strains of Escherichia coli, Staphylococcus aureus, Bacillus subtilis, and Candida albicans, using broth microdilution425, 426 and disk diusion427 assays in each case.196 This work was done in collaboration with Professor Roy Robins-Browne and Andrea Bigham, in the Department of Microbiology and Immunology at the University of Melbourne. No activity was apparent against any of the test organisms at 100 g/mL or 100 g/disk. These data extend the stringent structure-activity requirements of 29a.217 In particular, a hydrogen bond donor at C-18 seems to be a necessary, but not sucient, condition of activity.217 To probe this further, we decided to screen (+)-hardwickiic acid (ent-29a). Ent-29a proved active against Staph. aureus (minimum inhibitory concentration (MIC) 25 g/mL) and B. subtilis (MIC 12.5 g/mL, 10 mm zone of inhibition), but much less potent than its enantiomer (MIC 0.78 g/mL against B. subtilis).217 The assays are shown in Figure 4.3. A very small zone of inhibition is apparent around ent-29a in the disk diusion assay. An undergraduate investigation published without peer review429 reported that
4.3. ANTIMICROBIAL ACTIVITY.

streptomycin sulphate (positive control) acetone (negative control)
165
(+)-hardwickiic acid (ent-29a) g/mL 100 50 25 12.5 6.25 3.12 1.56 0.78 0.39 0.19
E. coli
Staph. aureus
MIC
MIC
C. albicans
B. subtilis
(+)-hardwickiic acid (ent-29a)
Crude extract
B. subtilis
E. coli
S. aureus
Figure 4.3: Disk and microdilution assays for ent-29a and crude extract.428 the acetone extract of S. divinorum was active against a wide range of bacteria. We were unable to conrm these results. The acetone extract of the commercial material, as well as 1a, showed no activity at 100 g/mL or 100 g/disk. Apparently insucient 29a and oleanolic acid (31) were present to elicit an eect (31, like 29a, is active against B. subtilis and Staph. aureus).245
4.3.2
HIV-1.
Several opioids have been shown to inhibit HIV replication in vitro in several human cell types.430 They appear to act as viral entry inhibitors, by causing downregulation of the CXCR4 co-receptor, used by the virus to attach to the cell.431 HIV entry inhibitors are currently the focus of intensive research.432 The use of synergistic drug cocktails has already produced dramatic progress, and the addition of a further mechanism is expected to oer still greater synergy. KOR antagonists, which do not themselves inhibit viral replication, have been shown to enhance the eects of the standard therapy azidovudine (AZT).433 Beside their eect on viral replication, opioids also appear to counteract the
166
neurotoxic eects of HIV in infected cells.434, 435 Finally, since nausea is a common side eect of standard therapies,436 the possible antiemetic activity437, 438 of opioids may be advantageous.
4.3.2.1
NL4.3 and AD8 Strains.

107
HIV copies / 106 cells
HIV strain: NL4.3 AD8
106 105 104 103 102 101 0 2 4

Days post infection
1a (10-6 M) Me2SO only control
Figure 4.4: HIV-1 replication assays (NL43 and AD8 strains). Based on the above rationale, we submitted salvinorin A (1a) for testing against HIV in vitro. Testing was performed by Dr Sharon Lewin and Ajantha Solomon at Monash Universitys medical department (Alfred Hospital). Two strains of HIV-1 were tested: NL4.3 (T cell tropic) and AD8 (macrophage tropic). HIV was quantitated using real-time PCR.439, 440 Incubation of peripheral blood mononuclear cells (PBMCs) with 1 M salvinorin A had no eect on viral replication after one week (Figure 4.4). There was also no change in expression of CXCR4, CCR5 or CD4 receptors after one day, as determined using FACS staining.440
4.3.2.2
ROJO and TEKI Strains.
Two other HIV-1 strains were tested for the US National Institute of Allergy and Infectious Diseases (NIAID), under the direction of Dr Stephen Turk.
4.3. ANTIMICROBIAL ACTIVITY.
167
The isolates tested were ROJO (syncytium inducing/lymphocyte tropic) and TEKI (non-syncytium inducing/monocyte tropic). Again, salvinorin A (up to 230 M) had no eect on viral replication in PBMCs after one week of incubation (eg. Figure 4.5).441 The positive control, AZT, strongly inhibited viral replication (IC50 = 13 nM). Tests were performed in triplicate.
PERCENT OF VIRUS CONTROL
160 140 120 100 80 60 160
PERCENT OF CELL CONTROL
salvinorin A (1a)
140 120 100 80 60
% VIRUS CONTROL
40 20
% CELL CONTROL
40 20
0.00
140
0.02
0.07
0.23
0.73
2.30
7.27
23.0
72.7
230
140
CONCENTRATION (M)
PERCENT OF VIRUS CONTROL
120 100 80 60 40 20 0
AZT
PERCENT OF CELL CONTROL
120 100 80 60
% VIRUS CONTROL % CELL CONTROL
40 20 0 1000
0.00
0.10
0.32
1.00
3.16
10.0
31.6
100
316.2
CONCENTRATION (nM)
Figure 4.5: HIV-1 replication assays (ROJO isolate).
4.3.2.3
Discussion.
These results were surprising given the established activity of the opioids U50,488 and U69,593, which are of comparable potency to 1a. However, the extremely potent endogenous opioid dynorphin A showed no activity;442, 443 this has been speculatively attributed to metabolism by peptidases in the cell culture. Consistent with this, the potent fragment dynorphin A113 , which has a half life in plasma of less than one minute,444 caused negligible inhibition of HIV-1 replication.442 This suggests that the rapid metabolism87, 88 of 1a may account for its lack of activity. However, this hypothesis cannot account for other evidence. The intact peptide dynorphin A117 , which has a half life in plasma of three hours,444 also caused
168
no inhibition.442 Moreover, only brief exposure to U50,488 (< 30 minutes) is necessary to inhibit HIV-1 replication.431 Thus, there appear to be other factors at work. Wang et al found that 1a causes much less opioid receptor internalisation than U50,488, despite their similar potencies.81 Such dierences may involve the apparent subtypes445 of opioid receptors. There is an anecdotal report of improved health following S. divinorum use by an AIDS patient.446 While the above data suggest that 1a itself has no activity against HIV, oleanolic acid (31) is known to be active by two complementary mechanisms,243, 242 as discussed in Section 2.2.4.1 on page 96. Given a sucient dosage, therefore, the crude plant extract would probably exhibit some activity. No relevant data was located regarding other compounds in the plant.
4.4
NCI Anticancer Screen.
Salvinorins B and C (1b and 1c) and divinatorins A-C (28a-28c) were tested by the US National Cancer Institute, in the standard in vitro assay against 60 tumour cell lines.447 The standard measure of tumour cell growth inhibition is the GI50 the drug concentration at which growth is reduced to 50% of the control value. In most cases, this degree of inhibition was not achieved. Where substantial inhibition occurred, the GI50 was above 105 M in all cases (Figure 4.6). In the one apparent exception, 28b exhibited a GI50 of just under 106 M against the SF-539 brain tumour cell line. However, a signicant inhibition of growth was already apparent at 108 M (59% of control), which scarcely increased up to 105 M (45%), a 1000-fold increase in concentration (Figure 4.7). The GI50 value was therefore clearly artefactual. In all other cases, inhibition of growth increased sharply at higher doses: see Figure 4.6. GI50 values this high are strongly predictive of poor performance in subsequent assays,448 and the NCI therefore did not select the compounds for further testing. Salvinorin B (1b) completely inhibited growth, giving negative growth rates,
4.4. NCI ANTICANCER SCREEN.

All Cell Lines
100
169
50
divinatorin A (28a)
0
(NSC D737768)
-50
-100
All Cell Lines

100 100
-8
-7
-6
-5
-4
% Growth
50
50
salvinorin B (1b)
0 0
divinatorin B (28b)
(NSC D737715)
-50 -50
(NSC D737769)
-100 100 -8 -7 -6 -5 -4
-100 -8 100 -7 -6 -5 -4
50 % Growth
50
salvinorin C (1c)
0 0
divinatorin C (28c)
(NSC D737716)
-50 -50
(NSC D737770)
-100 -8 -7 -6 -5 Log10 Concentration (M) -4
Figure 4.6: NCI 60 cell line results for salvinorins and divinatorins in 8 cell lines. The Total Growth Inhibition concentration (TGI) was above 105 M in each case. No other compound gave this degree of inhibition. Only one cell line gave below -50% growth, the standard threshold for clear cytotoxicity: the SNB-75 CNS tumour line, with an LC50 slightly below 104 M (Figure 4.7). It is interesting that CNS tumour cells responded more strongly to 1b than the other categories. Nonetheless, the potencies even in these cases are unremarkable. Salvinorin A (1a) was not accepted even for in vitro testing. Given the ap-
170
100 100
50 % Growth
50
divinatorin B (28b)
0 0
salvinorin B (1b) (NSC D737715)

-50
(NSC D737769)
-50
-100 -8 -7 -6 -5 Log10 Concentration (M) SF-268 SNB-75 -4
CNS tumour cell line:
SF-295 U251
SF-539 SNB-19
Figure 4.7: CNS cell line results for divinatorin B and salvinorin B. parently rapid deacetylation of 1a in cell culture, however, it seems likely that the results would in any case have been close to those of 1b. Surprisingly, the reason for the rejection of 1a was that it had already been tested in vivo more than 20 years earlier.449 The results have apparently not been published previously. Using a standard protocol,450 mice were implanted with P388 tumour cells (lymphocytic leukaemia); test animals were injected with 1a daily for 5 days. Doses of 1a up to 76 mg/kg had no eect, while 152 mg/kg was slightly toxic (survival time 79% of control).451 The value of 0% given on the NCI website449 is a typographical error.451
4.5
Activity at the Opioid Receptor.
Previous work found that 1b was inactive at the opioid receptor, while replacement of the acetoxy group in 1a with more hindered esters dramatically reduced binding anity.135 No information was available about other functional groups. In order to further explore salvinorin As structure-activity relationships, isolated compounds and derivatives were tested in vitro for binding anity at cloned opioid receptors using radioligand assays. Those compounds with submicromolar anity were also screened for agonist potency and ecacy using
4.5. ACTIVITY AT THE OPIOID RECEPTOR.
171
a functional assay (see Experimental Section for details). The testing was done under the direction of Dr Bryan Roth, under the auspices of the National Institute of Mental Healths Psychoactive Drug Screening Program. No compound showed anity for or subtypes (K i > 1 M), and the following discussion will therefore discuss the opioid receptor exclusively. The raw data are shown in Table 4.2; structures can be found in the discussion below. The following discussion will express anities and potencies relative to 1a: Ki (x) Ki (1a) EC50 (x) EC50 (1a)
Krel (x) = and ECrel (x) =
A value of 10 thus signies a potency or anity 10 lower than 1a.

O
O O
O O
1a Ki 4 1 nM EC50 46 8 nM
Figure 4.8: KOR binding anity and potency of salvinorin A.
4.5.1
4.5.1.1
Other Salvinorins and Divinatorins.

Radioligand Binding Results.
There have been conicting results on salvinorin B (1b). Initial work by the Roth group found it inactive (K i > 10 M),135 but subsequent work by Bguin et al reported high anities (K i = 66,137 111271 or 155125 nM). Retesting by the Roth group indicated very weak anity (3.1 M versus 1a control 44 nM).452
172 Ki s.e.m. nM 1a 4 1 1b 3,153* 1c 1,022 262 1d >10,000 1e >10,000 28a >10,000 28b >10,000 28c >10,000 46 18 2 50 37a 163 36e 1,125 36c >10,000 51 156 18 11 35 59 49 6 1 50 6 2 53 77 347 81a 18 2 59 >10,000 60a 2,900 400 * = direct comparison K rel EC50 nM 46
CHAPTER 4. BIOASSAYS. s.e.m. ECrel E max % 100 s.e.m.
8 19 71* 255 >2,500 >2,500 >2,500 >2,500 >2,500 4.5 315 35 6.8 108 11 41 244 102 5.3 78 17 281 >2,500 39 126 36 2.7 108 14 15 78 21 1.7 107 5 1.5 223 60 4.8 103 13 1.5 624 200 14 116 10 87 >10,000 >217 4.5 141 43 3 122 27 >2,500 725 to 1a (44 nM) against [3 H]diprenorphine
Table 4.2: KOR radioligand and functional assay results.

O O O
O HO
O O R2O
OR1 H
O O
R1
H R2
O Krel
O R2 1c 1d 1e
O R1 Krel 255 >2,500 >2,500 28a 28b 28c
O R1
OR3 R2 R3 Krel
1b
71
Ac Ac H Ac Ac H
OH H H >2,500 OH OH Me >2,500 H OAc H >2,500
Figure 4.9: KOR binding anities of salvinorins and divinatorins. The discrepancy is not as great as it appears, since the two groups obtained dierent absolute anities for 1a. Bguin et al reported values of K rel = 66,137 85271 or 120.125 The Roth groups latest result (K rel = 71)452 is concordant.
173
The nonconcordant earlier result (K rel > 500) may have resulted from the use of a dierent radioligand [3 H]bremazocine rather than [3 H]diprenorphine. Moreover, in all cases the binding anity of 1b was negligible relative to 1a. Salvinorin C (1c) also showed negligible binding anity compared to 1a (K rel = 255, Figure 4.9). Salvinorins D (1d) and E (1e) and divinatorins A-C (28a28c) showed no anity (K rel > 2,500). Based on these results, we tentatively concluded270 that 1a is the sole opioid present in the plant. Recently, Lee et al have reported that salvinorins B (1b) and G (1g) and divinatorin D (28d, Figure 2.51 on page 118) bind to the KOR, but with much lower anities and potencies than 1a ( K rel = 66 418).137 Since the concentrations of these compounds are also orders of magnitude lower, their contribution to the activation of the KOR by S. divinorum is negligible. Nine other isolated compounds were inactive. Thus, these results strongly suggest that 1a is eectively the sole active principle of S. divinorum. The weak activity of 28d shows that the C ring can be cleaved without total loss of anity. Apart from this greater conformational freedom, 28d not only lacks the 2-acetoxy function, but also possesses a 1-hydroxy group, both of which drastically reduce activity in analogues of 1a, as will be discussed below. The structure-activity relationships of this compound are thus markedly dierent from those of 1a. Given that its deacetyl analogues 28b and 28e137 are inactive, perhaps the 17-acetoxy function substitutes for the furan ring in binding. However 28c, which also possesses this function, is inactive. Further exploration of this productive series of compounds is clearly warranted.
4.5.2
Modication of the Ketone.
Reduction of the ketone to an -hydroxy group (giving 36e) dramatically reduced binding anity (Figure 4.10). Acetylation (giving 36c) abolished binding entirely. These results were surprising, since molecular modelling had indicated that these modications would not aect binding.67 Although the very low anities of 1e, 1c, 36e, and 36c might appear to suggest that the
174
O O O O OH H H O O O O O H H O O
O
O O O H H
O O
36e Krel 281
36c Krel >2,500
81a Krel 4.5 ECrel 3
Figure 4.10: KOR activity after ketone modications. ketone is part of the pharmacophore, the relatively high anity and potency of 1-deoxy compound 81a show that it is not. This suggests that 1-hydroxy or acetoxy groups interact unfavourably with the KOR. Curiously, saturating the 3,4-double bond in 1e (inactive) gives 36e (active), while the opposite is true of 1c (active) and 36c (inactive). However, none of these compounds show appreciable (sub-micromolar) anity. The eect of this double bond itself therefore remains unclear.
4.5.3
Modication of the Acetoxy Group.

O O O
O O
O O O
OH
O O O O
O O
46 Krel 4.5 ECrel 6.8
59 Krel >2,500
60a Krel 725
Figure 4.11: KOR activity after acetoxy group modications. It had previously been shown that substituting more hindered esters for the 2-acetoxy group reduced binding anity.135 This suggested that the less-
175
hindered formate 46 (Figure 4.11 on the preceding page) might prove more potent, but in fact both anity and potency were reduced. The acetoxy group therefore appears to be the optimal alkyl chain length. The autoxidation products 59 and 60a were also screened. 59 was inactive, as Lee et al reported for the incorrect structure 62 (Figure 3.8 on page 139).271 Surprisingly, 60a showed weak anity at the KOR (K rel = 725). This provides further evidence that the 2-acyloxy function in 1a is not essential for binding, but that modifying this function usually reduces anity dramatically.
4.5.4
Modication of the Methyl Ester.

O O
O O
O O
O O
O O
OH
77 Krel 87 ECrel >217
OH
67a -
Figure 4.12: KOR activity after methyl ester modications. The role of the methyl ester in binding was strongly conrmed by the results for 18-hydroxy derivative 77, which appeared to be an antagonist, since it bound but did not activate the receptor. However, no functional tests of antagonist potency were performed. The precursor acid 67a was not tested, since the 8-epimers streaked on silica, so separation was not attempted. Lee et al subsequently separated the epimers using repeated chromatography. They reported, surprisingly, that the natural H-8 epimer was inactive at the KOR, while the H-8 epimer had high anity (K i = 48 nM).361
176
4.5.5
Modication of the Lactone.

O O O
O O
O OH
O O
O O
35 Krel 15 ECrel 1.7
49 Krel 1.5 ECrel 4.8
50 Krel 1.5 ECrel 14
Figure 4.13: KOR activity after lactone modications. The high anities of lactol 35, ether 49, and enol ether 50 (all full agonists) show that the lactone carbonyl is not essential for binding or activity. However, the latter two compounds, especially 50, showed reduced potency in the functional assay.
4.5.6
Modication of the Furan Ring.

O
13
O H O O H H H O O
O O
O
8
O O
37a Krel 41 ECrel 5.3
51 Krel 39 ECrel 2.7
Figure 4.14: KOR activity after furan modications. The substantially reduced anity of tetrahydrofuran 51 suggests that the furan ring is involved in binding. As described earlier, one 13-epimer of 51 was isolated for characterisation. The binding anity did not dier significantly from the 1:1 mixture (123 nM vs 156 nM). The reduced anity of
177
8-epi-salvinorin A (37a) also supports the role of the furan; since the lactone carbonyl is evidently not necessary for binding, the orientation of the furan is the salient dierence between this compound and 1a. The reduction in potency of these compounds was far smaller than the reduction in anity, particularly in 51.
4.5.7
Incorporation into a Revised Binding Model.

H H O N O Gln 115
Tyr 313
H O O O H H O O H O
Tyr 312
Tyr 139
Figure 4.15: Westkaempers original binding model. The original report of the activity of 1a at the KOR67 included a proposed model of its interactions with the receptor (Figure 4.15). As shown, the furan oxygen and the acetyl, methyl ester and lactone carbonyls act as hydrogen bond acceptors. The model, created by Dr Richard Westkaemper, was derived from a model of U69,593 binding reported previously. While the proposed roles of the furan, acetoxy group and methyl ester were consistent with our data above, the proposed interaction with the lactone seemed unlikely given the high anities of 35, 49 and 50 mentioned above. Westkaemper subsequently proposed a revised model,453 which drew upon the above data on the roles of each functional group in binding, although this was regrettably not acknowledged in the nished paper. Additional data on the role of the 2-acetoxy group was provided by others.453 Complementing these chemical data, site-directed mutagenesis in the Roth group allowed selective
178
modication of the receptor itself, identifying the residues involved in binding to 1a.453 The revised model is shown in Figure 4.16, visualised from coordinates kindly provided by Dr Westkaemper. For clarity, hydrogen atoms are not shown.
Tyr 313 Tyr 119
Tyr 320 - - - H-bonds*
Tyr 313 Tyr 119
Tyr 320 - - - H-bonds*
Ile 294 - - - hydrophobic interactions Glu 297 Glu 297
Ile 294 - - - hydrophobic interactions
*not present simultaneously
*not present simultaneously
Figure 4.16: Westkaempers revised binding model (stereoview).
The revised model has only one residue in common with the original: tyrosine 313. The proposed interaction is a hydrophobic one with the acetyl methyl group, however, rather than the original H bond to the carbonyl oxygen. This was suggested by the surprising result that mutation of this tyrosine residue to phenylalanine (lacking the hydroxyl) did not aect binding anity.453 Three of the ve proposed interactions in the new model are hydrophobic. The mutagenesis data suggest that the furan can form an H bond to either of tyrosines 199 or 320.453 While these bonds are both shown in Figure 4.16 for clarity, they cannot be present simultaneously, since only one nonbonded pair is available on the furan oxygen. One interesting aspect of this model is that three of the ve residues involved (isoleucine 294, glutamic acid 297 and tyrosine 313) are unique to the subtype. This provides a plausible explanation for the striking selectivity of
179
1a and analogues; very few derivatives show submicromolar anity or subtypes, and none of these are selective.25, 82, 136
4.5.8
Subsequent Results.
O O O O HO O HO H O O H H O
O RO
O O
O O
O O
O O
O O 84 O 85 0.31 0.13 O
OH
67a (H-8b)
34a ECrel 102
Krel >770 O O 86 Ki (m) 12 Krel 0.13 (m/k)
Krel ECrel
6 4
Figure 4.17: KOR binding anities and potencies of recent derivatives. Since publication of these results, numerous 1a analogues have been tested in vitro by other groups. These results again conrm that replacement of the 2-acetoxy group (with other esters, ethers, carbonates, amides etc) almost invariably reduces binding anity.95, 271, 82 There are exceptions, however. The high anity and potency of ethyl ether 8495 indicate that the acetyl carbonyl is not essential for activity. This result provides strong support for the hydrophobic interaction at C-2 in Westkaempers revised model (Figure 4.17). Another notable derivative is methoxymethyl ether 85,271 to date the only derivative signicantly more potent than 1a. Surprisingly, benzoate 8682 and some related derivatives activate the opioid receptor, but are not selective. Other results indicate that epimerisation at C-2 does not abolish activity in all derivatives.125 There is less data on modications at C-18,361, 95 but all esters and amides
180
tested showed large reductions in anity and potency. The loss of anity on O-demethylation (giving 67a)361 strongly supports the carbinol hydrophobic interaction in Westkaempers model over the more intuitive carbonyl H bond. Even less data has been reported on furan modications, but the greatly reduced potency of salvinicin A (34a)25 conrms that hindrance in this region interferes with binding. Some synthetic modications of the furan ring have been prepared,136 but biological data have not been reported.
Chapter 5 Experimental.
5.1
5.1.1
General Conditions
Instruments and Procedures.
Flash Chromatography: Flash column chromatography was performed according to Leonards194 procedure using Scharlau silica gel 60 (particle size 0.04 - 0.06 mm) or Merck silica gel 60. Mass ratios up to 400:1 versus crude product were used for dicult separations (hRf < 5). The activated carbon used was Merck aktivkohle 2183 powder (20:1 mass ratio versus crude product). HPLC: Spherex 5 m silica column (250 10 mm), ow rate 2 mL min1 with refractive index detection. Solvent front reached detector at 6.8 min. HRESIMS: Bruker 4.7T BiOAPEX FTMS. InChIs: IUPAC International Chemical Identiers454 were created using winChI version 1.455 IR: Bio-Rad FTS 165 FT-IR (running WIN-IR v 4.14) and Shimadzu FTIR 8400 (running HYPER IR v 1.57), using thin lms on NaCl discs. 181
182
CHAPTER 5. EXPERIMENTAL.
LCMS: Phenomenex Luna 3 m C-18 column (150 2 mm). Gradient elution: 0.05 % HCO2 H in 20-100% MeCN/H2 O over 40 minutes. ESI interface with ion trap detector. Detailed conditions have been published elsewhere.59 NMR: Varian Inova 400, Inova 500 and Unity Plus 400 (running Vnmr 6.1; some processing on VnmrJ 1.1D, some on iNMR). Bruker US2 800 (running TopSpin 1.3). 1 H NMR are 400 MHz, and less otherwise stated.
13 13
C NMR 100 MHz, un-
C multiplicities are based on DEPT experiments.
Assignments are given according to the standard32 clerodane numbering scheme below. Where 13 C assignments are given, all assignments (1 H and
13 1
C) are based on 2D NMR data (COSY, HMQC, HMBC). In other cases,
H assignments were made by comparison with related compounds for
which such data was available. Stereochemical assignments were based on coupling constants where possible, and NOESY data elsewhere (eg. pro-R vs. pro-S in rotatable methylenes). Peaks whose stereochemistry could not be unambiguously assigned on these bases are listed as a and b. Complex rst-order multiplets were analysed using Hoyes algorithms,456, 457 with resolution enhancement using the line-broadening window function where necessary (not shown in the reproduced spectra). Abbreviations: s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, quin = quintet, sext =sextet, sept = septet, b = broad. The values of solvent residual peaks (eg. CHCl3 = 7.26 ppm) and impurities were taken from Gottlieb et al.458
15 14 13 11 1 2 3 4 19 18 10 9 12 17 16
8 7
5 20 6
Polarimetry: JASCO DIP-1000. Concentration c is in g/l00 mL; the units
5.1. GENERAL CONDITIONS of the specic rotation are ( mLg1 dm1 ).
183
SEM: Samples were sputter-coated with gold. Fresh leaves were xed in 2.5% glutaraldehyde in 0.1 M phosphate buer at rt for 24 h. The samples were rinsed in the buer for 30 min ( 3) and then dehydrated in increasing concentrations of aq. EtOH for 30 min each (10, 30, 50, 70, 90, and 100%). After two further rinses in fresh 100% EtOH, the leaf samples were dried in a critical point dryer. TLC: Merck silica gel 60 F254 plates, visualised with phosphomolybdic acid in EtOH and heated unless otherwise indicated. hRf = Rf 100. UV: Shimadzu UV-2401PC (quartz cell).
5.1.2
Reagents.
DMPU and EtSH (Aldrich, 98%) were stored over 4 sieves. Petrol refers to the fraction boiling at 40-60 C. Reactions under Ar were performed using freshly distilled solvents unless otherwise indicated.
5.1.3
Plant Materials.
Commercial material Dried S. divinorum leaves, cultivated in Oaxaca Mexico, were purchased from Salvia Space Ethnobotanicals (Berkeley, California). Voucher specimens were deposited at the National Herbarium of Victoria (accession number MEL 2101361) and the University of Melbourne Herbarium (MELU s.n.) Australian material Additional S. divinorum plants were cultivated in Melbourne. A voucher specimen of the dried leaves was deposited at the National Herbarium of Victoria (accession number MEL 2145478).
184
Copaiba balsam was donated by Australian Botanical Products (Hallam, Victoria).
5.1.4
Assays.
Radioligand Binding Assays performed as previously detailed459, 67 using cloned receptors stably expressed in HEK 293 cells. : rat KORs with [3 H]diprenorphine (50 Ci/mmol, PerkinElmer Inc) or [3 H]U69,593 (41.4 Ci/mmol, PerkinElmer Inc) as radioligand. : human DORs with [3 H]DADLE (51.5 Ci/mmol, PerkinElmer Inc) as radioligand. : human MORs with [3 H]diprenorphine as radioligand. K i values were calculated using Prism 4.01 (GraphPad Software, Inc) as the mean SEM of quadruplicate (n 4) determinations. Nonspecic binding was dened using 10 M naloxone. Calcium Flux Functional Assay. Performed as previously detailed135, 460 using cloned rat KORs stably expressed in HEK 293 cells, cotransfected with the universal G protein G16 . Ca2+ mobilisation was quantied using a 96-well FlexStationII with the calcium ux assay kit (Molecular Devices Corp, Sunnyvale, CA). EC50 and Emax values were calculated using Prism 4.01 (GraphPad Software, Inc), as the mean SEM of quadruplicate (n 4) determinations. Antimicrobial Tests. The extract and compounds were tested against Escherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 29213), Bacillus subtilis (ATCC 6633), and Candida albicans (ATCC 90028) using standard broth microdilution426, 425 (100 - 0.19 g/mL using two-fold serial dilutions) and disc-diusion427 assays (100 g/disk). All measurements were performed in duplicate. Streptomycin sulfate and amphotericin B were used as positive controls.
5.2. ISOLATION
185
Insect Antifeedant Tests. A standard choice assay employing sweetened breglass discs423 was used, except that the discs were moistened. Final instar H. armigera larvae were selected on the basis of head size. Squares of Whatman GF/A lter (3.6 cm2 ) were treated with 100 L of 0.05 M sucrose solution. Test discs were then treated with 10 g of test compound in acetone, usually423 expressed as 100 L of a 100 ppm (100 mg L1 ) solution. Control discs were treated with the same volume of acetone. Discs were dried, numbered and weighed. Before performing the assay, the discs were moistened with distilled water. After more than half of one disc had been eaten, the discs were dried and reweighed. Tests were performed with 5 replicates per compound. HIV-1 Replication Assays.440 Alfred Hospital: PBMC were stimulated for 72 hours, and then infected with NL4.3 or AD8 (30 ng/ml virus) for 2 hours. Cells were then washed and seeded in 24 well plates. Salvinorin A was dissolved in Me2 SO before addition. Me2 SO at the same concentration was also compared to the control. Cells were lysed on day 0, 2, 5 and 7 of infection. HIV was quantitated using real-time PCR.439 NIAID/Southern Research Institute: detailed assay conditions can be found elsewhere.441
5.2
5.2.1
Isolation
Extraction of Commercial S. divinorum.
Dried S. divinorum leaves (860 g) were powdered and steeped for 1 h in acetone (3 1 L). Filtration and evaporation under reduced pressure gave a dark green tar (30.5 g). This was puried by ash column chromatography on an equal mixture of activated carbon and diatomite lter aid, eluting with a stepwise gradient from acetone to petrol, to give an amber semicrystalline syrup (5.73 g). Recrystallisations from MeOH and EtOH gave 1a (2.64 g). The mother liquor was puried by ash column chromatography on silica gel
186
(5-50% acetone/CH2 Cl2 gradient). This was divided, based on TLC (10% acetone/CH2 Cl2 ), into four series: A (656 mg), B (150 mg), C (359 mg) and D (77 mg). TLC: hRf of series 10% acetone/CH2 Cl2 A B C D 12
62 36-53 18
Series A: ash column chromatography, eluting with a gradient from 50-80% Et2 O/petrol, gave 1c (total yield 219 mg, 0.25 g/kg) and additional 1a, which was recrystallised from EtOH (total yield 2.9 g, 3.4 g/kg). Further elution gave 29a (7 mg). Series B: ash column chromatography on silica gel (35 g) in 70-90% Et2 O/petrol, and recrystallisation from MeOH, gave 1b (13 mg, 0.015 g/kg). Series C: Trituration in hot Et2 O gave 1d (75 mg). Flash column chromatography of the mother liquor (60-100% Et2 O/petrol) gave four fractions based on TLC (70% Et2 O/petrol): C1 (55 mg), C2 (119 mg), C3 (57 mg) and C4 (39 mg). TLC: hRf of fraction 70% Et2 O/petrol C1 C2 C3 17 C4 13
40-43 28-21
Fraction C1: Repeated ash column chromatography (20% acetone/petrol and 40-60% Et2 O/petrol) gave 28c (23 mg) and 31 (3 mg). Fraction C2: Repeated ash column chromatography (25% acetone/petrol and 60-100% Et2 O/petrol) gave 28b (41 mg). Fraction C3: Extensive ash column chromatography (Et2 O/petrol, acetone/petrol and EtOAc/petrol) gave additional 28b (total yield 41 mg) and a mixture of 1e and 1f. Final purication by HPLC (60% EtOAc/petrol) gave 1e (2.8 mg) and 1f (1.1 mg). Fraction C4 gave additional 1d (total yield 114 mg, 0.13 g/kg). Series D: Repeated ash column chromatography (60% Et2 O/petrol and 4% MeOH/CH2 Cl2 ) gave 28a (36 mg).
5.2. ISOLATION
187
5.2.2
Extraction of Australian S. divinorum.
Dried, powdered S. divinorum leaves (224 g) were steeped in acetone for 30 min (3 250 mL). Filtration and evaporation under reduced pressure gave a dark green tar (7 g). This was puried by vacuum ltration through a 1:1 mixture of activated carbon (75 g) and diatomite lter aid, eluting with a gradient from 50-20% EtOAc/petrol, to give series E (97 mg) and F (279 mg) based on TLC (70% Et2 O/petrol). TLC: hRf of series 70% Et2 O/petrol E F
55-68 16
Series E: Repeated ash column chromatography (1% acetone/CH2 Cl2 and 20% Et2 O/petrol) gave 33 (1 mg). Further ash column chromatography (0.75% MeOH/CH2 Cl2 and 1% EtOH/CHCl3 ) gave 32 (23 mg) and 30 (12 mg). Series F: Two recrystallisations from MeOH gave 1a (126 mg). yield (mg/kg) Salvinorin A B C D E F Divinatorin A B C ()-Hardwickiic acid Oleanolic acid Presqualene alcohol Peplusol (E)-Phytol 1a 1b 1c 1d 1e 1f 28a 28b 28c 29a 31 32 33 30 3,400 15 254 132 3 1 42 48 27 8 3 3 4 3 70% Et2 O/ 10% acetone/ petrol CH2 Cl2 24 14 31 18 23 24 37 31 50 64 66 44 75 59 57 37 60 25 47 40 15 31 39 45 65 34 73 58
Table 5.1: Yields and TLC data (hRf ) of isolated compounds.
188
5.2.3
Salvinorin A (1a).
InChI=1/C23H28O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8,11,14-17,19H,5,7,9-10H2,14H3/t14-,15-,16-,17-,19-,22-,23-/m0/s1
O O
O O
1a
colourless crystals, mp (from EtOH) 236-238 C; lit. (from MeOH) 238-240 C;22 TLC: See Table 5.1 on the previous page. UV (MeCN): max (log ) 208 (3.76) nm; lit.23 UV (MeOH): max (log ) 211 (3.72) nm;
1
H NMR (800 MHz, CDCl3 ): 7.41 (1H, dt, J = 1.7, 0.9 Hz, H-16), 7.38 (1H,
t, J = 1.7 Hz, H-15), 6.37 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.52 (1H, ddd, J = 11.7, 5.2, 0.8 Hz, H-12), 5.14 (1H, ddt, J 11.9, 9.0, 0.9 Hz, H-2), 3.72 (3H, s, CO2 CH 3 ), 2.74 (1H, dd, J 11.3, 5.6 Hz, H-4), 2.50 (1H, dd, J = 13.5, 5.2 Hz, H-11), 2.31-2.28 (2H, m, H-3), 2.18 (1H, br s, H-10), 2.17-2.14 (1H, m, H-7), 2.16 (3H, s, OCOCH 3 ), 2.07 (1H, dd, J = 12.0, 3.1 Hz, H-8), 1.79 (1H, dt, J = 13.4, 3.1 Hz, H-6), 1.64 (1H, tdd, J = 13.5, 12.1, 3.4 Hz, H-7), 1.58 (1H, td, J = 13.5, 0.9 Hz, H-6), 1.57 (1H, ddd, J = 13.5, 11.7, 0.8 Hz, H-11), 1.45 (3H, s, H-20), 1.11 (3H, s, H-19);
13
C NMR (CDCl3 ) data matched previously reported values.22
5.2. ISOLATION
189
5.2.4
Salvinorin B (1b).
InChI=1/C21H26O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7,10,12-15,17,22H,4,6,8-9H2,1-3H3/ t12-,13-,14-,15-,17-,20-,21-/m0/s1

O
O HO
O O
1b
colourless crystals, mp (from hot MeOH) 239-240 C; (from cold Et2 O/petrol) 244-245 C; lit. (from hot MeOH) 213-216 C;23 251-254 C;15, 260 (from cold MeOH) 211214 C;169
1
H and
13
5.2.5
Salvinorin C (1c).
InChI=1/C25H30O9/c1-13(26)32-18-10-17(22(28)30-5)24(3)8-6-16-23(29)34-19 (15-7-9-31-12-15)11-25(16,4)21(24)20(18)33-14(2)27/h7,9-10,12,16,18-21H,6,8, 11H2,1-5H3/t16-,18-,19-,20-,21-,24-,25-/m0/s1

O O O O O H H O O
1c
190 clear resin; TLC: See Table 5.1 on page 187. []16 +70 (c 0.6, CHCl3 ); D []22 +49 (c 0.6, CHCl3 ) lit;27 D UV (MeCN): max (log ) 208 (4.10) nm;
FTIR (lm): max 3145, 2952, 1740, 1643, 1506, 1435, 1372, 1315, 1226, 1174, 1142, 1075, 1041, 961, 950, 910, 875, 789, 754, 695, 667 cm1 ;
1
H and
13
5.2.6
Salvinorin D (1d).
InChI=1/C23H28O8/c1-12(24)30-18-16(25)9-15(20(26)28-4)22(2)7-5-14-21(27) 31-17(13-6-8-29-11-13)10-23(14,3)19(18)22/h6,8-9,11,14,16-19,25H,5,7,10H2, 1-4H3/t14-,16-,17-,18-,19-,22-,23-/m0/s1

O O O HO H H O O
1d
ne colourless crystals, mp (from cold Et2 O) 185-187 C; TLC: See Table 5.1 on page 187. []17 +67 (c 1.0, CH2 Cl2 ); D FTIR (lm): max 3475, 3146, 2952, 2861, 1723, 1505, 1435, 1371, 1315, 1228, 1195, 1174, 1142, 1071, 1056, 1027, 949, 875, 788, 744, 699 cm1 ;
1
H NMR (CDCl3 ): 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
6.54 (1H, dd, J = 2.4, 1.3 Hz, H-3), 6.40 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.70
5.2. ISOLATION
191
(1H, dt, J = 5.5, 1.3 Hz, H-1), 5.53 (1H, dd, J = 11.2, 5.8 Hz, H-12), 4.44 (1H, ddd, J = 6.7, 5.5, 2.4 Hz, H-2), 3.74 (3H, s, CO2 CH 3 ), 2.56 (1H, dt, J = 13.3, 3.4 Hz, H-6), 2.54 (1H, dd, J = 13.1, 5.9 Hz, H-11), 2.17-2.09 (1H, m, H-7), 2.15 (3H, s, OCOCH 3 ), 2.13 (1H, dd, J = 13.5, 3.5 Hz, H-8), 2.01 (1H, br d, J = 6.7 Hz, OH ), 1.78 (1H, dtd, J = 15.1, 13.3, 3.5 Hz, H-7), 1.69 (3H, s, H-19), 1.64 (1H, ddd J = 13.1, 11.2, 1.0 Hz, H-11), 1.42 (1H, br s, H-10), 1.22 (3H, s, H-20), 1.20 (1H, td, J = 13.5, 3.5 Hz, H-6);
13
C NMR (CDCl3 ): 171.57 (C, C-17/OC OCH3 ), 171.55 (C, C-17/OC OCH3 ),
166.2 (C, C-18), 143.8 (CH, C-15), 141.2 (C, C-4), 139.4 (CH, C-16), 135.7 (CH, C-3), 125.4 (C, C-13), 108.4 (CH, C-14), 71.6 (CH, C-12), 68.6 (CH, C-2), 66.5 (CH, C-1), 52.5 (CH, C-10), 51.8 (CH, C-8), 51.7 (CH3 , CO2 C H3 ), 43.9 (CH2 , C-11), 37.6 (C, C-5), 37.03 (C, C-9), 36.97 (CH2 , C-6), 21.6 (CH3 , C-19), 21.2 (CH3 , OCOC H3 ), 18.3 (CH2 , C-7), 15.6 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 455.1672 (calcd for C23 H28 O8 Na+ , 455.1676).
5.2.7
Salvinorin E (1e).
InChI=1/C23H28O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8-9,11,14,16-19,25H,5,7,10H2, 1-4H3/t14-,16-,17-,18-,19-,22-,23-/m0/s1

O
O O
OH H
O O
1e
clear resin; TLC: See Table 5.1 on page 187. HPLC: t R (min) 60% EtOAc/petrol 1e 1f
9.8 10.7
192 []17 +46 (c 0.14, CHCl3 ); D
FTIR (lm): max 3510, 3144, 2952, 2858, 1722, 1641, 1505, 1436, 1374, 1316, 1228, 1142, 1070, 1029, 949, 935, 897, 875, 805, 788, 763, 708, 690, 679 cm1 ;
1
6.43 (1H, dd, J = 2.4, 1.6 Hz, H-3), 6.41 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.60 (1H, ddd, J = 11.0, 5.8, 0.9 Hz, H-12), 5.40 (1H, dd, J = 4.9, 2.4 Hz, H-2), 4.46 (1H, ddd, J = 4.7, 1.6, 1.3 Hz, H-1), 3.73 (3H, s, CO2 CH 3 ), 2.52 (1H, ddd, J = 12.5, 3.0, 2.6 Hz, H-6), 2.46 (1H, dd, J = 13.1, 6.0 Hz, H-11), 2.17 (3H, s, OCOCH 3 ), 2.18-2.07 (2H, m, H-7 & 8), 1.94 (1H, dd, J = 2.3, 1.5 Hz, OH ), 1.84 (1H, dtd, J = 14.2, 12.0, 3.7 Hz, H-7), 1.72 (3H, s, H-19), 1.62 (1H, dd, J = 13.1, 11.2 Hz, H-11), 1.47 (3H, s, H-20), 1.30 (1H, br s, H-10), 1.19 (1H, td, J = 13.3, 3.7 Hz, H-6);
13
C NMR (CDCl3 ): 171.8 (C, OC OCH3 ), 169.8 (C, C-17), 166.0 (C, C-18),
143.9 (CH, C-15), 143.4 (C, C-4), 139.3 (CH, C-16), 131.5 (CH, C-3), 125.8 (C, C-13), 108.4 (CH, C-14), 72.3 (CH, C-2), 71.7 (CH, C-12), 64.3 (CH, C1), 54.0 (CH, C-10), 51.8 (CH3 , CO2 C H3 ), 51.7 (CH, C-8), 44.4 (CH2 , C-11), 37.8 (C, C-5), 37.5 (C, C-9), 37.0 (CH2 , C-6), 21.9 (CH3 , C-19), 21.0 (CH3 , OCOC H3 ), 18.4 (CH2 , C-7), 16.2 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 455.1687 (calcd for C23 H28 O8 Na+ , 455.1676).
5.2.8
Salvinorin F (1f).
InChI=1/C21H26O6/c1-20-8-6-14-19(24)27-16(12-7-9-26-11-12)10-21(14,2)17 (20)15(22)5-4-13(20)18(23)25-3/h4,7,9,11,14-17,22H,5-6,8,10H2,1-3H3/t14-,15 -,16-,17-,20-,21-/m0/s1
5.2. ISOLATION
O
193
OH H
O O
1f
clear resin; TLC: See Table 5.1 on page 187. HPLC: see table on page 191. []16 20 (c 0.05, CHCl3 ); D FTIR (lm): max 3514, 3147, 2951, 2857, 1712, 1637, 1505, 1436, 1372, 1318, 1232, 1194, 1144, 1070, 1028, 978, 947, 896, 875, 797, 756, 686 cm1 ;
1
6.67 (1H, ddd, J = 4.7, 3.0, 1.0 Hz, H-3), 6.41 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.60 (1H, ddd, J = 11.2, 5.8, 0.9 Hz, H-12), 4.51 (1H, br dd, J = 5.5, 4.7 Hz, H-1), 3.72 (3H, s, CO2 CH 3 ), 2.60 (1H, ddd, J = 20.1, 5.5, 3.0 Hz, H-2), 2.53 (1H, dt, J = 13.2, 3.2 Hz, H-6), 2.46 (1H, dd, J = 13.2, 5.8 Hz, H-11), 2.35 (1H, ddt, J = 20.1, 4.6, 1.1 Hz, H-2), 2.17-2.08 (2H, m, H-7, 8), 1.82 (1H, dtd, J = 15.0, 13.2, 3.5 Hz, H-7), 1.71 (3H, s, H-19), 1.62 (1H, ddd, J = 13.2, 11.4, 0.9 Hz, H-11), 1.48 (3H, s, H-20), 1.29 (1H, dd, J = 4.7, 0.7 Hz, OH ), 1.25 (1H, br s, w 1 = 3.8 Hz, H-10), 1.18 (1H, tdd, J = 13.3, 3.6, 0.9 Hz,
2
H-6);
13
C NMR (CDCl3 ): 172.1 (C, C-17), 166.9 (C, C-18), 143.8 (CH, C-15), 140.6
(C, C-4), 139.3 (CH, C-16), 133.4 (CH, C-3), 125.9 (C, C-13), 108.4 (CH, C14), 71.7 (CH, C-12), 63.9 (CH, C-1), 54.8 (CH, C-10), 52.2 (CH, C-8), 51.5 (CH3 , CO2 C H3 ), 44.4 (CH2 , C-11), 38.0 (CH2 , C-2), 37.7 (C, C-9), 37.3 (CH2 , C-6), 36.6 (C, C-5), 21.6 (CH3 , C-19), 18.6 (CH2 , C-7), 16.4 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 397.1610 (calcd for C21 H26 O6 Na+ , 397.1622).
194
5.2.9
Divinatorin A (28a).
InChI=1/C20H28O4/c1-13-6-9-20(3)15(18(22)23)4-5-16(21)17(20)19(13,2)107-14-8-11-24-12-14/h4,8,11-13,16-17,21H,5-7,9-10H2,1-3H3,(H,22,23)/t13-,16+, 17-,19+,20+/m1/s1/f/h22H
O
OH H
OH
28a
amber resin; TLC: See Table 5.1 on page 187. []19 53 (c 1.8, CH2 Cl2 ); D FTIR (lm): max 3392, 2927, 2874, 2648, 1684, 1634, 1503, 1456, 1411, 1386, 1262, 1245, 1221, 1162, 1102, 1066, 1025, 1003, 966, 924, 894, 874, 782, 760, 703, 670 cm1 ;
1
H NMR (CDCl3 ): 7.36 (1H, t, J = 1.6 Hz, H-15), 7.20 (1H, td, J = 2.5,
1.0 Hz, H-16), 6.90 (1H, ddd, J = 4.8, 2.7, 1.0 Hz, H-3), 6.25 (1H, dt, J = 1.8, 1.0 Hz, H-14), 4.49 (1H, br d, J = 4.8 Hz, H-1), 2.56 (1H, ddd, J = 20.1, 5.1, 2.8 Hz, H-2), 2.43-2.33 (2H, m, H-2, 6), 2.34 (1H, br td, J = 13.5, 4.2 Hz, H-12-pro-R), 2.05 (1H, br ddd, J = 14.3, 12.9, 4.7 Hz, H-12-pro-S ), 1.85 (1H, ddd, J = 14.7, 12.8, 4.7 Hz, H-11-pro-S ), 1.67 (1H, ddd, J = 14.8, 12.8, 4.4 Hz, H-11-pro-R), 1.64 (3H, s, H-19), 1.60-1.54 (2H, m, H-7, 8), 1.47-1.42 (1H, m, H-7), 1.45 (1H, br s, H-10), 1.23-1.17 (1H, m, H-6), 1.15 (3H, s, H-20), 0.84 (3H, d, J = 6.2 Hz, H-17);
13
C NMR (CDCl3 ): 171.8 (C, C-18), 142.8 (CH, C-15), 140.8 (C, C-4), 138.4
(CH, C-16), 136.2 (CH, C-3), 125.2 (C, C-13), 110.9 (CH, C-14), 64.7 (CH, C-1), 49.0 (CH, C-10), 39.7 (C, C-9), 39.1 (CH2 , C-11), 38.6 (CH2 , C-6), 38.1
5.2. ISOLATION
195
(CH2 , C-2), 37.4 (C, C-5), 37.1 (CH, C-8), 27.4 (CH2 , C-7), 21.4 (CH3 , C-19), 19.8 (CH3 , C-20), 18.2 (CH2 , C-12), 15.7 (CH3 , C-17); HRESIMS: [M + Na]+ m/z 355.1864 (calcd for C20 H28 O4 Na+ , 355.1880).
5.2.10
Divinatorin B (28b).
InChI=1/C21H30O5/c1-20(9-6-14-8-11-26-13-14)15(12-22)7-10-21(2)16(19(2 4)25-3)4-5-17(23)18(20)21/h4,8,11,13,15,17-18,22-23H,5-7,9-10,12H2,1-3H3/ t15-,17-,18+,20-,21-/m0/s1

O
OH H
H OH
28b
amber resin; TLC: See Table 5.1 on page 187. []20 54 (c 2.1, CHCl3 ); D FTIR (lm): max 3434, 2930, 2881, 1714, 1503, 1461, 1437, 1385, 1356, 1236, 1196, 1162, 1097, 1067, 1025, 979, 943, 921, 874, 781, 759, 734 cm1 ;
1
H NMR (500 MHz, CDCl3 ): 7.35 (1H, t, J = 1.7 Hz, H-15), 7.20 (1H, dd, J
= 1.6, 1.0 Hz, H-16), 6.65 (1H, ddd, J = 4.9, 2.8, 1.0 Hz, H-3), 6.25 (1H, dd, J = 1.8, 0.9 Hz, H-14), 4.46 (1H, dq, J = 5.1, 1.3 Hz, H-1), 3.84 (1H, dd, J = 10.5, 3.5 Hz, H-17-pro-R), 3.71 (3H, s, CO2 CH 3 ), 3.38 (1H, dd, J = 10.5, 8.0 Hz, H-17-pro-S ), 2.53 (1H, ddd, J = 19.9, 5.1, 2.8 Hz, H-2), 2.42 (1H, dddd, J = 14.5, 12.7, 4.7, 1.1 Hz, H-12-pro-R), 2.36 (1H, dt, J = 13.0, 3.4 Hz, H-6), 2.31 (1H, ddt, J = 19.9, 4.9, 1.4 Hz, H-2), 2.08 (1H, dddd, J = 14.5, 12.5, 4.8, 1.1 Hz, H-12-pro-S ), 1.90 (1H, ddd, J = 15.0, 12.4, 4.9 Hz, H-11-pro-S ), 1.85 (1H, dq, J = 13.2, 3.4 Hz, H-7), 1.77 (1H, ddd, J = 15.0, 12.6, 4.6 Hz,
196
H-11-pro-R), 1.66 (3H, s, H-19), 1.64-1.49 (2H, m, H-8 & 7), 1.49 (2H, br s, OH ), 1.45 (1H, br s, H-10), 1.19-1.13 (1H, m, H-6), 1.18 (3H, s, H-20);
13
C NMR (CDCl3 ): 167.3 (C, C-18), 142.8 (CH, C-15), 141.4 (C, C-4), 138.4
(CH, C-16), 133.2 (CH, C-3), 124.9 (C, C-13), 110.8 (CH, C-14), 64.3 (CH, C1), 63.9 (CH2 , C-17), 51.3 (CH3 , CO2 C H3 ), 48.7 (CH, C-10), 44.8 (CH, C-8), 39.1 (C, C-9), 38.8 (CH2 , C-11), 37.97 (CH2 , C-2/6), 37.96 (CH2 , C-2/6), 37.1 (C, C-5), 21.9 (CH2 , C-7), 21.4 (CH3 , C-19), 20.9 (CH3 , C-20), 18.2 (CH2 , C-12); HRESIMS: [M + Na]+ m/z 385.1988 (calcd for C21 H30 O5 Na+ , 385.1985).
5.2.11
Divinatorin C (28c).
InChI=1/C22H30O5/c1-15(23)27-14-17-8-11-22(3)18(20(24)25)5-4-6-19(22)21 (17,2)10-7-16-9-12-26-13-16/h5,9,12-13,17,19H,4,6-8,10-11,14H2,1-3H3,(H,24, 25)/t17-,19+,21-,22-/m0/s1/f/h24H

O
H O
OH
28c
amber resin; TLC: See Table 5.1 on page 187. []25 110 (c 1.1, CHCl3 ); D FTIR (lm): max 2960, 2938, 2873, 2650, 1738, 1681, 1629, 1502, 1459, 1420, 1385, 1367, 1236, 1161, 1064, 1025, 1000, 975, 873, 785, 759, 730 cm1 ;
1
H NMR (500 MHz, CDCl3 ): 7.35 (1H, t, J = 1.6 Hz, H-15), 7.22 (1H, dd,
J = 1.6, 0.8 Hz, H-16), 6.89 (1H, dd, J = 4.8, 2.9 Hz, H-3), 6.28 (1H, dd, J
5.2. ISOLATION
197
= 1.8, 0.9 Hz, H-14), 4.26 (1H, dd, J = 11.0, 4.1 Hz, H-17-pro-R), 3.79 (1H, dd, J = 11.0, 8.4 Hz, H-17-pro-S ), 2.53 (1H, dt, J = 13.2, 3.2 Hz, H-6), 2.40 (1H, td, J = 13.8, 4.2 Hz, H-12-pro-R), 2.35 (1H, dt, J = 20.2, 5.3 Hz, H-2), 2.24-2.16 (2H, m, H-2 & 12-pro-S ), 2.03 (3H, s, OCOCH 3 ), 1.82-1.67 (4H, m, H-1, 7, 8, 11-pro-S ), 1.64 (1H, ddd, J = 15.0, 12.6, 4.3 Hz, H-11-pro-R), 1.54-1.44 (2H, m, H-1 & 7), 1.42 (1H, br d, J = 12.1 Hz, H-10), 1.27 (3H, s, H-19), 1.15 (1H, td, J = 13.2, 3.6 Hz, H-6), 0.83 (3H, s, H-20);
13
C NMR (CDCl3 ): 171.9 (C, C-18), 171.2 (C, OC OCH3 ), 142.8 (CH, C-15),
141.2 (C, C-4), 140.3 (CH, C-3), 138.5 (CH, C-16), 125.2 (C, C-13), 110.9 (CH, C-14), 66.1 (CH2 , C-17), 46.8 (CH, C-10), 40.9 (CH, C-8), 38.9 (CH2 , C-11), 38.4 (C, C-9), 37.4 (C, C-5), 35.2 (CH2 , C-6), 27.4 (CH2 , C-2), 22.3 (CH2 , C-7), 21.0 (CH3 , OCOC H3 ), 20.5 (CH3 , C-19), 19.0 (CH3 , C-20), 18.3 (CH2 , C-12), 17.0 (CH2 , C-1); HRESIMS: [M + Na]+ m/z 397.1989 (calcd for C22 H30 O5 Na+ , 397.1985).
5.2.12
()-Hardwickiic Acid (29a) and methyl ester 29b.
InChI=1/C20H28O3/c1-14-7-10-20(3)16(18(21)22)5-4-6-17(20)19(14,2)11-815-9-12-23-13-15/h5,9,12-14,17H,4,6-8,10-11H2,1-3H3,(H,21,22)/t14-,17-,19+, 20+/m1/s1
O
H R 29a 29b H Me
O
1
OR
H NMR (CDCl3 ) matched previously reported values;216, 217 the spectrum
was superimposable with that of authentic (+)-hardwickiic acid (ent-29a), prepared as detailed in Section 5.3.5 on page 202. Methylation with CH2 N2 in Et2 O gave the methyl ester 29b;
198
InChI=1/C21H30O3/c1-15-8-11-21(3)17(19(22)23-4)6-5-7-18(21)20(15,2)129-16-10-13-24-14-16/h6,10,13-15,18H,5,7-9,11-12H2,1-4H3/t15-,18-,20+,21+ /m1/s1 []18 86 (c 0.04, CHCl3 ); D []23 104 (c 1.1, CHCl3 ) lit;217 D
1
H and
13
C NMR (CDCl3 ), FTIR (lm) and EIMS (70 eV) matched previ-
ously reported values.218, 217 The 1 H NMR spectrum was superimposable with that of authentic (+)-methyl hardwickiate (ent-29b), prepared as detailed in Section 5.3.5 on page 202.
5.2.13
Oleanolic Acid (31).
InChI=1/C30H48O3/c1-25(2)14-16-30(24(32)33)17-15-28(6)19(20(30)18-25) 8-9-22-27(5)12-11-23(31)26(3,4)21(27)10-13-29(22,28)7/h8,20-23,31H,9-18H2, 1-7H3,(H,32,33)/t20-,21-,22+,23-,27-,28+,29+,30-/m0/s1
H OH H HO H O 31
HRESIMS: [M + Na]+ m/z 479.3516 (calcd for C30 H48 O3 Na+ , 479.3496);
1
H and
13
C NMR (CDCl3 ) matched previously reported values.221
5.2.14
Presqualene Alcohol (32).
InChI=1/C30H50O/c1-23(2)13-9-15-25(5)17-11-18-27(7)21-28-29(22-31)30(28, 8)20-12-19-26(6)16-10-14-24(3)4/h13-14,17,19,21,28-29,31H,9-12,15-16,18,20,22 H2,1-8H3/b25-17+,26-19+,27-21+/t28-,29-,30-/m1/s1
5.3. SYNTHESIS
H
199
OH
32
[]21 +45 (c 1.2, CHCl3 ); D []20 +49 (c 4.8, CHCl3 ) lit;223 D

1
H NMR (CDCl3 ),222
13
C NMR (C6 D6 )223 and FTIR (lm)222 matched previ-
ously reported values. The 1 H NMR spectrum in C6 D6 was superimposable with a previously published spectrum.223
5.2.15
Peplusol (33).
InChI=1/C30H50O/c1-24(2)13-9-15-26(5)17-11-18-28(7)21-22-30(23-31)29(8) 20-12-19-27(6)16-10-14-25(3)4/h13-14,17,19,21,30-31H,8-12,15-16,18,20,22-23 H2,1-7H3/b26-17+,27-19+,28-21+/t30-/m0/s1
OH
33
[]20 6 (c 0.07, CHCl3 ); D []25 18 (c 0.74, i PrOH) lit;224 D

1
H and
13
C NMR (CDCl3 ) and FTIR (lm) matched previously reported val-
ues.224
5.3
5.3.1
Synthesis
Salvinorin C (1c) via acetylation of salvinorin D (1d).
Ac2 O (250 L, 2.6 mmol) was added to a solution of 1d (11.0 mg, 25.4 mol) in dry pyridine (2.5 mL) under Ar. After stirring for 3.5 h, TLC
200
(10% acetone/CH2 Cl2 ) indicated completion. The reaction mixture was diluted with ice water and extracted with Et2 O ( 3). The organic phase was washed with saturated NaHCO3 , saturated CuSO4 , water and brine. Drying (MgSO4 ), evaporation in vacuo and ash column chromatography on silica gel (55% Et2 O/petrol) gave 1c as a clear resin (8.4 mg, 70%); TLC: See Table 5.1 on page 187. Cospotted with isolated material. []16 +69 (c 0.4, CHCl3 ). D Other spectra (FTIR, 1 H and isolated material.
13
C NMR) superimposable with those of the
5.3.2
Salvinorins D (1d) and E (1e) via acetylation of 1h.
1h (2.8 mg, 7.2 mol), Ac2 O (7 L, 74 mol) and catalytic DMAP (
1 mg)
were stirred in pyridine (3 mL) for 3 h, when TLC (10% acetone/CH2 Cl2 , visualised with KMnO4 ) showed no starting material. Evaporation in vacuo gave a 1:2 mixture of 1d and 1e. TLC: See Table 5.1 on page 187 and table on the facing page. Cospotted with isolated materials.
1
H NMR spectrum superimposable with those of the isolated materials.
5.3.3
Salvinorins C (1c) and E (1e) via acetylation of 1h.
1h (4.1 mg, 10.5 mol), Ac2 O (8 L, 85 mol) and catalytic DMAP (
mg) were stirred in pyridine (500 L) at 45 C for 75 min, when TLC (10% acetone/CH2 Cl2 , visualised with KMnO4 /H2 SO4 dip) showed no starting material. Evaporation in vacuo and ash column chromatography (2.5 10% acetone/CH2 Cl2 ) gave 1c and 1e; the latter was contaminated by an inseparable byproduct.
5.3. SYNTHESIS
201
TLC: See Table 5.1 on page 187 and table on this page. Cospotted with isolated materials.
1
H NMR spectra superimposable with those of the isolated materials.
5.3.4
Dideacetylsalvinorin C (1h) from 1c.
To a solution of 1c (5.8 mg, 12.2 mol) and Na2 CO3 (5.1 mg, 41.1 mol) in CH2 Cl2 (1 mL) was added MeOH (1 mL), and the solution stirred at rt for 2 h, when TLC (10% acetone/ CH2 Cl2 ) showed considerable starting material. After heating at 45 C for a further 90 min, TLC indicated completion. The solution was partitioned between brine (acidied with 10% HCl) and CH2 Cl2 ( 3). Drying (MgSO4 ), evaporation in vacuo, and ash column chromatography (loaded in CH2 Cl2 , eluted with 33 - 50% EtOAc/petrol, then 25% MeOH/ CH2 Cl2 ) gave 1h as a resin (4.1 mg, 86%); InChI=1/C21H26O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7-8,10,12,14-17,22-23H,4,6,9H2,1-3H3 /t12-,14-,15-,16-,17-,20-,21-/m0/s1
O
HO
OH H
O O
1h
TLC: hRf 10% acetone/CH2 Cl2
1c 60
1h 18
[]18 +27 (c 0.2, CH2 Cl2 ); D FTIR (lm): max 3456, 2951, 1714, 1504, 1435, 1379, 1314, 1229, 1177, 1144, 1075, 1049, 1027, 949, 875, 788, 736, 685 cm1 ;
202
1
H NMR (CDCl3 ): 7.43 (1H, m, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15), 6.48
(1H, dd, J = 2.4, 1.6 Hz, H-3), 6.40 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.60 (1H, dd, J = 11.1, 5.9 Hz, H-12), 4.32 (1H, br d, J = 5.1 Hz, H-1), 4.28 (1H, dd, J = 5.1, 2.4 Hz, H-2), 3.73 (3H, s, CO2 CH 3 ), 2.49 (1H, dd, J = 13.2, 6.0 Hz, H-11), 2.50-2.45 (1H, m, H-6), 2.40-2.30 (2H, m, OH ), 2.14-2.10 (1H, m, H-8), 2.09 (1H, dq, J = 14.6, 3.6 Hz, H-7), 1.82 (1H, dtd, J = 15.0, 13.2, 3.3 Hz, H-7), 1.70 (3H, s, H-19), 1.60 (1H, ddd, J = 13.0, 11.1, 0.8 Hz, H-11), 1.47 (3H, s, H-20), 1.22 (1H, d, J = 1.0 Hz, H-10), 1.16 (1H, tdd, J = 13.3, 3.6, 0.9 Hz, H-6);
13
C NMR (CDCl3 ): 172.0 (C, C-17), 166.6 (C, C-18), 143.9 (CH, C-15), 142.2
(C, C-4), 139.3 (CH, C-16), 135.3 (CH, C-3), 125.8 (C, C-13), 108.4 (CH, C14), 71.8 (CH, C-12), 69.6 (CH, C-2), 65.6 (CH, C-1), 54.1 (CH, C-10), 51.8 (CH, C-8), 51.7 (CH3 , CO2 C H3 ), 44.4 (CH2 , C-11), 37.49 (C, C-5/9), 37.45 (C, C-5/9), 37.0 (CH2 , C-6), 22.1 (CH3 , C-19), 18.4 (CH2 , C-7), 16.3 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 413.1588 (calcd for C21 H26 O7 Na+ , 413.1571).
5.3.5
(+)-Hardwickiic acid (ent-29a).
Following a modied version of Costa et als procedure,218 the acid fraction of copaiba balsam was methylated with CH2 N2 in Et2 O, and the methyl ester (ent-29b) was isolated. This ester (32 mg, 97 mol) was dissolved in acetone. KF/Al2 O3 (220 mg, 40% w/w KF) was added and the acetone evaporated under reduced pressure. This powder was irradiated in a microwave oven (650 W) at 100% power for 8 minutes, then cooled. Minimal water was added and stirred for 5 minutes, then ltered, and the lter cake was rinsed with water ( 2). Rinsing of the lter cake with CHCl3 ( 3) gave, after drying (MgSO4 ) and evaporation, starting material ent-29b (11 mg). The aqueous ltrate was acidied with 10% HCl and extracted with CHCl3 ( 4). The pooled organic extracts were dried (MgSO4 ) and evaporated to give 14 mg crude product.
After D2 O exchange.
5.3. SYNTHESIS
203
Flash column chromatography in 80% Et2 O/petrol gave ent-29a (5 mg, 16 mol) as a semicrystalline lm; InChI=1/C20H28O3/c1-14-7-10-20(3)16(18(21)22)5-4-6-17(20)19(14,2)11-815-9-12-23-13-15/h5,9,12-14,17H,4,6-8,10-11H2,1-3H3,(H,21,22)/t14-,17-,19+, 20+/m0/s1/f/h21H
O
H R
OR
ent-29a H ent-29b Me
ent-29a 31
ent-29b 76
[]19 +81 (c 0.2, CHCl3 ); D []23 85 (CHCl3 ) lit. value for 29a;217 D FTIR, 1 H and
13
C NMR data217 matched reported values.
5.3.6
Salvinorin A lactol (35).
Following the published procedure,260 1a (15.8 mg, 36.5 mol) was warmed in THF (1 mL) under Ar until fully dissolved, then cooled to -78 C. i Bu2 AlH (1M in THF, 0.5 mL, 500 mol) was added dropwise. The solution was stirred for 25 minutes, then quenched (sat. aq. NH4 Cl dropwise), evaporated in vacuo until thick, diluted in water and extracted into Et2 O ( 3). Washing (brine), drying (MgSO4 ), evaporation in vacuo, and ash column chromatography (610% acetone/CH2 Cl2 gradient) gave 35260 as a clear resin (10.4 mg, 65% (81% borsm)); InChI=1/C23H30O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8,11,14-17,19,21,27H,5,7,9-10H 2,1-4H3/t14-,15-,16-,17-,19-,21u,22-,23-/m0/s1
204
O
O O
O OH
35
1a 57
35 25
FTIR (lm): max 3446, 2953, 2256, 1730, 1503, 1438, 1379, 1272, 1237, 1203, 1161, 1123, 1053, 1010, 978, 914, 875, 785, 732, 647 cm1 ;
1
H NMR, major (17) anomer (CDCl3 ): 7.36 (1H, br s, H-16), 7.34 (1H, t,
J = 1.8 Hz, H-15), 6.38 (1H, dd, J = 1.8, 0.9 Hz, H-14), 5.14-5.09 (1H, m, H-2), 4.87 (1H, dd, J = 11.6, 2.4 Hz, H-12), 4.80 (1H, d, J = 8.7 Hz, H-17), 3.70 (3H, s, CO2 CH 3 ), 2.76-2.72 (1H, m, H-4), 2.28-2.23 (2H, m, H-3), 2.14 (3H, s, OCOCH 3 ), 2.11 (1H, dd, J = 13.2, 2.4 Hz, H-11), 2.07 (1H, d, J = 0.9 Hz, H-10), 1.80 (1H, dq, J = 13.9, 3.3 Hz, H-7), 1.70 (1H, dt, J = 13.5, 3.2 Hz, H-6), 1.64-1.57 (1H, m, H-6), 1.38 (3H, s, H-20), 1.43-1.32 (1H, m, H-7), 1.21 (1H, ddd, J = 13.2, 11.6, 0.9 Hz, H-11), 1.15-1.09 (1H, m, H-8), 1.08 (3H, s, H-19); these data are broadly consistent with incomplete data published previously;260
13
C NMR, major (17) anomer (CDCl3 ): 202.5 (C, C-1), 171.9 (C, C-18),
169.9 (C, OC OCH3 ), 143.0 (CH, C-15), 139.1 (CH, C-16), 126.2 (C, C-13), 108.8 (CH, C-14), 94.2 (CH, C-17), 75.0 (CH, C-2), 66.2 (CH, C-12), 65.4 (CH, C-10), 53.6 (CH, C-4), 52.1 (CH, C-8), 51.8 (CH3 , CO2 C H3 ), 44.7 (CH2 , C-11), 42.4 (C, C-5), 38.8 (CH2 , C-6), 35.6 (C, C-9), 30.8 (CH2 , C-3), 20.6 (CH3 , OCOC H3 ), 17.6 (CH2 , C-7), 16.7 (CH3 , C-19), 15.0 (CH3 , C-20).
5.3. SYNTHESIS
205
5.3.7
(4R)-3,4-Dihydrosalvinorin C (36c).
Following a published procedure,27 formation of the orthoacetate of 36h (which did not go to completion in 10 h), followed by acid-catalyzed hydrolysis, gave 36d which was used in the next reaction without purication; InChI=1/C23H30O8/c1-12(24)30-18-16(25)9-15(20(26)28-4)22(2)7-5-14-21(27) 31-17(13-6-8-29-11-13)10-23(14,3)19(18)22/h6,8,11,14-19,25H,5,7,9-10H2,1-4H3 /t14-,15-,16-,17-,18-,19-,22-,23-/m0/s1
O O O HO H H O O
36d
TLC: hRf Et2 O

1
36h 33
Orth.Ac. 60
36d 29
6.41 (1H, br d, J = 2.0 Hz, H-14), 5.61 (1H, br s, H-1), 5.47 (1H, dd, J = 11.5, 5.3 Hz, H-12), 3.76-3.72 (1H, m, H-2), 3.69 (3H, s, CO2 CH 3 ), 2.49 (1H, dd, J = 13.1, 5.6 Hz, H-11), 2.22-2.08 (3H, m), 2.16 (3H, s, OCOCH 3 ), 1.86-1.77 (3H, m), 1.71-1.61 (2H, m), 1.35 (3H, s, H-20), 1.35-1.24 (1H, m), 1.18 (3H, s, H-19), 1.13 (1H, d, J = 1.9 Hz, H-10);
1
H NMR [(CD3 )2 CO] matched values previously reported for CDCl3 .27
Acetylation of 36d with Ac2 O/pyridine27 and purication by HPLC (40% EtOAc/ petrol) gave 36c; InChI=1/C25H32O9/c1-13(26)32-18-10-17(22(28)30-5)24(3)8-6-16-23(29)3419(15-7-9-31-12-15)11-25(16,4)21(24)20(18)33-14(2)27/h7,9,12,16-21H,6,8,1011H2,1-5H3/t16-,17-,18-,19-,20-,21-,24-,25-/m0/s1
206
O O O O O H H O O
36c
36d 33
36c 77 (developed in KMnO4 ).
HPLC: t R = 18.5 min (40% EtOAc/petrol).

1
H NMR (CDCl3 ): 7.45 (1H, br s, H-16), 7.41 (1H, t, J = 1.8 Hz, H-15), 6.41
(1H, dd, J = 2.0, 0.9 Hz, H-14), 5.67 (1H, dt, J = 3.5, 1.5 Hz, H-1), 5.46 (1H, dd, J = 11.6, 5.5 Hz, H-12), 4.79-4.74 (1H, m, H-2), 3.70 (3H, s, CO2 CH 3 ), 2.43 (1H, dd, J = 13.3, 5.4 Hz, H-11), 2.31-2.24 (2H, m), 2.16-2.09 (2H, m), 2.14 (3H, s, OCOCH 3 ), 1.98 (3H, s, OCOCH 3 ), 1.84-1.79 (2H, m), 1.72-1.53 (m, obscured by H2 O), 1.69 (1H, ddd, J = 13.4, 11.6, 0.9 Hz, H-11), 1.40-1.31 (2H, m), 1.38 (3H, s, H-20), 1.20 (1H, d, J = 1.9 Hz, H-10), 1.17 (3H, s, H-19); these data are broadly consistent with those published previously82 (assignments dier). 1 H NMR [(CD3 )2 CO]23 and viously reported values.
13
C NMR (CDCl3 )27 matched pre-
5.3.8
(4R)-3,4-Dihydrosalvinorin E (36e).
Following the published procedure,23 acetylation in Ac2 O/pyridine27 of 36h (which had been puried by HPLC in EtOAc) gave 36e; InChI=1/C23H30O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8,11,14-19,25H,5,7,9-10H2,1-4 H3/t14-,15-,16-,17-,18-,19-,22-,23-/m0/s1
5.3. SYNTHESIS
O
207
O O
OH H
O O
36e
TLC: hRf 50% EtOAc/petrol

1
36h 20
36e 48 (developed in KMnO4 ).
6.41 (1H, dd, J = 1.8, 0.9 Hz, H-14), 5.55 (1H, dd, J = 11.4, 5.4 Hz, H-12), 4.70 (1H, ddd, J = 11.7, 4.6, 3.2 Hz, H-2), 4.29 (1H, br s, H-1), 3.68 (3H, s, CO2 CH 3 ), 2.41 (1H, dd, J = 13.2, 5.4 Hz, H-11), 2.31 (1H, q, J = 12.6 Hz, H-3), 2.20(1H, dd, J = 13.2, 2.4 Hz, H-4), 2.14-2.08 (1H, m, H-8), 2.09 (3H, s, OCOCH 3 ), 1.90 (1H, br s, w 1 = 11 Hz, OH ), 1.82 (1H, dddd, J = 12.5,
2
4.9, 2.5, 1.1 Hz, H-3), 1.76 (1H, dq, J = 13.3, 3.2 Hz, H-7), 1.74-1.69 (1H, m, H-6a), 1.64 (1H, ddd, J = 13.0, 11.6, 0.9 Hz, H-11), 1.45 (3H, s, H-20), 1.38 (3H, s, H-19), 1.38-1.26 (2H, m, H-6b,7), 1.00 (1H, br s, H-10); these data are consistent with incomplete data published previously.23 NMR (CDCl3 ) matched previously reported values.27
13
5.3.9
(4R)-Dideacetyl-3,4-dihydrosalvinorin C (36h).
A slight modication of a published procedure was used:23, 260 to 1a (37.6 mg, 86.9 mol) and NaBH4 (4.4 mg, 116 mol) was added CH2 Cl2 (200 L), followed by EtOH (3 mL), and the cloudy solution stirred under Ar at 40 C. After 4 h, TLC (Et2 O) indicated completion. The solution was cooled to 0
C, and 0.5% H2 SO4 /MeOH added dropwise until eervescence ceased. The
solution was concentrated to 500 L in vacuo, then partitioned between brine (acidied with 10% HCl) and CH2 Cl2 ( 3). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (40-66% EtOAc/petrol) gave 36h15, 23 (19.1 mg, 56%);
208
InChI=1/C21H28O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7,10,12-17,22-23H,4,6,8-9H2,1-3H3/t 12-,13-,14-,15-,16-,17-,20-,21-/m0/s1

O
HO
OH H
O O
36h
TLC: hRf Et2 O
1a 45
36h 14
38h 25
HPLC: t R = 8.7 min (EtOAc).

1
6.41 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.56 (1H, ddd, J = 11.5, 6.1, 0.9 Hz, H-12), 4.20 (1H, br s, H-1), 3.68 (3H, s, CO2 CH 3 ), 3.58 (1H, ddd, J = 11.6, 5.2, 3.4 Hz, H-2), 2.48 (1H, dd, J = 13.2, 5.6 Hz, H-11), 2.25-2.07 (4H, m), 1.91 (2H, br s, w 1 = 70 Hz, OH ), 1.77-1.69 (3H, m), 1.62 (1H, ddd, J = 13.3,
2
11.6, 1.0 Hz, H-11), 1.46 (3H, s, H-20), 1.37 (3H, s, H-19), 1.29 (1H, tdd, J = 13.6, 3.1, 1.0 Hz, H-6), 0.92 (1H, d, J = 1.9 Hz, H-10);
1
H NMR ([CD3 ]2 CO) matched previously reported values.23
5.3.10
8-epi-Salvinorin A (37a).
Distilled DMPU (60 C / 0.1 mmHg) was added to 1a (21.4 mg, 49.5 mol) and NaHCO3 (30.1 mg, 358 mol), and stirred at 150 C for 2 h. The amber solution was diluted with EtOAc, neutralised dropwise with 10% HCl, and washed (10% HCl 4, then brine). Drying (MgSO4 ) and evaporation in vacuo followed by ash column chromatography (30% - 50% EtOAc/petrol gradient) monitored by TLC (Et2 O) gave 37a as a clear resin (10.8 mg, 51% (81% borsm));
5.3. SYNTHESIS
209
InChI=1/C23H28O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8,11,14-17,19H,5,7,9-10H2,1-4 H3/t14-,15+,16+,17+,19+,22+,23+/m1/s1

O
O O
H
8
O O
37a
TLC: hRf Et2 O vanillin/H2 SO4
1a 39 purple
37a 50 blue
[]13 53 (c 0.6, CHCl3 ); D FTIR (lm): max 3018, 2951, 2883, 1732, 1504, 1450, 1437, 1375, 1323, 1238, 1202, 1161, 1124, 1084, 1047, 1024, 997, 970, 937, 876, 783, 756, 667, 601 cm1 ;
1
6.37 (1H, d, J = 1.7 Hz, H-14), 5.25 (1H, dd, J = 12.0, 2.2 Hz, H-12), 5.09 (1H, dd, J 10.9, 9.3 Hz, H-2), 3.69 (3H, s, CO2 CH 3 ), 2.79-2.72 (1H, m, H-4), 2.45 (1H, dd, J = 5.0, 2.2 Hz, H-8), 2.36 (1H, dd, J = 15.0, 2.2 Hz, H-11), 2.29-2.22 (2H, m, H-3), 2.24 (1H, br s, H-10), 2.19 (1H, dq, J = 14.3, 3.1 Hz, H-7), 2.15 (3H, s, OCOCH 3 ), 2.00 (1H, td, J = 13.7, 3.9 Hz, H-6), 1.83 (1H, tdd, J = 14.2, 5.0, 3.9 Hz, H-7), 1.62 (3H, s, H-20), 1.54 (1H, dt, J = 13.7, 3.4 Hz, H-6), 1.50 (1H, dd, J = 15.0, 12.0 Hz, H-11), 1.07 (3H, s, H-19);
13
C NMR (CDCl3 ): 202.3 (C, C-1), 173.4 (C, C-17), 171.8 (C, C-18), 169.8
(C, OC OCH3 ), 143.6 (CH, C-15), 139.7 (CH, C-16), 123.3 (C, C-13), 108.5 (CH, C-14), 75.2 (CH, C-2), 70.1 (CH, C-12), 64.1 (CH, C-10), 52.9 (CH, C4), 51.8 (CH3 , CO2 C H3 ), 48.0 (CH2 , C-11), 45.2 (CH, C-8), 42.2 (C, C-5), 34.7 (C, C-9), 33.9 (CH2 , C-6), 30.6 (CH2 , C-3), 24.6 (CH3 , C-20), 20.5 (CH3 , OCOC H3 ), 17.6 (CH2 , C-7), 15.2 (CH3 , C-19);
210
HRESIMS: [M + Na]+ m/z 455.1683 (calcd for C23 H28 O8 Na+ , 455.1676).
5.3.11
8-epi-Salvinorin B (37b).
MeOH (2 mL) was reuxed under Ar for 15 minutes, and cooled to rt. 1a (32.3 mg, 74.7 mol) and Na2 CO3 (30.4 mg, 245 mol) were added, and the resulting suspension stirred under Ar at rt for 4.5 h, when TLC indicated completion. The reaction was quenched with 10% HCl, then partitioned between 10% HCl and CH2 Cl2 ( 4). Drying (MgSO4 ) and evaporation in vacuo gave an owhite powder. Addition of minimal MeOH ( 0.3 mL) and centrifugation gave 1b as a white powder (21.5 mg, 74%). The supernatant was evaporated under reduced pressure to give 37b as an amber resin (6.6 mg, 23%); InChI=1/C21H26O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7,10,12-15,17,22H,4,6,8-9H2,1-3H3/t 12-,13+,14+,15+,17+,20+,21+/m1/s1
O
O HO
H
8
O O
37b
TLC: hRf Et2 O 50% EtOAc/petrol 10% acetone/CH2 Cl2 vanillin/H2 SO4
1a 40 37 69 purple
1b 24 23 52 purple
37b 40 37 54 blue
[]13 27 (c 0.3, CHCl3 ); D FTIR (lm): max 3475, 2952, 1728, 1504, 1438, 1386, 1274, 1201, 1156, 1120, 1051, 1026, 997, 970, 876, 786, 755 cm1 ;
5.3. SYNTHESIS
1
211
(1H, dd, J = 1.9, 0.8 Hz, H-14), 5.29 (1H, dd, J = 11.8, 2.2 Hz, H-12), 4.00 (1H, ddd, J = 11.9, 7.6, 1.3 Hz, H-2), 3.69 (3H, s, CO2 CH 3 ), 2.70 (1H, dd, J = 13.5, 3.2 Hz, H-4), 2.47 (1H, dd, J = 4.8, 2.2 Hz, H-8), 2.44 (1H, ddd, J = 13.5, 7.6, 3.2 Hz, H-3), 2.42 (1H, dd, J = 14.8, 2.2 Hz, H-11), 2.23 (1H, d, J = 1.3 Hz, H-10), 2.21 (1H, dq, J = 14.4, 3.1 Hz, H-7), 2.01 (1H, td, J = 13.5, 11.9 Hz, H-3), 1.99 (1H, td, J = 13.8, 0.8 Hz, H-6), 1.85 (1H, tdd, J = 14.0, 4.8, 3.7 Hz, H-7), 1.65 (3H, s, H-20), 1.54 (1H, dt, J = 13.5, 3.5 Hz, H-6), 1.46 (1H, ddd, J = 14.8, 11.7, 0.7 Hz, H-11), 1.06 (3H, s, H-19); these data are inconsistent with those of Harding et al 82 and Lee et al.271
13
C NMR (CDCl3 ): 209.1 (C, C-1), 173.4 (C, C-17), 172.1 (C, C-18), 143.7
(CH, C-15), 139.6 (CH, C-16), 123.5 (C, C-13), 108.4 (CH, C-14), 74.5 (CH, C2), 70.0 (CH, C-12), 63.7 (CH, C-10), 52.4 (CH, C-4), 51.7 (CH3 , CO2 C H3 ), 48.3 (CH2 , C-11), 45.3 (CH, C-8), 42.7 (C, C-5), 34.6 (C, C-9), 34.3 (CH2 , C-3), 33.9 (CH2 , C-6), 24.7 (CH3 , C-20), 17.6 (CH2 , C-7), 15.3 (CH3 , C-19); HRESIMS: [M + Na]+ m/z 413.1573 (calcd for C21 H26 O7 Na+ , 413.1571).
5.3.12
8-epi-Salvinorin C (37c).
1 mg)
37d (1.7 mg, 3.9 mol), Ac2 O (0.25 mL) and catalytic DMAP (
were stirred in pyridine (0.5 mL) at 50 C for 90 min, when TLC (10% acetone/CH2 Cl2 , visualised with KMnO4 /H2 SO4 dip) showed no starting material. In a separate ask, 37e ( 2.8 mg, 6.5 mol) was subjected to the same conditions for 4 h, when TLC (same conditions) showed no starting material. The crude products, which cospotted by TLC, were pooled and evaporated in vacuo. Flash column chromatography (2-4% acetone/CH2 Cl2 ) gave 37c (4.7 mg, 95%) as a colourless resin; InChI=1/C25H30O9/c1-13(26)32-18-10-17(22(28)30-5)24(3)8-6-16-23(29)3419(15-7-9-31-12-15)11-25(16,4)21(24)20(18)33-14(2)27/h7,9-10,12,16,18-21H,6, 8,11H2,1-5H3/t16-,18+,19+,20+,21+,24+,25+/m1/s1
212
O O O O O H H
8
O O
37c
TLC: hRf 3% acetone/CH2 Cl2 vanillin/H2 SO4
37c 35 blue
37d 8 blue
37e 26 blue
[]23 +9 (c 0.2, CH2 Cl2 ); D FTIR (lm): max 2926, 2854, 1741, 1716, 1645, 1556, 1504, 1435, 1371, 1332, 1229, 1198, 1172, 1159, 1093, 1032, 962, 875, 802 cm1 ;
1
H NMR (CDCl3 ): 7.50 (1H, dd, J = 1.6, 0.8 Hz, H-16), 7.43 (1H, t, J =
1.7 Hz, H-15), 6.42 (2H, m, H-3 & 14), 5.59 (1H, ddd, J = 4.7, 1.3, 0.9 Hz, H-1), 5.49 (1H, dd, J = 4.8, 2.3 Hz, H-2), 5.28 (1H, br d, J = 11.0 Hz, H-12), 3.72 (3H, s, CO2 CH 3 ), 2.49 (1H, dd, J = 5.3, 2.3 Hz, H-8), 2.33 (1H, dt, J = 13.4, 3.4 Hz, H-6), 2.26-2.16 (1H, m, H-7), 2.14 (1H, dd, J = 14.3, 1.6 Hz, H-11), 2.10 (3H, s, OCOCH 3 ), 2.02 (3H, s, OCOCH 3 ), 1.97 (1H, tdd, J = 14.8, 5.5, 3.9 Hz, H-7), 1.70 (3H, s, H-19), 1.61-1.50 (m, H-11, 6, H2 O), 1.35 (3H, s, H-20), 1.25 (1H, br s, H-10);
13
C NMR (CDCl3 ): 173.6, 170.5, 169.7, 165.7, 143.7, 142.6, 139.7, 132.1,
123.4, 108.4, 69.7, 69.6, 65.1, 51.74, 51.72, 48.9, 45.6, 38.2, 36.0, 33.5, 25.1, 21.14, 21.13, 20.6, 18.2; HRESIMS: [M + Na]+ m/z 497.1788 (calcd for C25 H30 O9 Na+ , 497.1782).
5.3.13
8-epi-Salvinorin D (37d).
Standing 37e in CDCl3 at rt overnight gave a 1:1 mixture with 37d, which was isolated by ash column chromatography (5-10% acetone/CH2 Cl2 ) as an amber resin;
5.3. SYNTHESIS
213
InChI=1/C23H28O8/c1-12(24)30-18-16(25)9-15(20(26)28-4)22(2)7-5-14-21(2 7)31-17(13-6-8-29-11-13)10-23(14,3)19(18)22/h6,8-9,11,14,16-19,25H,5,7,10 H2,1-4H3/t14-,16+,17+,18+,19+,22+,23+/m1/s1
O O O HO H H
8
O O
37d
TLC: See table on the preceding page. []23 +15 (c 0.1, CH2 Cl2 ); D FTIR (lm): max 3402, 2953, 2926, 1738, 1716, 1503, 1462, 1440, 1382, 1368, 1333, 1302, 1228, 1199, 1173, 1158, 1099, 1055, 1025, 981, 875, 842, 798, 784, 732 cm1 ;
1
H NMR (500 MHz, CDCl3 ): 7.49 (1H, br s, H-16), 7.43 (1H, t, J = 1.8
Hz, H-15), 6.49 (1H, dd, J = 2.3, 1.5 Hz, H-3), 6.41 (1H, dd, J = 2.0, 0.8 Hz, H-14), 5.53 (1H, br d, J = 5.0 Hz, H-1), 5.29 (1H, br d, J = 11.5 Hz, H-12), 4.41 (1H, dd, J = 5.1, 1.4 Hz, H-2), 3.72 (3H, s, CO2 CH 3 ), 2.48 (1H, dd, J = 5.3, 2.3 Hz, H-8), 2.30 (1H, dt, J = 13.5, 3.8 Hz, H-6), 2.22 (1H, dtd, J = 14.7, 3.5, 2.1 Hz, H-7), 2.17 (1H, dd, J = 14.1, 1.5 Hz, H-11), 2.13 (3H, s, OCOCH 3 ), 1.96 (1H, tdd, J = 14.2, 5.3, 3.9 Hz, H-7), 1.70-1.64 (1H, m, H-11), 1.67 (3H, s, H-19), 1.62-1.52 (m, H-6, H2 O), 1.37 (3H, s, H-20), 1.25 (1H, br s, H-10);
13
C NMR (CDCl3 ): 173.7, 171.8, 166.1, 143.7, 141.4, 139.7, 135.0, 123.6,
108.4, 69.6, 69.3, 67.6, 51.7, 51.6, 49.1, 45.7, 37.8, 36.0, 33.5, 25.2, 21.3, 21.0, 18.1; HRESIMS: [M + Na]+ m/z 455.1681 (calcd for C23 H28 O8 Na+ , 455.1676).
214
5.3.14
8-epi-Salvinorin E (37e).
37h (6.6 mg, 16.9 mol) was dissolved in pyridine (2 mL) and Ac2 O (250 L), and stirred at rt for 3 h, when TLC (10% acetone/CH2 Cl2 ) showed no starting material. H2 O (10 mL) was added, and the solution extracted with Et2 O ( 2). The pooled organic layers were washed (sat. NaHCO3 , sat. CuSO4 , and brine). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (80% Et2 O/petrol) gave 37e (5.3 mg, 72%) as an amber resin; InChI=1/C23H28O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h6,8-9,11,14,16-19,25H,5,7,10H2,1 -4H3/t14-,16+,17+,18+,19+,22+,23+/m1/s1
O
O O
OH H
H
8
O O
37e
TLC: See tables on page 212 and on the next page. []22 +14 (c 0.1, CH2 Cl2 ); D FTIR (lm): max 3514, 2951, 1737, 1722, 1503, 1461, 1435, 1373, 1322, 1257, 1227, 1199, 1155, 1126, 1092, 1047, 1031, 931, 900, 837, 843, 806, 780, 735 cm1 ;
1
H NMR (500 MHz, CDCl3 ): 7.49 (1H, br s, H-16), 7.42 (1H, t, J = 1.9 Hz,
H-15), 6.41 (1H, dd, J = 2.0, 1.0 Hz, H-14), 6.39 (1H, dd, J = 2.4, 1.5 Hz, H-3), 5.35 (1H, dd, J = 4.7, 2.3 Hz, H-2), 5.30 (1H, br d, J = 11.5 Hz, H-12), 4.29 (1H, dd, J = 4.4, 1.7 Hz, H-1), 3.71 (3H, s, CO2 CH 3 ), 2.49 (1H, dd, J = 5.2, 2.1 Hz, H-8), 2.36 (1H, dt, J = 13.3, 3.6 Hz, H-6), 2.21 (1H, dtd, J = 14.5, 3.7, 2.1 Hz, H-7), 2.14 (3H, s, OCOCH 3 ), 2.12 (1H, dd, J = 13.7, 1.5 Hz, H-11), 1.99 (1H, tt, J = 14.3, 4.4 Hz, H-7), 1.89 (1H, dd, J = 2.5, 1.3
5.3. SYNTHESIS
215
Hz, OH ), 1.70 (3H, s, H-19), 1.64 (1H, dd, J = 13.9, 11.5 Hz, H-11), 1.62 (3H, s, H-20), 1.54 (1H, td, J = 13.1, 3.4 Hz, H-6), 1.30 (1H, t, J = 0.9 Hz, H-10);
13
C NMR (CDCl3 ): 174.0, 169.7, 166.0, 143.7, 143.4, 139.6, 131.2, 123.8,
108.4, 72.5, 69.6, 65.0, 53.1, 51.6, 49.4, 45.6, 37.9, 36.2, 33.4, 25.8, 21.2, 21.0, 18.1; HRESIMS: [M + Na]+ m/z 455.1675 (calcd for C23 H28 O8 Na+ , 455.1676).
5.3.15
8-epi-Dideacetylsalvinorin C (37h).
1c (11.0 mg, 23.2 mol) and KCN (7.5 mg, 115 mol) were dissolved in MeOH (1.5 mL) and stirred at rt for 30 min, when TLC (10% acetone/CH2 Cl2 ) showed no starting material. The solution was evaporated in vacuo, and the residue partitioned between H2 O and CH2 Cl2 ( 3). Recrystallisation from acetone, repeated on the mother liquor, gave 37h (6.6 mg, 73%) as colourless crystals; InChI=1/C21H26O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7-8,10,12,14-17,22-23H,4,6,9H2,1-3H3 /t12-,14+,15+,16+,17+,20+,21+/m1/s1
O
HO
OH H
H
8
O O
37h
TLC: hRf 10% acetone/CH2 Cl2 50% EtOAc/petrol vanillin/H2 SO4
1c 63 purple
37h 23 28 blue
1h 18 19 purple
Use of this poisonous reagent is not recommended. K2 CO3 and other bases are equally eective.
216 mp (from acetone) 243-245 C; []22 +15 (c 0.04, CH2 Cl2 ); D
FTIR (lm): max 3461, 2952, 2926, 1716, 1507, 1459, 1437, 1378, 1259, 1228, 1200, 1180, 1156, 1095, 1052, 875, 803 cm1 ;
1
(1H, dd, J = 2.4, 1.6 Hz, H-3), 6.40 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.30 (1H, br d, J = 11.5 Hz, H-12), 4.25 (1H, dd, J = 4.9, 2.4 Hz, H-2), 4.16 (1H, ddd, J = 4.9, 1.6, 0.7 Hz, H-1), 3.71 (3H, s, CO2 CH 3 ), 2.48 (1H, dd, J = 5.5, 2.2 Hz, H-8), 2.23 (1H, dt, J = 13.5, 2.2 Hz, H-6), 2.23-2.17 (1H, m, H-7), 2.15 (1H, dd, J = 13.9, 1.5 Hz, H-11), 1.98 (1H, tdd, J = 14.5, 5.3, 3.8 Hz, H-7), 1.81 (2H, br s, w 1 50 Hz, OH ), 1.68 (3H, s, H-19), 1.64 (3H, s, H-20), 1.62
2
(1H, dd, J = 13.8, 12.1 Hz, H-11), 1.51 (1H, td, J = 12.9, 3.8 Hz, H-6), 1.21 (1H, br s, H-10);
13
C NMR (CDCl3 ): 174.2, 166.4, 143.7, 142.2, 139.6, 134.9, 123.9, 108.4,
69.9, 69.6, 66.5, 53.2, 51.6, 49.4, 45.6, 37.6, 36.2, 33.4, 26.0, 21.3, 18.1; HRESIMS: [M + Na]+ m/z 413.1571 (calcd for C21 H26 O7 Na+ , 413.1571).
5.3.16
Salvinorin B formate (46).
A mixture of Ac2 O (0.25 mL) and HCO2 H (0.7 mL) was stirred at 45 C for 40 minutes. Meanwhile 1b (18.0 mg, 46.1 mol) was added to dry pyridine (1 mL), warmed to 45 C until fully dissolved, then cooled to 0 C. The cooled anhydride mixture was added dropwise by pipette, with violent bubbling. The solution was warmed to room temperature and stirred for 30 minutes, when TLC (20% acetone/CH2 Cl2 , visualised in KMnO4 ) indicated completion. The reaction mixture was cooled to 0 C, diluted dropwise with water, and extracted into EtOAc. The organic layer was washed (1% HCl, water, 5% NaHCO3 and brine). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (2-4% MeOH/CH2 Cl2 gradient) gave 46 as a clear resin (13.8 mg, 72%);
5.3. SYNTHESIS
217
InChI=1/C22H26O8/c1-21-6-4-13-20(26)30-16(12-5-7-28-10-12)9-22(13,2)18 (21)17(24)15(29-11-23)8-14(21)19(25)27-3/h5,7,10-11,13-16,18H,4,6,8-9H2,1-3 H3/t13-,14-,15-,16-,18-,21-,22-/m0/s1

O
O O
O O
46
1b 65
46 85
[]26 54 (c 0.6, CH2 Cl2 ); D FTIR (lm): max 3147, 2952, 2854, 1726, 1504, 1451, 1439, 1396, 1372, 1278, 1224, 1163, 1108, 1072, 1024, 1008, 941, 899, 875, 786, 735, 701 cm1 ;
1
H NMR (CDCl3 ): 8.14 (1H, s, CH O), 7.41 (1H, br s, H-16), 7.40 (1H, t, J =
1.8 Hz, H-15), 6.38 (1H, dd, J = 1.8, 0.9 Hz, H-14), 5.54 (1H, dd, J = 11.7, 5.2 Hz, H-12), 5.25 (1H, ddt, J = 11.2, 8.7, 1.1 Hz, H-2), 3.73 (3H, s, CO2 CH 3 ), 2.79-2.74 (1H, m, H-4), 2.51 (1H, dd, J = 13.5, 5.2 Hz, H-11), 2.37-2.32 (2H, m, H-3), 2.18 (1H, d, J = 0.8 Hz, H-10), 2.19-2.15 (1H, m, H-7), 2.08 (1H, dd, J = 11.5, 3.0 Hz, H-8), 1.80 (1H, dt, J 12.9, 2.9 Hz, H-6), 1.65 (1H, tdd, J = 13.3, 11.8, 3.2 Hz, H-7), 1.63-1.58 (1H, m, H-6), 1.60 (1H, ddd, J = 13.5, 11.7, 0.8 Hz, H-11), 1.46 (3H, s, H-20), 1.13 (3H, s, H-19);
13
C NMR (CDCl3 ): 200.8 (C, C-1), 171.4 (C, C-18), 171.0 (C, C-17), 159.4
(CH, C HO), 143.7 (CH, C-15), 139.4 (CH, C-16), 125.1 (C, C-13), 108.3 (CH, C-14), 74.5 (CH, C-2), 72.0 (CH, C-12), 64.1 (CH, C-10), 53.5 (CH, C-4), 52.0 (CH3 , CO2 C H3 ), 51.3 (CH, C-8), 43.4 (CH2 , C-11), 42.1 (C, C-5), 38.1 (CH2 , C-6), 35.5 (C, C-9), 30.6 (CH2 , C-3), 18.1 (CH2 , C-7), 16.4 (CH3 , C-19), 15.2 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 441.1525 (calcd for C22 H26 O8 Na+ , 441.1520).
218
5.3.17
Dideacetylsalvinorin C 2-O-(4-bromobenzoate) (47).
To a solution of 1h (3.8 mg, 9.7 mol), 4-bromobenzoyl chloride (8.7 mg, 39.6 mol) and DMAP (3.1 mg, 25.4 mol) in dry CH2 Cl2 (1.5 mL) was added NEt3 (100 L). The bright yellow solution was stirred under Ar for 90 min, when TLC (Et2 O, visualised in KMnO4 ) indicated completion. The solution was diluted in Et2 O, washed with 10% HCl ( 3), sat. NaHCO3 and brine. Drying (MgSO4 ) and evaporation in vacuo gave 47 as a white powder (5.0 mg, 90%); InChI=1/C28H29BrO8/c1-27-10-8-18-26(33)37-21(16-9-11-35-14-16)13-28(18, 2)23(27)22(30)20(12-19(27)25(32)34-3)36-24(31)15-4-6-17(29)7-5-15/h4-7,9,11 -12,14,18,20-23,30H,8,10,13H2,1-3H3/t18-,20-,21-,22-,23-,27-,28-/m0/s1
O
O O Br O
OH H
O O
47
1h 40
47 78
mp (from cold Et2 O) 161-166 C; []18 +86 (c 0.4, CH2 Cl2 ); D FTIR (lm): max 3509, 2950, 1717, 1590, 1504, 1484, 1434, 1398, 1269, 1233, 1175, 1143, 1104, 1071, 1012, 946, 875, 849, 786, 758, 679 cm1 ;
1
H NMR (CDCl3 ): 7.93 (2H, d, J = 8.6 Hz, H-2), 7.62 (2H, d, J = 8.6 Hz,
H-3), 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.9 Hz, H-15), 6.54 (1H, t, J = 2.0 Hz, H-3), 6.42 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.63 (1H, dd, J = 4.8, 2.4 Hz, H-2), 5.59 (1H, dd, J = 10.9, 5.8 Hz, H-12), 4.61 (1H, br d, J = 4.6 Hz, H-1), 3.74 (3H, s, CO2 CH 3 ), 2.54 (1H, dt, J = 13.4, 3.6 Hz, H-6), 2.48 (1H,
5.3. SYNTHESIS
219
dd, J = 13.3, 6.1 Hz, H-11), 2.17 (1H, dd, J = 12.0, 3.5 Hz, H-8), 2.12 (1H, dq, J = 12.0, 3.5 Hz, H-7), 1.96 (1H, br s, w 1 = 14.6 Hz, OH ), 1.92-1.80
2
(1H, m, H-7), 1.76 (3H, s, H-19), 1.67 (1H, dd, J = 12.9, 11.2 Hz, H-11), 1.47 (3H, s, H-20), 1.38 (1H, s, H-10), 1.22 (1H, td, J = 13.1, 3.7 Hz, H-6);
13
C NMR (CDCl3 ): 171.8, 166.0, 164.9, 143.9, 143.6, 139.3, 132.0, 131.4,
131.2, 129.0, 128.0, 125.7, 108.4, 73.2, 71.7, 64.4, 53.9, 51.8, 51.7, 44.4, 37.9, 37.6, 37.0, 22.0, 18.4, 16.2; HRESIMS: [M + Na]+ m/z 595.0938, 597.0920 (calcd for C28 H29 O8 BrNa+ , 595.0938, 597.0917).
5.3.18
17-Deoxysalvinorin A (49).
Et3 SiH (20 L, 125 mol) was added to a stirred solution of lactol 35 (18.3 mg, 42.1 mol) in dry CH2 Cl2 (1 mL) under Ar. The solution was cooled to 0 C, and BF3 Et2 O (10 L, 79 mol) was added. The light brown solution was stirred at 0 C for 2 h, when TLC (10% acetone/CH2 Cl2 ) showed no 35. The reaction was quenched (0.5 mL sat. NaHCO3 ), and partitioned between Et2 O and brine. Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (0-4% acetone/CH2 Cl2 gradient) gave enol ether 50 (4.1 mg, 23%) along with 49 as a clear resin (8.4 mg, 48%); InChI=1/C23H30O7/c1-13(24)30-17-9-16(21(26)27-4)22(2)7-5-15-12-29-18(14 -6-8-28-11-14)10-23(15,3)20(22)19(17)25/h6,8,11,15-18,20H,5,7,9-10,12H2,14H3/t15-,16-,17-,18-,20-,22-,23-/m0/s1
O
O O
49
220 TLC: hRf 10% acetone/CH2 Cl2 2% acetone/CH2 Cl2 35 38 49 50 72 24 72 42
[]20 81 (c 0.4, CH2 Cl2 ); D FTIR (lm): max 2950, 2927, 2857, 1730, 1506, 1236, 1201, 1161, 1107, 1042, 1018, 932, 889, 875, 784, 736 cm1 ;
1
H NMR (CDCl3 ): 7.33 (2H, m, H-15 & 16), 6.34 (1H, t, J = 1.4 Hz, H-14),
5.14 (1H, dd, J = 10.7, 9.4 Hz, H-2), 4.70 (1H, dd, J = 11.6, 2.5 Hz, H-12), 3.70 (3H, s, CO2 CH 3 ), 3.58 (2H, d, J = 7.6 Hz, H-17), 2.79-2.75 (1H, m, H-4), 2.29-2.24 (2H, m, H-3), 2.15 (3H, s, OCOCH 3 ), 2.12 (1H, dd, J = 13.1, 2.6 Hz, H-11), 2.09 (1H, d, J = 1.0 Hz, H-10), 1.68-1.62 (2H, m, H-6), 1.54-1.43 (1H, m, H-8), 1.38 (3H, s, H-20), 1.35-1.26 (2H, m, H-7), 1.19 (1H, ddd, J = 13.1, 11.6, 1.0 Hz, H-11), 1.08 (3H, s, H-19);
13
C NMR (CDCl3 ): 202.5 (C, C-1), 171.9 (C, C-18), 169.9 (C, OC OCH3 ),
142.9 (CH, C-15), 138.9 (CH, C-16), 127.0 (C, C-13), 108.7 (CH, C-14), 75.0 (CH, C-2), 67.5 (CH, C-12), 67.1 (CH2 , C-17), 65.6 (CH, C-10), 53.8 (CH, C-4), 51.8 (CH3 , CO2 C H3 ), 46.4 (CH, C-8), 45.4 (CH2 , C-11), 42.7 (C, C-5), 39.0 (CH2 , C-6), 34.6 (C, C-9), 30.8 (CH2 , C-3), 20.6 (CH3 , OCOC H3 ), 19.6 (CH2 , C-7), 16.8 (CH3 , C-19), 13.7 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 441.1881 (calcd for C23 H30 O7 Na+ , 441.1884).
5.3.19
8,17-Didehydro-17-deoxysalvinorin A (50).
Et3 SiH (25 L, 156 mol) and Amberlyst 15 resin (22 mg) were added to a solution of lactol 35 (15.1 mg, 34.7mol) in CH2 Cl2 (1 mL). The sealed ask was stirred at rt for 24 h. Filtration, evaporation and ash column chromatography (1% acetone/CH2 Cl2 ) gave 50 as a clear resin (11.0 mg, 76%); InChI=1/C23H28O7/c1-13(24)30-17-9-16(21(26)27-4)22(2)7-5-15-12-29-18(14-
5.3. SYNTHESIS
221
6-8-28-11-14)10-23(15,3)20(22)19(17)25/h6,8,11-12,16-18,20H,5,7,9-10H2,1-4H 3/t16-,17-,18-,20-,22-,23-/m0/s1
O
O O
50
TLC: See table on the facing page. []23 60 (c 0.5, CH2 Cl2 ); D FTIR (lm): max 2927, 2855, 1731, 1664, 1504, 1437, 1380, 1275, 1237, 1203, 1164, 1113, 1098, 1044, 1024, 1000, 962, 944, 875, 786, 737 cm1 ;
1
6.36 (1H, dd, J = 1.9, 0.8 Hz, H-14), 6.26 (1H, d, J = 1.8 Hz, H-17), 5.12 (1H, dd, J = 11.0, 9.0 Hz, H-2), 4.78 (1H, dd, J = 11.7, 2.1 Hz, H-12), 3.70 (3H, s, CO2 CH 3 ), 2.74-2.69 (1H, m, H-4), 2.33 (1H, dd, J = 13.6, 2.1 Hz, H-11), 2.32-2.24 (1H, m, H-7), 2.29-2.24 (2H, m, H-3), 2.16 (3H, s, OCOCH 3 ), 2.12 (1H, d, J = 0.8 Hz, H-10), 1.94 (1H, ddd, J = 14.6, 4.6, 2.6 Hz, H-7), 1.69 (1H, ddd, J = 13.2, 4.4, 2.6 Hz, H-6), 1.52 (3H, s, H-20), 1.50 (1H, td, J = 13.2, 4.6 Hz, H-6), 1.40 (1H, ddd, J = 13.6, 11.6, 0.8 Hz, H-11), 1.15 (3H, s, H-19);
13
143.2 (CH, C-15), 139.3 (CH, C-16), 137.1 (CH, C-17), 125.8 (C, C-13), 117.0 (C, C-8), 108.7 (CH, C-14), 75.2 (CH, C-2), 66.6 (CH, C-12), 65.5 (CH, C-10), 53.5 (CH, C-4), 51.7 (CH3 , CO2 C H3 ), 44.4 (CH2 , C-11), 42.8 (C, C-5), 40.1 (CH2 , C-6), 34.1 (C, C-9), 30.6 (CH2 , C-3), 22.8 (CH3 , C-20), 22.6 (CH2 , C-7), 20.6 (CH3 , OCOC H3 ), 15.4 (CH3 , C-19); HRESIMS: [M + Na]+ m/z 439.1728 (calcd for C23 H28 O7 Na+ , 439.1727).
222
5.3.20
13,14,15,16-Tetrahydrosalvinorin A (51).
5% Rh/C (25.3 mg) was added to a solution of 1a (20.3 mg, 46.9 mol) in 50% CH2 Cl2 /MeOH (6 mL). The suspension was agitated under H2 (4 atm) at room temperature for 90 minutes, when TLC indicated completion (hRf = 46 (51), 74 (1a) in 20% acetone/CH2 Cl2 ). The solution was ltered through diatomite lter aid and evaporated in vacuo. Flash column chromatography (10% acetone/CHCl3 ) gave 51 (13-epimers, 1:1) as a clear resin (12 mg, 59%). For characterisation, a portion of the less polar epimer was separated by HPLC in EtOAc; InChI=1/C23H32O8/c1-12(24)30-16-9-15(20(26)28-4)22(2)7-5-14-21(27)31-17 (13-6-8-29-11-13)10-23(14,3)19(22)18(16)25/h13-17,19H,5-11H2,1-4H3/t13u,1 4-,15-,16-,17-,19-,22-,23-/m0/s1
O
13
O O
H O H H
H O O
51
TLC: hRf 20% acetone/CH2 Cl2 HPLC: t R (min) EtOAc
51 46 51
1a 74
11.4, 11.75
[]19 39 (c 0.4, CHCl3 ); D FTIR (lm): max 2953, 2879, 1730, 1452, 1438, 1379, 1277, 1235, 1203, 1165, 1110, 1078, 1050, 1005, 951, 912, 895, 755 cm1 ;
1
H NMR (CDCl3 ): 5.14 (1H, dd, J = 11.4, 8.6 Hz, H-2), 4.44 (1H, ddd, J
= 11.7, 6.9, 5.0 Hz, H-12), 3.85 (1H, td, J = 8.5, 5.0 Hz, H-15a), 3.78 (1H, dd, J = 9.0, 7.6 Hz, H-16a), 3.74 (1H, dt, J = 8.5, 7.4 Hz, H-15b), 3.72 (3H,
5.3. SYNTHESIS
223
s, CO2 CH 3 ), 3.54 (1H, dd, J = 9.0, 6.7 Hz, H-16b), 2.73 (1H, dd, J 11.4, 5.4 Hz, H-4), 2.43 (1H, sext, J = 7.4 Hz, H-13), 2.31-2.26 (2H, m, H-3), 2.19 (1H, dd, J = 13.3, 5.0 Hz, H-11), 2.17 (3H, s, OCOCH 3 ), 2.17-2.12 (1H, m, H-7), 2.12 (1H, br s, H-10), 2.08 (1H, dddd, J = 12.6, 8.7, 7.4, 5.0 Hz, H-14a), 1.95 (1H, dd, J = 11.5, 3.1 Hz, H-8), 1.81 (1H, ddt, J = 12.6, 8.3, 7.4 Hz, H-14b), 1.78-1.74 (1H, m, H-6), 1.65-1.56 (1H, m, H-7), 1.60-1.50 (1H, m, H-6), 1.35 (3H, s, H-20), 1.26 (1H, ddd, J = 13.3, 11.7, 0.8 Hz, H-11), 1.09 (3H, s, H-19);
13
C NMR (CDCl3 ): 202.0 (C, C-1), 171.5 (C, C-18), 171.3 (C, C-17), 169.9
(C, OC OCH3 ), 78.2 (CH, C-12), 75.0 (CH, C-2), 68.8 (CH2 , C-16), 68.0 (CH2 , C-15), 64.0 (CH, C-10), 53.5 (CH, C-4), 52.0 (CH3 , CO2 C H3 ), 51.4 (CH, C-8), 45.1 (CH, C-13), 42.0 (C, C-5), 41.2 (CH2 , C-11), 38.1 (CH2 , C-6), 35.1 (C, C-9), 30.8 (CH2 , C-3), 28.2 (CH2 , C-14), 20.6 (CH3 , OCOC H3 ), 18.1 (CH2 , C-7), 16.3 (CH3 , C-19), 15.1 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 459.1984 (calcd for C23 H32 O8 Na+ , 459.1989).
5.3.21
Autoxidation of 1a in KOH/MeOH.
Oxygen was bubbled through a solution of KOH in MeOH (1 M, 2 mL) for 5 min. This was then added to a solution of 1a (21.4 mg, 49.5 mol) in minimal CH2 Cl2 ( 250 L), and oxygen bubbled through the resulting orange solution for 20 minutes, when TLC (Et2 O) showed no 1a or 1b. 10% HCl was added dropwise until an opaque white colour persisted. The solution was diluted in 0.05 M NaOH and extracted into CH2 Cl2 ( 3). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (20 50% Et2 O/petrol) gave enedione 59 as a resin (9.0 mg, 47%); InChI=1/C21H22O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7-8,10,12,15,23H,4,6,9H2,1-3H3/t12-, 15-,20-,21-/m0/s1
224
O
OH O
O O
59
TLC: hRf Et2 O
1a 54
1b 42
59 69
[]14 +58 (c 0.7, CH2 Cl2 ); D UV (CH3 CN): max (log ) 215 (4.30), 249 (3.77), 324 (3.57) nm; FTIR (lm): max 3373, 3149, 2955, 1726, 1651, 1601, 1504, 1456, 1435, 1408, 1380, 1331, 1244, 1203, 1162, 1068, 1022, 911, 875, 793, 736, 703 cm1 ;
1
6.99 (1H, s, H-3), 6.91 (1H, s, OH ), 6.42 (1H, dd, J = 2.0, 0.9 Hz, H-14), 5.44 (1H, dd, J = 12.3, 2.9 Hz, H-12), 3.85 (3H, s, CO2 CH 3 ), 3.11 (1H, ddd, J = 14.8, 2.9, 1.2 Hz, H-11), 2.98 (1H, ddd, J = 9.7, 5.4, 1.2 Hz, H-8), 2.53 (1H, ddd, J = 14.1, 7.7, 6.3 Hz, H-6a), 2.24 (1H, dtd, J = 14.6, 7.4, 5.3 Hz, H-7a), 2.02 (1H, dd, J = 15.0, 12.2 Hz, H-11), 1.98 (1H, dddd, J = 14.3, 9.7, 7.7, 6.4 Hz, H-7b), 1.77-1.67 (1H, m, H-6b), 1.72 (3H, s, H-19), 1.67 (3H, s, H-20);
13
C NMR (CDCl3 ): 180.7 (C, C-2), 173.2 (C, C-17), 165.4 (C, C-18), 157.5
(C, C-4), 145.0 (C, C-1), 143.6 (CH, C-15), 140.0 (C, C-10), 139.6 (CH, C-16), 128.2 (CH, C-3), 124.5 (C, C-13), 108.4 (CH, C-14), 70.9 (CH, C-12), 52.6 (CH3 , CO2 C H3 ), 44.9 (CH, C-8), 42.3 (C, C-5), 37.7 (C, C-9), 36.8 (CH2 , C-11), 30.3 (CH3 , C-19), 28.3 (CH2 , C-6), 24.4 (CH3 , C-20), 21.9 (CH2 , C-7); HRESIMS: [M + Na]+ m/z 409.1265 (calcd for C21 H22 O7 Na+ , 409.1258). The aqueous layer was acidied with 10% HCl until opaque white, then extracted into CH2 Cl2 ( 3), which was dried (MgSO4 ) and ltered. MeOH (10 mL) and Me3 SiCHN2 in Et2 O (2.0 M, 200 L) were added. The yellow solution was stirred for 30 min, then evaporated under reduced pressure. Flash column
5.3. SYNTHESIS
225
chromatography (50% EtOAc/petrol) gave a mixture of the seco triesters 60a, 60b and 60c (12.0 mg, 53%). Repeated HPLC (36% EtOAc/petrol) gave a compound which decomposed before full characterisation but was tentatively identied, based on HRESIMS and NMR (1 H and COSY), as 60c; InChI=1/C23H30O9/c1-22(15(19(25)29-4)10-17(24)28-3)8-6-14-20(26)32-16 (13-7-9-31-12-13)11-23(14,2)18(22)21(27)30-5/h7,9,12,14-16,18H,6,8,10-11H2, 1-5H3/t14-,15-,16-,18+,22-,23-/m0/s1
O
O O O O O
O O
10
60c (tentative)
TLC: hRf
59
60a 41
60b 41
60c 41
80% Et2 O/petrol 50 HPLC: t R (min) 36% EtOAc/petrol

1
60c
60a
60b
17.0 17.4 18.3
6.42 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.19 (1H, dd, J = 11.7, 5.4 Hz, H-12), 3.72 (3H, s, CO2 C H3 ), 3.67 (3H, s, CO2 C H3 ), 3.65 (3H, s, CO2 C H3 ), 3.55 (1H, dd, J = 12.3, 2.9 Hz, H-4), 2.80 (1H, dd, J = 16.6, 12.3 Hz, H-3a), 2.59 (1H, dd, J = 14.9, 5.5 Hz, H-11), 2.56 (1H, dd, J = 16.6, 2.9 Hz, H-3b), 2.53 (1H, br s, H-10), 2.39 (1H, dd, J = 12.6, 3.7 Hz, H-8), 1.94-1.83 (1H, m, H-7), 1.86 (1H, dd, J = 14.9, 11.8 Hz, H-11), 1.78 (1H, dq, J = 14.4, 3.8 Hz, H-7), 1.60-1.41 (2H, m, H-6 [obscured by H2 O]), 1.41 (3H, s, H-19), 1.20 (3H, s, H-20); HRESIMS: [M + Na]+ m/z 473.1781 (calcd for C23 H30 O9 Na+ , 473.1782). Further elution gave 60a;
226
InChI=1/C23H30O9/c1-22(15(19(25)29-4)10-17(24)28-3)8-6-14-20(26)32-16 (13-7-9-31-12-13)11-23(14,2)18(22)21(27)30-5/h7,9,12,14-16,18H,6,8,10-11H2, 1-5H3/t14-,15-,16-,18-,22-,23-/m0/s1

O
O O O O O
O O
60a
TLC, HPLC: see table on the previous page. []24 +6 (c 0.1, CH2 Cl2 ); D FTIR (lm): max 2953, 1732, 1506, 1436, 1393, 1373, 1261, 1226, 1202, 1163, 1137, 1079, 1025, 875, 792 cm1 ;
1
H-15), 6.41 (1H, m, H-14), 5.46 (1H, dd, J = 11.6, 5.4 Hz, H-12), 3.69 (3H, s, CO2 C H3 ), 3.68 (3H, s, CO2 C H3 ), 3.65 (3H, s, CO2 C H3 ), 2.82 (1H, dd, J = 15.9, 11.8 Hz, H-3a), 2.72 (1H, dd, J = 11.8, 2.0 Hz, H-4), 2.43 (1H, dd, J = 15.9, 2.0 Hz, H-3b), 2.25 (1H, s, H-10), 2.17-2.08 (2H, m, H-7, 8), 2.01 (1H, dd, J = 13.7, 5.5 Hz, H-11), 1.84 (1H, ddd, J = 13.7, 12.0, 0.8 Hz, H-11), 1.74-1.63 (2H, m, H-6a, 7), 1.52-1.48 (1H, m, H-6b), 1.38 (3H, s, H-19), 1.30 (3H, s, H-20);
13
C NMR (CDCl3 ): 174.0 (C, C-2/18), 172.5 (C, C-2/18), 171.7 (C, C-1),
171.0 (C, C-17), 143.8 (CH, C-15), 139.6 (CH, C-16), 125.1 (C, C-13), 108.5 (CH, C-14), 71.6 (CH, C-12), 58.6 (CH, C-10), 52.0 (CH3 , CO2 C H3 ), 51.9 (CH, C-4), 51.7 (CH3 , CO2 C H3 ), 51.4 (CH3 , CO2 C H3 ), 50.5 (CH, C-8), 44.6 (CH2 , C-11), 38.7 (C, C-5), 36.9 (C, C-9), 35.1 (CH2 , C-6), 32.4 (CH2 , C-3), 18.8 (CH3 , C-19), 18.2 (CH2 , C-7), 15.6 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 473.1783 (calcd for C23 H30 O9 Na+ , 473.1782).
5.3. SYNTHESIS Further elution gave 60b;
227
InChI=1/C23H30O9/c1-22(15(19(25)29-4)10-17(24)28-3)8-6-14-20(26)32-16 (13-7-9-31-12-13)11-23(14,2)18(22)21(27)30-5/h7,9,12,14-16,18H,6,8,10-11H2, 1-5H3/t14-,15+,16+,18+,22+,23+/m1/s1

O
O O O O O
H
8
O O
60b
TLC, HPLC: see table on page 225. []23 +4 (c 0.4, CH2 Cl2 ); D FTIR (lm): max 2953, 1733, 1504, 1436, 1392, 1361, 1247, 1200, 1166, 1091, 1062, 1025, 998, 968, 909, 875, 849, 794, 735 cm1 ;
1
H-15), 6.42 (1H, dd, J = 1.8, 0.7 Hz, H-14), 5.26 (1H, dd, J = 12.0, 2.1 Hz, H-12), 3.67 (3H, s, CO2 C H3 ), 3.66 (3H, s, CO2 C H3 ), 3.64 (3H, s, CO2 C H3 ), 2.84 (1H, dd, J = 15.9, 11.9 Hz, H-3a), 2.76 (1H, dd, J = 11.9, 1.6 Hz, H-4), 2.53 (1H, dd, J = 4.6, 2.8, H-8), 2.39 (1H, dd, J = 16.0, 1.8 Hz, H-3b), 2.28 (1H, br s, H-10), 2.19-2.14 (1H, m, H-7), 2.02 (1H, dd, J = 14.5, 11.9 Hz, H-11), 1.94-1.86 (2H, m, H-6, 7), 1.80 (1H, dd, J = 14.4, 2.1 Hz, H-11), 1.48 (3H, s, H-20), 1.32-1.27 (1H, m, H-6), 1.29 (3H, s, H-19);
13
C NMR (CDCl3 ): 173.8 (C, C-18), 173.2 (C, C-17), 172.6 (C, C-2), 172.2
(C, C-1), 143.7 (CH, C-15), 139.7 (CH, C-16), 123.6 (C, C-13), 108.4 (CH, C-14), 69.8 (CH, C-12), 57.8 (CH, C-10), 51.9 (CH3 , CO2 C H3 ), 51.7 (CH3 , CO2 C H3 ), 51.7 (CH, C-4), 51.5 (CH3 , CO2 C H3 ), 46.5 (CH2 , C-11), 44.4 (CH, C-8), 38.5 (C, C-5), 35.8 (C, C-9), 32.2 (CH2 , C-3), 31.6 (CH2 , C-6), 25.1 (CH3 , C-20), 18.2 (CH3 , C-19), 18.2 (CH2 , C-7); HRESIMS: [M + Na]+ m/z 473.1780 (calcd for C23 H30 O9 Na+ , 473.1782).
228
5.3.22
NaBH4 reduction of 59.
Enedione 59 (41.3 mg, 107 mol) and NaBH4 (10 mg, 264 mol) were dissolved in CH2 Cl2 (500 L), followed by EtOH (2 mL), and stirred under Ar at 40
C. The initial orange colour faded to faint yellow within 1 h. After 4 h,
TLC (Et2 O) indicated completion. The solution was cooled to 0 C, and 0.5% H2 SO4 /MeOH added dropwise until eervescence ceased. The solution was concentrated to 500 L in vacuo, then partitioned between brine (acidied with 10% HCl) and CH2 Cl2 ( 3). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (70-100% Et2 O/petrol) gave 38h27 (15.7 mg, 37%); InChI=1/C21H28O7/c1-20-6-4-12-19(25)28-15(11-5-7-27-10-11)9-21(12,2)17 (20)16(23)14(22)8-13(20)18(24)26-3/h5,7,10,12-17,22-23H,4,6,8-9H2,1-3H3/t 12-,13+,14+,15+,16+,17+,20+,21+/m1/s1
O
HO
OH H
H
8
O O
38h
TLC: hRf Et2 O

1
38h 53
59 69
H NMR (CDCl3 ): 7.48 (1H, dt, J = 1.7, 0.9 Hz, H-16), 7.42 (1H, t, J =
1.7 Hz, H-15), 6.41 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.29 (1H, dd, J = 11.7, 1.5 Hz, H-12), 4.07 (1H, br s, H-1), 3.65 (3H, s, CO2 CH 3 ), 3.54 (1H, ddd, J = 11.1, 4.7, 3.2 Hz, H-2), 2.45 (1H, br d, J = 4.7 Hz, H-8), 2.23-2.10 (4H, m), 1.90 (1H, tdd, J = 14.4, 5.5, 4.0 Hz, H-7), 1.73-1.53 (m, obscured by H2 O & OH ), 1.66 (3H, s, H-20), 1.32 (3H, s, H-19), 0.90 (1H, d, J = 1.6 Hz, H-10);
1
H NMR ([CD3 ]2 CO) and
13
C NMR (CDCl3 ) matched literature values.27
5.3. SYNTHESIS
229
5.3.23
O-Demethyl-18-deoxysalvinorin A (77).
To dry EtSH (1.5 mL, 20.2 mmol) at 0 C, stirred rapidly under a stream of Ar, was added n BuLi in hexanes (2.1 M, 8 mL, 16.9 mmol). Immediate, violent gas evolution was accompanied by the sudden formation of a white solid. Rapid stirring was continued while the remaining n BuLi was added, swirling the ask when necessary to free the stir bar. After warming to room temperature, the solution was evaporated under reduced pressure and dried under high vacuum at 50 C for 30 min, giving LiSEt as a white powder. This was stored at room temperature in a sealed ask under Ar for up to a year without losing activity. 1a (42.2 mg, 97.6 mol) and LiSEt (13.7 mg, 201 mol) under Ar were dissolved in DMPU (1 mL). The yellow solution was stirred at 55 C for 23 h, when TLC (1% H2 SO4 /10% MeOH/CH2 Cl2 ) indicated consumption of 1a and intermediate 1b. The cooled orange solution was diluted with EtOAc and washed (10% HCl 3, then water), then extracted into 1% NaHCO3 ( 3). The pooled aqueous fractions were acidied at 0 C with 10% HCl, then extracted into CH2 Cl2 ( 3). Drying (MgSO4 ) and evaporation in vacuo gave an amber resin, which was treated with Ac2 O (0.4 mL) in pyridine (1 mL) and catalytic DMAP at room temperature for 17 h. After cooling to 0 C and quenching (water), the solution was diluted in EtOAc and washed (10% HCl and sat. NH4 Cl). Drying (MgSO4 ) and evaporation in vacuo gave the mixed acids 67a (H-8 : , 1.4:1) as an amber resin (29.7 mg, 73% over two steps); InChI=1/C22H26O8/c1-11(23)29-15-8-14(19(25)26)21(2)6-4-13-20(27)30-16 (12-5-7-28-10-12)9-22(13,3)18(21)17(15)24/h5,7,10,13-16,18H,4,6,8-9H2,13H3,(H,25,26)/t13u,14-,15-,16-,18-,21-,22-/m0/s1/f/h25H
O
O O
O O
OH
67a
230 TLC: hRf 1% H2 SO4 /10% MeOH/CH2 Cl2 1% NEt3 /EtOAc Et2 O vanillin/H2 SO4
1
CHAPTER 5. EXPERIMENTAL. 1a 71 1b 63 67a 0-20 0-30 56 20 purple purple purple 56 30 blue 77 78
H NMR (CDCl3 ): 7.43 (1H, br s, H-16 ), 7.41 (1H, br s, H-16), 7.39-7.38
(2H, m, H-15, 15 ), 6.37 (2H, m, H-14, 14 ), 5.52 (1H, dd, J = 11.6, 5.2 Hz, H-12), 5.26 (1H, dd, J = 11.9, 1.8 Hz, H-12 ), 5.16 (1H, dd, J = 12.4, 7.6 Hz, H-2), 5.11 (1H, dd, J = 12.4, 7.2 Hz, H-2 ), 2.80 (1H, dd, J = 5.2, 3.5 Hz, H-4), 2.76 (1H, dd, J = 5.2, 3.5 Hz, H-4 ), 2.49 (1H, dd, J = 13.4, 5.2 Hz, H-11), 2.45 (1H, dd, J = 4.9, 2.0 Hz, H-8 ), 2.37 (1H, dd, J = 14.8, 2.0 Hz, H-11), 2.37-1.94 (m), 2.28 (1H, br s, H-10 ), 2.20 (1H, br s, H-10), 2.17 (3H, s, OCOCH 3 ), 2.15 (3H, s, OCOCH 3 ), 1.84 (1H, tt, J = 14.2, 4.2 Hz, H-7), 1.73 (1H, dtd, J = 13.7, 3.5, 0.8 Hz, H-6), 1.67-1.52 (m), 1.63 (3H, s, H-20 ), 1.50 (1H, dd, J = 15.1, 12.1 Hz, H-11 ), 1.45 (3H, s, H-20), 1.13 (3H, s, H-19), 1.09 (3H, s, H-19 ). The mixed acids 67a (35.2 mg, 84.0 mol) were stirred in dry THF (1 mL) at rt, under Ar, in a ask tted with a reux condenser. BH3 THF (1.0 M, 110 L, 110 mol) was added dropwise; the solution was heated to 55 C and stirred at this temperature for 90 minutes, when TLC (1% NEt3 /EtOAc) indicated completion. The solution was cooled to room temperature, quenched with water (dropwise), and evaporated under reduced pressure to a cloudy paste. This was diluted with sat. NaHCO3 and extracted into CH2 Cl2 (3). Drying (MgSO4 ), evaporation in vacuo and ash column chromatography (10-25% acetone/Et2 O gradient) monitored by TLC (Et2 O) gave 78 as a clear resin (8.4 mg, 25%); InChI=1/C22H28O7/c1-12(24)28-16-8-14(10-23)21(2)6-4-15-20(26)29-17(135-7-27-11-13)9-22(15,3)19(21)18(16)25/h5,7,11,14-17,19,23H,4,6,8-10H2,13H3/t14-,15+,16-,17-,19-,21-,22-/m0/s1
Unnatural (H-8) epimer.
5.3. SYNTHESIS
O
231
O O
H
8
O O
OH
78
TLC: see table on the preceding page. []24 31 (c 0.3, CH2 Cl2 ); D FTIR (lm): max 3547, 3148, 2942, 2880, 1725, 1504, 1468, 1450, 1440, 1376, 1238, 1202, 1159, 1085, 1050, 1024, 950, 934, 875, 802, 784, 734, 703 cm1 ;
1
6.38 (1H, dd, J = 1.9, 0.9 Hz, H-14), 5.25 (1H, dd, J = 12.0, 2.2 Hz, H-12), 5.09 (1H, ddd, J = 12.3, 7.2, 1.0 Hz, H-2), 3.93 (1H, dd, J = 10.5, 3.9 Hz, H-18a), 3.46 (1H, ddd, J = 10.8, 8.0, 0.6 Hz, H-18b), 2.52 (1H, ddd, J = 12.3, 7.0, 2.8 Hz, H-3a), 2.45-2.43 (1H, m, H-8), 2.36 (1H, dd, J = 15.0, 2.2 Hz, H-11), 2.24 (1H, d, J = 0.9 Hz, H-10), 2.22-2.09 (2H, m, H-6,7), 2.15 (3H, s, OCOCH 3 ), 1.91-1.72 (4H, m, H-3b, 4, 6, 7), 1.64 (3H, s, H-20), 1.52 (1H, dd, J = 14.8, 12.0 Hz, H-11), 0.92 (3H, s, H-19);
13
C NMR (CDCl3 ): 204.0, 173.6, 169.9, 143.6, 139.7, 123.4, 108.6, 76.2, 70.1,
64.5, 61.7, 50.3, 48.1, 45.4, 42.1, 34.6, 33.9, 31.6, 24.8, 20.6, 17.7, 15.5; HRESIMS: [M + Na]+ m/z 427.1725 (calcd for C22 H28 O7 Na+ , 427.1727). Further elution gave 77 as a clear resin (7.7 mg, 23%); InChI=1/C22H28O7/c1-12(24)28-16-8-14(10-23)21(2)6-4-15-20(26)29-17(135-7-27-11-13)9-22(15,3)19(21)18(16)25/h5,7,11,14-17,19,23H,4,6,8-10H2,13H3/t14-,15-,16-,17-,19-,21-,22-/m0/s1
232
O
O O
O O
OH
77
TLC: see table on page 230. []25 19 (c 0.3, CH2 Cl2 ); D FTIR (lm): max 3468, 2944, 1727, 1504, 1452, 1376, 1237, 1162, 1095, 1053, 1023, 952, 875, 794, 735, 703 cm1 ;
1
(1H, dd, J = 1.9, 0.8 Hz, H-14), 5.52 (1H, dd, J = 11.7, 5.3 Hz, H-12), 5.15 (1H, dd, J = 12.3, 7.6 Hz, H-2), 3.94 (1H, dd, J = 10.3, 3.9 Hz, H-18a), 3.49 (1H, dd, J = 10.3, 8.0 Hz, H-18b), 2.54 (1H, ddd, J = 12.3, 7.0, 2.6 Hz, H-3a), 2.49 (1H, dd, J = 13.3, 5.3 Hz, H-11), 2.17 (1H, d, J = 0.9 Hz, H-10), 2.16 (3H, s, OCOCH 3 ), 2.20-2.13 (1H, m, H-7), 2.06 (1H, dd, J = 12.1, 3.0 Hz, H-8), 1.99 (1H, dt, J = 13.3, 3.3 Hz, H-6), 1.94-1.87 (1H, m, H-4), 1.80 (1H, q, J = 12.3 Hz, H-3b), 1.64 (1H, dddd, J = 14.0, 13.5, 12.1, 3.4 Hz, H-7), 1.58 (1H, ddd, J = 13.3, 11.7, 0.9 Hz, H-11), 1.45 (3H, s, H-20), 1.43 (1H, td, J = 13.2, 3.7 Hz, H-6), 0.96 (3H, s, H-19);
13
143.7 (CH, C-15), 139.4 (CH, C-16), 125.2 (C, C-13), 108.4 (CH, C-14), 76.0 (CH, C-2), 72.1 (CH, C-12), 64.5 (CH, C-10), 61.7 (CH2 , C-18), 51.5 (CH, C-8), 50.8 (CH, C-4), 43.4 (CH2 , C-11), 41.9 (C, C-5), 38.1 (CH2 , C-6), 35.3 (C, C-9), 31.8 (CH2 , C-3), 20.6 (CH3 , OCOC H3 ), 18.1 (CH2 , C-7), 16.7 (CH3 , C-19), 15.3 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 427.1729 (calcd for C22 H28 O7 Na+ , 427.1727).
5.3. SYNTHESIS
233
5.3.24
1-Deoxysalvinorin A (81a).
A solution of NaBH4 (11.6 mg, 307 mol) and 1a (108 mg, 249 mol) in dry EtOH (9 mL)/CH2 Cl2 (2 mL) was stirred under Ar at 40 C. The cloudy solution gradually cleared. At 4 h, TLC (Et2 O) indicated completion (for hRf values, see table on page 208). The solution was evaporated in vacuo, and the residue partitioned between brine and CH2 Cl2 ( 3). Drying (MgSO4 ) and evaporation in vacuo gave an approximately equal mixture of 36h23 and 38h27 (total 103 mg), which was used without purication. This crude mixture and 1,1-thiocarbonyldiimidazole (118 mg, 662 mol) were dissolved in dry DMF and stirred under Ar at 90 C for 6 h, when TLC (40% acetone/ CH2 Cl2 ) indicated completion. The cooled solution was diluted in EtOAc/Et2 O and washed (10% HCl 3, then brine). Drying (MgSO4 ) and evaporation in vacuo followed by ash column chromatography (1-5% acetone/CH2 Cl2 gradient) gave the cyclic thionocarbonates 80 (H-8 : , 1:2), as a clear resin (72 mg, 67% over two steps); InChI=1/C22H26O7S/c1-21-6-4-12-19(24)27-15(11-5-7-26-10-11)9-22(12,2)17 (21)16-14(28-20(30)29-16)8-13(21)18(23)25-3/h5,7,10,12-17H,4,6,8-9H2,1-3H3 /t12u,13-,14-,15-,16-,17-,21-,22-/m0/s1
O S O O H H O O
80
TLC: hRf Et2 O 10% acetone/CH2 Cl2 40% acetone/CH2 Cl2 vanillin/H2 SO4
36h/38h
80
83 47
81b 40 20
82b 47 20
81a 74
55 62 86
70
purple
blue
purple
234
1
H NMR (CDCl3 ): 7.49 (1H, br s, H-16 ), 7.45 (1H, br s, H-16), 7.43-7.42
(2H, m, H-15, 15 ), 6.40 (1H, br s, H-14, 14 ), 5.58 (1H, dd, J = 11.2, 5.2 Hz, H-12), 5.34 (1H, br d, J = 10.8 Hz, H-12 ), 5.03 (1H, dd, J = 6.4, 2.8 Hz, H-1), 4.93 - 4.78 (3H, m, H-2, 1 , 2 ), 3.70 (3H, s, CO2 CH 3 ), 3.67 (3H, s, CO2 CH 3 ), 2.56 (1H, dd, J = 5.3, 1.9 Hz, H-8 ), 2.45 (1H, dd, J = 12.8, 5.2 Hz, H-11), 2.27 - 2.10 (m), 1.94 (1H, tdd, J = 14.2, 4.8, 3.9 Hz, H-7 ), 1.82 - 1.69 (m), 1.71 (1H, dd, J = 13.5, 11.0 Hz, H-11), 1.64 (3H, s, H-20 ), 1.59-1.49 (m), 1.47 (3H, s, H-20), 1.37 - 1.20 (m), 1.33 (3H, s, H-19), 1.30 (3H, s, H-19);
13
C NMR (CDCl3 ): 191.0, 190.9, 173.4, 171.6, 171.4, 170.8, 144.0, 143.8,
139.7, 139.4, 125.1, 123.3, 108.22, 108.19, 79.8, 79.0, 78.9, 78.8, 71.5, 69.6, 53.9, 53.1, 52.1, 52.0, 51.85, 51.79, 51.2, 48.0, 45.9, 43.5, 38.7, 37.5, 36.4, 36.2, 35.9, 34.9, 26.4, 26.3, 26.2, 18.3, 17.6, 16.5, 16.2, 14.8.
n
Bu3 SnH was prepared by a published procedure408 with reduced reaction time
(10 minutes),409 distilled (80 C/0.5 mmHg) and stored in the dark under argon at -20 C (when used, it remained clear and unclouded). The thionocarbonate mixture was dissolved in dry toluene (2 mL). Ar was bubbled through the resulting cloudy solution for 2 min; it was then stirred under Ar at 80 C. A solution of n Bu3 SnH408 (150 L, 550 mol) and AIBN (5.4 mg, 33 mol) in dry toluene (2 mL, also deoxygenated) was added in small portions over 4 h; the solution gradually cleared. After a further 2 h, TLC (10% acetone/ CH2 Cl2 ) indicated completion. The solution was cooled and loaded directly onto silica gel, rinsing the ask with CH2 Cl2 ( 2). Repeated ash column chromatography (50-70% EtOAc/petrol gradient) monitored by TLC (Et2 O) gave two fractions, A and B. Fraction A was subjected to ash column chromatography (10% acetone/ CH2 Cl2 ) to give carbonate 83 as a clear resin (2.7 mg, 4%); InChI=1/C22H26O8/c1-21-6-4-12-19(24)28-15(11-5-7-27-10-11)9-22(12,2)17 (21)16-14(29-20(25)30-16)8-13(21)18(23)26-3/h5,7,10,12-17H,4,6,8-9H2,1-3H3 /t12-,13-,14-,15-,16-,17-,21-,22-/m0/s1
Unnatural (H-8) epimer.
5.3. SYNTHESIS
O O O O H H O O
235
83
TLC: see table on page 233. []22 +5 (c 0.1, CH2 Cl2 ) ; D FTIR (lm): max 2952, 1797, 1727, 1502, 1452, 1441, 1370, 1276, 1229, 1200, 1163, 1073, 1043, 1023, 956, 875, 790, 736 cm1 ;
1
(1H, dd, J = 1.9, 0.9 Hz, H-14), 5.58 (1H, dd, J = 11.4, 5.4, H-12), 4.90 (1H, dd, J = 6.3, 2.9 Hz, H-1), 4.67 (1H, tdd, J = 7.9, 6.2, 2.2 Hz, H-2), 3.70 (3H, s, CO2 CH 3 ), 2.46-2.41 (2H, m), 2.27-2.10 (4H, m), 1.87-1.67 (m, overlapping with H2 O), 1.42 (3H, s, H-20), 1.36-1.24 (m), 1.34 (3H, s, H-19), 1.27 (1H, d, J = 2.9 Hz, H-10);
13
C NMR (CDCl3 ): 171.6, 170.8, 154.1, 144.0, 139.4, 125.2, 108.2, 75.0, 74.6,
71.5, 54.5, 52.2, 52.0, 43.7, 38.8, 37.5, 35.9, 26.9, 24.3, 18.3, 16.4, 16.3; HRESIMS: [M + Na]+ m/z 441.1515 (calcd for C22 H26 O8 Na+ , 441.1520). Further elution gave 82b as a clear resin (15.5 mg, 25%); InChI=1/C21H28O6/c1-20-6-4-14-19(24)27-16(12-5-7-26-11-12)10-21(14,2)17 (20)9-13(22)8-15(20)18(23)25-3/h5,7,11,13-17,22H,4,6,8-10H2,1-3H3/t13-,14+, 15-,16-,17-,20-,21-/m0/s1
O
HO
H
8
O O
82b
236 TLC: see table on page 233. []21 4 (c 0.1, CH2 Cl2 ); D
FTIR (lm): max 3432, 2948, 2871, 1728, 1502, 1450, 1438, 1390, 1363, 1277, 1255, 1199, 1167, 1091, 1054, 1022, 1000, 875, 840, 785, 727 cm1 ;
1
(1H, dd, J = 2.0, 0.9 Hz, H-14), 5.24 (1H, dd, J = 11.8, 1.5 Hz, H-12), 3.69 3.60 (1H, m, H-2), 3.64 (3H, s, CO2 CH 3 ), 2.48 (1H, dd, J = 5.0, 1.6 Hz, H-8), 2.17 (1H, dtd, J = 14.4, 3.5, 2.1 Hz, H-7), 2.12 (1H, dd, J = 13.2, 3.6 Hz, H-4), 1.99 (1H, dd, J = 14.0, 1.6 Hz, H-11), 1.91 (1H, dddd, J = 12.9, 5.0, 3.3, 1.9 Hz, H-1), 1.87 - 1.58 (m), 1.81 (1H, dd, J = 13.0, 11.0 Hz, H-11), 1.52 (1H, dt, J = 13.6, 3.2 Hz, H-6), 1.43 (1H, td, J = 12.6, 10.7 Hz, H-1), 1.28-1.24 (1H, m, H-6), 1.24 (3H, s, H-19), 1.08 (3H, s, H-20), 1.00 (1H, dd, J = 12.7, 2.2 Hz, H-10);
13
C NMR (CDCl3 ): 174.1, 173.3, 143.7, 139.7, 123.8, 108.4, 70.1, 69.7, 53.8,
52.4, 51.3, 49.1, 44.8, 36.4, 35.8, 34.1, 34.0, 31.3, 24.1, 17.9, 14.1; HRESIMS: [M + Na]+ m/z 399.1771 (calcd for C21 H28 O6 Na+ , 399.1778). Fraction B gave 81b as a resin (13.5 mg, 22%); InChI=1/C21H28O6/c1-20-6-4-14-19(24)27-16(12-5-7-26-11-12)10-21(14,2)17 (20)9-13(22)8-15(20)18(23)25-3/h5,7,11,13-17,22H,4,6,8-10H2,1-3H3/t13-,14-, 15-,16-,17-,20-,21-/m0/s1
O
HO
O O
81b
TLC: see table on page 233.
5.3. SYNTHESIS
1
237
6.40 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.50 (1H, ddd, J = 11.4, 5.7, 0.8 Hz, H-12), 3.70-3.63 (1H, m, H-2), 3.67 (3H, s, CO2 CH 3 ), 2.32 (1H, dd, J = 13.4, 5.8 Hz, H-11), 2.13 (2H, br dd, J = 12.9, 3.2 Hz, H-4, 8), 2.09 (1H, dq, J = 14.1, 3.5 Hz, H-7), 1.96-1.82 (3H, m, H-1, 3), 1.74 (1H, dt, J = 13.6, 3.2 Hz, H-6), 1.65-1.58 (m, overlapping with H2 O), 1.44 (1H, td, J = 12.6, 10.9 Hz, H-1), 1.28 (1H, td, J = 13.6, 4.0 Hz, H-6), 1.10 (3H, s, H-19), 1.07 (3H, s, H-20), 1.04 (1H, dd, J = 12.8, 2.5 Hz, H-10);
13
C NMR (CDCl3 ): 173.0, 172.0, 143.8, 139.3, 125.7, 108.4, 71.8, 70.0, 54.4,
53.0, 51.5, 51.2, 44.0, 38.0, 36.9, 36.2, 33.7, 30.4, 18.2, 15.0, 14.6. This was fully characterised as the acetate by dissolving in pyridine (0.3 mL) and Ac2 O (0.3 mL) with a crystal of DMAP. The solution was stirred at rt for 2.5 h, when TLC (Et2 O, visualised in KMnO4 ) indicated completion, then evaporated in vacuo. Flash column chromatography (60% Et2 O/n-pentane) gave 81a as a clear resin (12.3 mg, 82%); InChI=1/C23H30O7/c1-13(24)29-15-9-17(20(25)27-4)22(2)7-5-16-21(26)30-18 (14-6-8-28-12-14)11-23(16,3)19(22)10-15/h6,8,12,15-19H,5,7,9-11H2,1-4H3/t15 -,16-,17-,18-,19-,22-,23-/m0/s1
O
O O
2
H
1 10
O O
81a
TLC: see table on page 233. []21 8 (c 0.1, CH2 Cl2 ); D FTIR (lm): max 2952, 1729, 1503, 1450, 1434, 1365, 1319, 1245, 1202, 1153, 1099, 1024, 954, 901, 875, 783, 734 cm1 ;
238
1
6.41 (1H, dd, J = 1.9, 1.0 Hz, H-14), 5.47 (1H, ddd, J = 11.2, 5.6, 0.7 Hz, H-12), 4.74 (1H, tt, J = 11.0, 5.5 Hz, H-2), 3.66 (3H, s, CO2 CH 3 ), 2.28 (1H, dd, J = 13.5, 5.9 Hz, H-11), 2.18 (1H, dd, J = 10.9, 3.5 Hz, H-4), 2.15 (1H, dd, J = 10.0, 3.5 Hz, H-8), 2.09 (1H, dq, J = 14.2, 3.4 Hz, H-7), 2.05 (3H, s, OCOCH 3 ), 1.95-1.89 (1H, m, H-1), 1.92-1.83 (2H, m, H-3), 1.74 (1H, dt, J = 13.5, 3.3 Hz, H-6), 1.68-1.57 (1H, m, H-7), 1.64 (1H, dd, J = 13.6, 11.5 Hz, H-11), 1.50 (1H, td, J = 12.8, 11.2 Hz, H-1), 1.31 (1H, td, J = 13.6, 3.7 Hz, H-6), 1.11 (3H, s, H-19), 1.10 (1H, dd, J = 12.9, 2.2 Hz, H-10), 1.06 (3H, s, H-20);
13
143.8 (CH, C-15), 139.4 (CH, C-16), 125.6 (C, C-13), 108.4 (CH, C-14), 71.82 (CH, C-2/12), 71.77 (CH, C-2/12), 54.2 (CH, C-4), 52.9 (CH, C-10), 51.5 (CH3 , CO2 C H3 ), 51.2 (CH, C-8), 43.9 (CH2 , C-11), 38.0 (CH2 , C-6), 36.9 (C, C-9), 36.2 (C, C-5), 29.6 (CH2 , C-3), 26.7 (CH2 , C-1), 21.3 (CH3 , OCOC H3 ), 18.2 (CH2 , C-7), 15.0 (CH3 , C-19), 14.6 (CH3 , C-20); HRESIMS: [M + Na]+ m/z 441.1893 (calcd for C23 H30 O7 Na+ , 441.1884).
Bibliography
[1] Photo by Max Hem, Melbourne. [2] Esquivel, B.; Snchez, A. A.; Aranda, E.: Natural products of agrochemical interest from Mexican Labiatae. In Phytochemicals and Phytopharmaceuticals, F. Shahidi; C.-T. Ho, eds., AOCS Press, Champaign, IL, pp. 371385 2000. [3] Rodriguez-Hahn, L.; Esquivel, B.; Cardenas, J.: Neo-clerodane diterpenoids from American Salvia species. Recent Adv. Phytochem. 1995, 29, 311332. [4] Epling, C.; Jtiva-M., C. D.: A New Species of Salvia from Mexico. Bot. Mus. Lea. Harvard Univ. 1962, 20, 7576. Unocial transcription available online. URL http://www.sagewisdom.org/epling&jativa.html [5] Wasson, R. G.: A New Mexican Psychotropic Drug from the Mint Family. Bot. Mus. Lea. Harvard Univ. 1962, 20, 7784. Unocial transcription available online. URL http://www.sagewisdom.org/wasson1.html [6] Reiseld, A. S.: The Botany of Salvia divinorum (Labiatae). Sida 1993, 15, 349366. HTML adaptation available online. URL http://www.sagewisdom.org/reiseld.html [7] Ott, J.: Ethnopharmacognosy and Human Pharmacology of Salvia divinorum and Salvinorin A. Curare 1995, 18, 103129. See pp. 112 (use by Mexican teenagers), 116 - 118 (pipiltzintzintli), 126 (failure of animal testing). Unocial transcription available online. URL http://www.sagewisdom.org/ott2.html [8] Photo by Slava Olcheski, Virginia. [9] Valds, III, L. J. J.: The early history of Salvia divinorum. Entheogen Rev. 2001, 10, 7375,80. URL http://www.sagewisdom.org/earlysdhistory.html [10] Hofmann, A.: LSD: My Problem Child. McGraw-Hill, New York 1980. Translated by Jonathan Ott. See Chapters 1 and 2 (animal tests of LSD), 239
240
BIBLIOGRAPHY 6 (S. divinorum). Unocial transcription available online. URL http://www.druglibrary.org/schaer/LSD/child.htm
[11] Valds, III, L. J. J.; Daz, J. L.; Paul, A. G.: Ethnopharmacology of ska Maria Pastora (Salvia divinorum, Epling and JtivaM.). J. Ethnopharmacol. 1983, 7, 287312. http://dx.doi.org/10.1016/ 0378-8741(83)90004-1. Unocial transcription available online. URL http://www.sagewisdom.org/valdes83.html [12] Siebert, D. J.: The History of the First Salvia divinorum Plants Cultivated Outside of Mexico. Entheogen Rev. 2003, 12, 117118. URL http://www.sagewisdom.org/salviahistory.html [13] Valds, III, L. J. J.; Hateld, G. M.; Koreeda, M.; Paul, A. G.: Studies of Salvia divinorum (Lamiaceae), an Hallucinogenic Mint from the Sierra Mazateca in Oaxaca, Central Mexico. Econ. Bot. 1987, 41, 283291. Unocial transcription available online. URL http://www.sagewisdom.org/valdes87.html [14] Valds, III, L. J. J.: Salvia divinorum and the Unique Diterpene Hallucinogen, Salvinorin (Divinorin) A. J. Psychoactive Drugs 1994, 26, 277283. Unocial transcription available online. URL http://www.sagewisdom.org/valdes94.html [15] Valds, III, L. J. J.: The Pharmacognosy of Salvia divinorum (Epling And Jtiva-M): an Investigation of Ska Mara Pastora. Ph.D. thesis, University Of Michigan, Ann Arbor, MI 1983. ProQuest Publication Number: AAT 8402393 (document ID: 750217621). See p. 49 - 50 (remedial use). URL http://wwwlib.umi.com/dissertations/fullcit/8402393 [16] Ref 15, pp. 80, 129. [17] Valds, III, L. J. J.: La Divina Pastora (The Divine Shepherdess), a Background to Studies of Salvia divinorum. Salvia Divinorum Magazine 2003, 1(2), 69. URL http://salvia.us/magazine.html [18] Siebert, D. J.: Salvia divinorum and salvinorin A: New pharmacologic ndings. J. Ethnopharmacol. 1994, 43, 5356. http://dx.doi.org/ 10.1016/0378-8741(94)90116-3. Free transcription available. URL http://www.sagewisdom.org/jep.html [19] Wasson, R. G.: Notes on the Present Status of Ololiuhqui and the Other Hallucinogens of Mexico. Bot. Mus. Lea. Harvard Univ. 1963, 20, 161 212. Unocial transcription available online. URL http://pages.prodigy.com/GBonline/liquix.htm
BIBLIOGRAPHY
241
[20] Daz, J. L.: Ethnopharmacology of Sacred Psychoactive Plants Used by the Indians of Mexico. Annu. Rev. Pharmacol. Toxicol. 1977, 17, 647 675. See p. 656. URL http://dx.doi.org/10.1146/annurev.pa.17.040177.003243 [21] Ref 15, pp. 4 - 5. [22] Ortega, A.; Blount, J. F.; Manchand, P. S.: Salvinorin, a New transNeoclerodane Diterpene from Salvia divinorum (Labiatae). J. Chem. Soc., Perkin Trans. 1 1982, pp. 25052508. URL http://dx.doi.org/10.1039/P19820002505 [23] Valds, III, L. J. J.; Butler, W. M.; Hateld, G. M.; Paul, A. G.; Koreeda, M.: Divinorin A, a Psychotropic Terpenoid, and Divinorin B from the Hallucinogenic Mexican Mint Salvia divinorum. J. Org. Chem. 1984, 49, 47164720. URL http://dx.doi.org/10.1021/jo00198a026 [24] Koreeda, M.; Brown, L.; Valds, III, L. J. J.: The absolute stereochemistry of salvinorins. Chem. Lett. 1990, pp. 20152018. [25] Harding, W. W.; Tidgewell, K.; Schmidt, M.; Shah, K.; Dersch, C. M.; Snyder, J.; Parrish, D.; Deschamps, J. R.; Rothman, R. B.; Prisinzano, T. E.: Salvinicins A and B, New Neoclerodane Diterpenes from Salvia divinorum. Org. Lett. 2005, 7, 30173020. URL http://dx.doi.org/10.1021/ol0510522 [26] Rodriguez-Hahn, L.; Esquivel, B.; Cardenas, J.: Clerodane diterpenes in Labiatae. Prog. Chem. Org. Nat. Prod. 1994, 63, 107196. See p. 110 (absolute stereochemistry). [27] Valds, III, L. J. J.; Chang, H. M.; Visger, D. C.; Koreeda, M.: Salvinorin C, a New Neoclerodane Diterpene from a Bioactive Fraction of the Hallucinogenic Mexican Mint Salvia divinorum. Org. Lett. 2001, 3, 39353937. URL http://dx.doi.org/10.1021/ol016820d [28] Valds, III, L. J. J.: Loliolide from Salvia divinorum. J. Nat. Prod. 1986, 49, 171. URL http://dx.doi.org/10.1021/np50043a031 [29] Giroud, C.; Felber, F.; Augsburger, M.; Horisberger, B.; Rivier, L.; Mangin, P.: Salvia divinorum: an hallucinogenic mint which might become a new recreational drug in Switzerland. Forensic Sci. Int. 2000, 112, 143150. URL http://dx.doi.org/10.1016/S0379-0738(00)00180-8 [30] Merritt, A. T.; Ley, S. V.: Clerodane diterpenoids. Nat. Prod. Rep. 1992, 9, 243287. URL http://dx.doi.org/10.1039/NP9920900243
242
BIBLIOGRAPHY
[31] Rogers, D.; Unal, G. G.; Williams, D. J.; Ley, S. V.; Sim, G. A.; Joshi, B. S.; Ravindranath, K. R.: The crystal structure of 3-epicaryoptin and the reversal of the currently accepted absolute conguration of clerodin. J. Chem. Soc., Chem. Commun. 1979, pp. 9799. URL http://dx.doi.org/10.1039/C39790000097 [32] McCrindle, R.; Overton, K. H.: The chemistry of cyclic diterpenoids. In Advances in Organic Chemistry: Methods and Results., R. A. Raphael; E. C. Taylor; H. Wynberg, eds., Wiley Interscience, New York, vol. 5, pp. 47105 1965. [33] Cane, D. E.: Isoprenoid Biosynthesis: Overview. In Comprehensive Natural Products Chemistry, D. E. Cane; D. H. R. Barton; K. Nakanishi; O. Meth-Cohn, eds., Elsevier, New York, vol. 2, pp. 114 1999. URL http://www.amazon.com/gp/reader/0080431542 [34] MacMillan, J.; Beale, M. H.: Diterpene Biosynthesis. In Comprehensive Natural Products Chemistry, D. E. Cane; D. H. R. Barton; K. Nakanishi; O. Meth-Cohn, eds., Elsevier, New York, vol. 2, pp. 221243 1999. URL http://www.amazon.com/gp/reader/0080431542 [35] Nabeta, K.; Ishikawa, T.; Kawae, T.; Okuyama, H.: Biosynthesis of heteroscyphic acid A in cell cultures of Heteroscyphus planus: nonequivalent labelling of C-5 units in diterpene biosynthesis. J. Chem. Soc., Chem. Commun. 1995, pp. 681682. URL http://dx.doi.org/10.1039/C39950000681 [36] Akhila, A.; Rani, K.; Thakur, R. S.: Biosynthesis of the clerodane furano-diterpene lactone skeleton in Tinospora cordifolia. Phytochemistry 1991, 30, 25732576. URL http://dx.doi.org/10.1016/0031-9422(91)85103-7 [37] Rohmer, M.: A Mevalonate-independent Route to Isopentenyl Diphosphate. In Comprehensive Natural Products Chemistry, D. E. Cane; D. H. R. Barton; K. Nakanishi; O. Meth-Cohn, eds., Elsevier, New York, vol. 2, pp. 4568 1999. URL http://www.amazon.com/gp/reader/0080431542 [38] Eisenreich, W.; Bacher, A.; Arigoni, D.; Rohdich, F.: Biosynthesis of isoprenoids via the non-mevalonate pathway. Cell. Mol. Life Sci. 2004, 61, 14011426. URL http://dx.doi.org/10.1007/s00018-004-3381-z [39] SEM image created in collaboration with Dr. Simon Crawford, University of Melbourne. [40] Siebert, D. J.: Localization of Salvinorin A and Related Compounds in Glandular Trichomes of the Psychoactive Sage, Salvia divinorum. Ann. Bot. 2004, 93, 763771. URL http://dx.doi.org/10.1093/aob/mch089
BIBLIOGRAPHY
243
[41] Daz, J. L.: Etnofarmacologa de algunos psicotrpicos vegetales de Mxico. In Etnofarmacologa de Plantas Alucingenas Latinoamericanas, J. L. Daz, ed., Centro Mexicano de Estudios en Farmacodependencia, Mexico, vol. 4 of Cuadernos Cientcos CEMEF, pp. 135201 1975. See pp. 149 - 153 (isolation and testing), pp. 153 - 154 (Leonurus sibiricus as Cannabis substitute), pp. 196 - 197 (English abstract). [42] Ref 15, pp. 20 - 26. Note: HAc is intended to refer to AcOH. [43] Valds, III, L. J. J. Email, 18th July 2003. [44] Habib, A. A.: False-positive alkaloid reactions. J. Pharm. Sci. 1980, 69, 3743. And references therein. [45] Ldy-Tenger, F.: Versuche zur mikrochemischen Identizierung einiger Pyridinderivate. Pharm. Acta Helv. 1953, 28, 2226. [46] McConnell Davis, T. W.; Farmilo, C. G.; Genest, K.: Analysis of an impure heroin seizure. Bull. Narc. 1962, 14, 4757. URL http://www.unodc.org/unodc/bulletin/bulletin_1962-01-01_3_ page006.html [47] Daz, J. L.: Ethnopharmacology and taxonomy of Mexican psychodysleptic plants. J. Psychedelic Drugs 1979, 11, 71101. [48] Scinder Scholar (Chemical Abstracts Service). Searched under the exact phrases salvia along with alkaloid or nitrogenous or nitrogencontaining. Accessed March 2006. URL http://www.cas.org/SCIFINDER/scicover2.html [49] Web of Science (Thomson Scientic). Searched under salvia along with alkaloid or nitrogenous or nitrogen-containing. Accessed March 2006. URL http://scientic.thomson.com/products/wos/ [50] Dictionary of Natural Products. Chapman and Hall/CRC Press, web version 2004(2) ed. 2004. Searched for N in molecular formula eld, and salvia in biological source eld. Accessed March 2006. URL http://www.chemnetbase.com/scripts/dnpweb.exe [51] Raauf, R. F.: Plant Alkaloids: A Guide to Their Discovery and Distribution. Food Products Press, Binghamton, NY 1996. See pp. 111 - 113 (S. apiana, azurea, ballotaeora and leucophylla). [52] Williams, C. H.; Hines, H. J. G.: The toxic properties of Salvia reexa. Aust. Vet. J. 1940, 16, 1420. [53] Raauf, R. F.: Handbook of Alkaloids and Alkaloid-Containing Plants. Wiley Interscience, New York 1970.
244
BIBLIOGRAPHY
[54] Werle, E.; Raub, A.: ber Vorkommen, Bildung Und Abbau Biogener Amine Bei Panzen Unter Besonderer Bercksichtigung Des Histamins. Biochem. Z. 1948, 318, 538553. English abstract available via Scinder (CAN 43:15521). [55] Don, M.-J.; Shen, C.-C.; Lin, Y.-L.; Syu, Wan, J.; Ding, Y.-H.; Sun, C.M.: Nitrogen-Containing Compounds from Salvia miltiorrhiza. J. Nat. Prod. 2005, 68, 10661070. URL http://dx.doi.org/10.1021/np0500934 [56] Ref 15, p. 131. [57] Valds, III, L. J. J.: Comments on the Report of Alkaloids in Salvia divinorum by J. L. Daz in 1975. Salvia divinorum Magazine 2003, 1(2), 11. URL http://salvia.us/magazine.html [58] Ref 15, pp. 104, 127 - 128. [59] Medana, C.; Massolino, C.; Pazzi, M.; Baiocchi, C.: Determination of salvinorins and divinatorins in Salvia divinorum leaves by liquid chromatography/multistage mass spectrometry. Rapid Commun. Mass Spectrom. 2006, 20, 131136. URL http://dx.doi.org/10.1002/rcm.2288 [60] Silverstein, R. M.; Webster, F. X.; Kiemle, D.: Spectrometric Identication of Organic Compounds. Wiley, New York, 7th ed. 2005. See p. 14 (nitrogen rule and corollary). [61] Ref 15, pp. 101, 103. [62] Ott, J.: Pharmacotheon: Entheogenic drugs, their plant sources and history. Natural Products Co., Kennewick, WA 1993. See esp. p. 153 (failure of animal testing). [63] Ref 15, pp. 103 - 105. [64] Coughenour, L. L.; McLean, J. R.; Parker, R. B.: A new device for the rapid measurement of impaired motor function in mice,. Pharmacol. Biochem. Behav. 1977, 6, 351353. URL http://dx.doi.org/10.1016/0091-3057(77)90036-3 [65] Ref 15, pp. 160 - 167. Data from Table 3, p. 164. Fractions plotted are divinorin A, TLC 2 (impure fraction), TLC 3 (much lower in 1a), CHCl3 /MeOH and 90% MeOH. The 1a contents of the mixed fractions were estimated as follows. 1) CHCl3 /MeOH fraction (actually CH2 Cl2 /MeOH): as noted on p. 165, this was the oily solids from Fig. 12, p. 122. Preparative TLC of 1.03 g this (Fig. 25, p. 158) gave TLC 2 (66 mg) and 3 (60 mg). The 1a was contained in these fractions; the level was much higher in TLC 2 than TLC 3 (p. 159). The CH2 Cl2 /MeOH
BIBLIOGRAPHY
245
fraction (oily solids) therefore contained between 60 and 126 mg of 1a (612%), probably closer to 6%. 2) 90% MeOH Fraction: as noted on p. 165, see the partitioning on p. 137. Chromatography on 9.36 g of this fraction gave 1.51 g of yellowish crystals, which yielded 892 mg of 1a after recrystallization. The 90% MeOH fraction therefore contained 9.516% of 1a. [66] Brimblecombe, R. W.; Green, A. L.: Eect of Monoamine Oxidase Inhibitors on the Behaviour of Rats in Halls Open Field. Nature 1962, 194, 983. URL http://dx.doi.org/10.1038/194983a0 [67] Roth, B. L.; Baner, K.; Westkaemper, R.; Siebert, D. J.; Rice, K. C.; Steinberg, S.; Ernsberger, P.; Rothman, R. B.: Salvinorin A: A potent naturally occurring nonnitrogenous -opioid selective agonist. Proc. Natl. Acad. Sci. U. S. A. 2002, 99, 1193411939. URL http://dx.doi.org/10.1073/pnas.182234399 [68] Ref 15, pp. 166 - 167. [69] Ref 15, pp. 144 - 148. [70] Ref 15, p. 140. [71] Pouton, C. W.: Lipid formulations for oral administration of drugs: non-emulsifying, self-emulsifying and self-microemulsifying drug delivery systems. Eur. J. Pharm. Sci. 2000, 11, S9398. URL http://dx.doi.org/10.1016/S0928-0987(00)00167-6 [72] ACD/Labs v8.14 for Solaris. See Scinder (Ref. 48). URL http://www.acdlabs.com [73] Ref 15. See the CH2 Cl2 /MeOH (p. 122) and 90% MeOH fractions (p. 135). [74] Hancock, B. C.; Parks, M.: What is the True Solubility Advantage for Amorphous Pharmaceuticals? Pharm. Res. 2000, 17, 397404. See also references therein. URL http://dx.doi.org/10.1023/A:1007516718048 [75] Griesser, U.; Stowell, J. G.: Solid-state analysis and polymorphism. In Pharmaceutical Analysis, D. C. Lee; M. Webb, eds., Blackwell, Oxford, pp. 240294 2003. URL http://www.amazon.com/gp/reader/0849328144 [76] Eder, M.; Mehnert, W.: Bedeutung Panzlicher Begleitstoe in Extrakten. Pharmazie 1998, 53, 285293. URL http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve& db=PubMed&dopt=Citation&list_uids=9631497
246
BIBLIOGRAPHY
[77] Ghafourian, T.; Zandasrar, P.; Hamishekar, H.; Nokhodchi, A.: The effect of penetration enhancers on drug delivery through skin: a QSAR study. J. Controlled Release 2004, 99, 113125. See also references therein. URL http://dx.doi.org/10.1016/j.jconrel.2004.06.010 [78] Zhang, Y.; Butelman, E.; Schlussman, S.; Ho, A.; Kreek, M.: Eects of the plant-derived hallucinogen salvinorin A on basal dopamine levels in the caudate putamen and in a conditioned place aversion assay in mice: agonist actions at kappa opioid receptors. Psychopharmacology 2005, 179, 551558. URL http://dx.doi.org/10.1007/s00213-004-2087-0 [79] Fantegrossi, W. E.; Kugle, K. M.; Valds, III, L. J. J.; Koreeda, M.; Woods, J. H.: Kappa-opioid receptor-mediated eects of the plantderived hallucinogen, salvinorin A, on inverted screen performance in the mouse. Behav. Pharmacol. 2005, 16, 627633. URL http://www.behaviouralpharm.com/pt/re/bpharm/abstract. 00008877-200512000-00005.htm [80] McCurdy, C. R.; Sufka, K. J.; Smith, G. H.; Warnick, J. E.; Nieto, M. J.: Antinociceptive prole of salvinorin A, a structurally unique kappa opioid receptor agonist. Pharmacol. Biochem. Behav. 2006, 83, 109113. URL http://dx.doi.org/10.1016/j.pbb.2005.12.011 [81] Wang, Y.; Tang, K.; Inan, S.; Siebert, D. J.; Holzgrabe, U.; Lee, D. Y.; Huang, P.; Li, J.-G.; Cowan, A.; Liu-Chen, L.-Y.: Comparison of Pharmacological Activities of Three Distinct -ligands (Salvinorin A, TRK820 and 3FLB) on Opioid Receptors in vitro and their Antipruritic and Antinociceptive Activities in vivo. J. Pharmacol. Exp. Ther. 2005, 312, 220230. URL http://dx.doi.org/10.1124/jpet.104.073668 [82] Harding, W. W.; Tidgewell, K.; Byrd, N.; Cobb, H.; Dersch, C. M.; Butelman, E. R.; Rothman, R. B.; Prisinzano, T. E.: Neoclerodane Diterpenes as a Novel Scaold for Opioid Receptor Ligands. J. Med. Chem. 2005, 48, 47654771. Mesylate = compound 15. URL http://dx.doi.org/10.1021/jm048963m [83] Butelman, E. R.; Harris, T. J.; Kreek, M. J.: The plant-derived hallucinogen, salvinorin A, produces -opioid agonist-like discriminative effects in rhesus monkeys. Psychopharmacology 2004, 172, 220224. URL http://dx.doi.org/10.1007/s00213-003-1638-0 [84] Schmidt, M. D.; Schmidt, M. S.; Butelman, E. R.; Harding, W. W.; Tidgewell, K.; Murry, D. J.; Kreek, M. J.; Prisinzano, T. E.: Pharmacokinetics of the plant-derived -opioid hallucinogen salvinorin A in nonhuman primates. Synapse 2005, 58, 208210. URL http://dx.doi.org/10.1002/syn.20191
BIBLIOGRAPHY
247
[85] Turner, D. M.: Salvinorin - The Psychedelic Essence of Salvia Divinorum. Panther Press, San Francisco, CA 1996. URL http://www.lavondyss.com/donut/scov.html [86] Gartz, J.: Salvia divinorum: Die Wahrsagesalbei. Nachtschatten Verlag, Solothurn, Switzerland 2001. See pp. 39 - 39 (extraction procedure), 62 - 63 (sublingual 1a in DMSO). [87] Schmidt, M. S.; Prisinzano, T. E.; Tidgewell, K.; Harding, W.; Butelman, E. R.; Kreek, M. J.; Murry, D. J.: Determination of Salvinorin A in body uids by high performance liquid chromatography-atmospheric pressure chemical ionization. J. Chromatogr. B. 2005, 818, 221225. URL http://dx.doi.org/10.1016/j.jchromb.2004.12.041 [88] Pichini, S.; Abanades, S.; Farr, M.; Pellegrini, M.; Marchei, E.; Pacici, R.; de la Torre, R.; Zuccaro, P.: Quantication of the plant-derived hallucinogen Salvinorin A in conventional and non-conventional biological uids by gas chromatography/mass spectrometry after Salvia divinorum smoking. Rapid Commun. Mass Spectrom. 2005, 19, 16491656. URL http://dx.doi.org/10.1002/rcm.1970 [89] Aardvark, D., ed.: Salvia Divinorum and Salvinorin A: the Best of the Entheogen Review, 1992-2000. Entheogen Review, Sacramento, 2nd ed. 2001. [90] Siebert, D. J. Email, 16th March 2004. [91] Shulgin, A. Email, 13th Dec. 2005. [92] Hofmann, A.: Mexikanische Zauberdrogen und ihre Wirkstoe. Planta Med. 1964, 12, 341352. URL http://www.erowid.org/references/refs_view.php?A=ShowDoc1& ID=2715 [93] Ref 15, p. 19. [94] Valds, III, L. J. J. Email, 19th Aug. 2002. [95] Bguin, C.; Richards, M. R.; Wang, Y.; Chen, Y.; Liu-Chen, L.-Y.; Ma, Z.; Lee, D. Y. W.; Carlezon, W. A.; Cohen, B. M.: Synthesis and in vitro pharmacological evaluation of salvinorin A analogues modied at C(2). Bioorg. Med. Chem. Lett. 2005, 15, 27612765. URL http://dx.doi.org/10.1016/j.bmcl.2005.03.113 [96] Armbruster, B. N.; Roth, B. L.: Mining the receptorome. J. Biol. Chem. 2005, 280, 51295132. URL http://dx.doi.org/10.1074/jbc.R400030200 [97] Sheer, D. J.; Roth, B. L.: Salvinorin A: the magic mint hallucinogen nds a molecular target in the kappa opioid receptor. Trends Pharmacol.
248
BIBLIOGRAPHY Sci. 2003, 24, 107109. URL http://dx.doi.org/10.1016/S0165-6147(03)00027-0
[98] Sneader, W.: Drug Discovery: A History. Wiley, New York 2005. See esp. pp. 90, 91, 115. URL http://books.google.com/books?ie=UTF-8& vid=ISBN0471899801 [99] Blakemore, P. R.; White, J. D.: Morphine, the Proteus of organic molecules. Chem. Commun. 2002, pp. 11591168. URL http://dx.doi.org/10.1039/b111551k [100] Buckingham, J.: Chasing the molecule. Sutton, Stroud, UK 2004. [101] Rees, D. C.; Hunter, J. C.: Opioid receptors. In Comprehensive medicinal chemistry : the rational design, mechanistic study and therapeutic application of chemical compounds, J. C. Emmett; C. Hansch; P. G. Sammes; J. B. Taylor, eds., Pergamon Press, New York, vol. 3, pp. 805846 1990. [102] Dhawan, B.; Cesselin, F.; Raghubir, R.; Reisine, T.; Bradley, P.; Portoghese, P.; Hamon, M.: International Union of Pharmacology. XII. Classication of opioid receptors. Pharmacol. Rev. 1996, 48, 567592. URL http://pharmrev.aspetjournals.org/cgi/reprint/48/4/567 [103] Poeaknapo, C.; Schmidt, J.; Brandsch, M.; Drager, B.; Zenk, M. H.: Endogenous formation of morphine in human cells. Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 1409114096. URL http://dx.doi.org/10.1073/pnas.0405430101 [104] Szmuszkovicz, J.: U-50,488 and the receptor. A personalized account covering the period 1973 to 1990. Prog. Drug Res. 1999, 52, 167195. URL http://books.google.com/books?ie=UTF-8& vid=ISBN376435979X [105] Szmuszkovicz, J.: U-50,488 and the receptor. Part 2. 1991-1998. Prog. Drug Res. 1999, 53, 151. URL http://books.google.com/books?ie=UTF-8& vid=ISBN3764360283 [106] van Ree, J. M.; Gerrits, M. A. F. M.; Vanderschuren, L. J. M. J.: Opioids, Reward and Addiction: An Encounter of Biology, Psychology, and Medicine. Pharmacol. Rev. 1999, 51, 341396. URL http://pharmrev.aspetjournals.org/cgi/content/abstract/51/2/ 341 [107] Carlezon Jr., W. A.; Bguin, C.; DiNieri, J. A.; Baumann, M. H.; Richards, M. R.; Todtenkopf, M. S.; Rothman, R. B.; Ma, Z.; Lee, D. Y. W.; Cohen, B. M.: Depressive-like Eects of the -Opioid Receptor Agonist Salvinorin A on Behavior and Neurochemistry in Rats. J.
BIBLIOGRAPHY Pharmacol. Exp. Ther. 2005, 316, 440447. And references therein. URL http://dx.doi.org/10.1124/jpet.105.092304
249
[108] Pande, A. C.; Pyke, R. E.; Greiner, M.; Wideman, G. L.; Benjamin, R.; Pierce, M. W.: Analgesic ecacy of enadoline versus placebo or morphine in postsurgical pain. Clin. Neuropharmacol. 1996, 19, 451456. URL http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve& db=PubMed&dopt=Citation&list_uids=8889289 [109] Walsh, S. L.; Strain, E. C.; Abreu, M. E.; Bigelow, G. E.: Enadoline, a selective kappa opioid agonist: comparison with butorphanol and hydromorphone in humans. Psychopharmacology 2001, 157, 151162. URL http://dx.doi.org/10.1007/s002130100788 [110] Chappell, P. B.; Leckman, J. F.; Scahill, L. D.; Hardin, M. T.; Anderson, G.; Cohen, D. J.: Neuroendocrine and behavioral eects of the selective kappa agonist spiradoline in Tourettes syndrome: a pilot study. Psychiatry Res. 1993, 47, 267280. URL http://dx.doi.org/10.1016/0165-1781(93)90084-T [111] Gadano, A.; Moreau, R.; Pessione, F.; Trombino, C.; Giuily, N.; Sinnassamy, P.; Valla, D.; Lebrec, D.: Aquaretic eects of niravoline, a -opioid agonist, in patients with cirrhosis. J. Hepatol. 2000, 32, 3842. URL http://dx.doi.org/10.1016/S0168-8278(00)80187-7 [112] Pfeier, A.; Brantl, V.; Herz, A.; Emrich, H. M.: Psychotomimesis mediated by opiate receptors. Science 1986, 233, 774776. URL http://dx.doi.org/10.1126/science.3016896 [113] Kumor, K. M.; Haertzen, C. A.; Johnson, R. E.; Kocher, T.; Jasinski, D.: Human psychopharmacology of ketocyclazocine as compared with cyclazocine, morphine and placebo. J. Pharmacol. Exp. Ther. 1986, 238, 960968. URL http://jpet.aspetjournals.org/cgi/content/abstract/238/3/960 [114] Greenwald, M. K.; Stitzer, M. L.: Butorphanol agonist eects and acute physical dependence in opioid abusers: comparison with morphine. Drug Alcohol Depend. 1998, 53, 1730. URL http://dx.doi.org/10.1016/S0376-8716(98)00104-5 [115] Barber, A.; Gottschlich, R.: Novel developments with selective, nonpeptidic kappa-opioid receptor agonists. Expert Opin. Invest. Drugs 1997, 6, 13511368. URL http://www.ingentaconnect.com/content/apl/eid/1997/00000006/ 00000010/art00004 [116] Bush, K. A.; Kirkham, B. W.; Walker, J. S.: The -opioid agonist, asimadoline, alters cytokine gene expression in adjuvant arthritis. Rheumatology (Oxford). 2001, 40, 10131021.
250
BIBLIOGRAPHY URL http://rheumatology.oxfordjournals.org/cgi/content/abstract/40/ 9/1013
[117] Riviere, P. J. M.: Peripheral kappa-opioid agonists for visceral pain. Br. J. Pharmacol. 2004, 141, 13311334. URL http://dx.doi.org/10.1038/sj.bjp.0705763 [118] Dortch-Carnes, J.; Potter, D. E.: Bremazocine: a -opioid agonist with potent analgesic and other pharmacologic properties. CNS Drug Rev. 2005, 11, 195212. URL http://www.nevapress.com/cnsdr/contents/1102.html [119] Sorbera, L. A.; Castaner, J.; Leeson, P. A.: Nalfurane hydrochloride. Drugs Fut. 2003, 28, 237242. URL http://dx.doi.org/10.1358/dof.2003.028.03.723584 [120] Wadenberg, M. L.: A review of the properties of spiradoline: a potent and selective -opioid receptor agonist. CNS Drug Rev. 2003, 9, 187 198. URL http://www.nevapress.com/cnsdr/contents/0902.html [121] Ukai, M.; Suzuki, M.; Mamiya, T.: Eects of U-50,488H, a -opioid receptor agonist, on the learned helplessness model of depression in mice. J. Neural Transm. 2002, 109, 12211225. URL http://dx.doi.org/10.1007/s00702-002-0764-x [122] Chourbaji, S.; Zacher, C.; Sanchis-Segura, C.; Dormann, C.; Vollmayr, B.; Gass, P.: Learned helplessness: Validity and reliability of depressivelike states in mice. Brain Res. Protoc. 2005, 16, 7078. URL http://dx.doi.org/10.1016/j.brainresprot.2005.09.002 [123] Broom, D. C.; Jutkiewicz, E. M.; Folk, J. E.; Traynor, J. R.; Rice, K. C.; Woods, J. H.: Nonpeptidic -opioid receptor agonists reduce immobility in the forced swim assay in rats. Neuropsychopharmacology 2002, 26, 744755. URL http://dx.doi.org/10.1016/S0893-133X(01)00413-4 [124] Mague, S. D.; Pliakas, A. M.; Todtenkopf, M. S.; Tomasiewicz, H. C.; Zhang, Y.; Stevens, W. C., J.; Jones, R. M.; Portoghese, P. S.; Carlezon, W. A., J.: Antidepressant-Like Eects of -Opioid Receptor Antagonists in the Forced Swim Test in Rats. J. Pharmacol. Exp. Ther. 2003, 305, 323330. URL http://dx.doi.org/10.1124/jpet.102.046433 [125] Bguin, C.; Carlezon, W.; Cohen, B. M.; He, M.; Lee, D. Y.-W.; Richards, M. R.: Salvinorin derivatives and uses thereof 2005. International Patent WO2005089745. See pp. 64 (hydrogenation), 65 (LiAlH4 reduction), 66 [Ba(OH)2 in MeOH]. URL http://www.patentmatic.com/downloadform.php
BIBLIOGRAPHY
251
[126] Hanes, K. R.: Antidepressant eects of the herb Salvia divinorum: a case report. J. Clin. Psychopharmacol. 2001, 21, 634635. URL http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve& db=PubMed&dopt=Citation&list_uids=11763023 [127] Hanes, K. R.: Salvia divinorum: Clinical and Research Potential. MAPS Bulletin 2003, 13, 1820. URL http://www.maps.org/news-letters/v13n1/13118han.pdf [128] Endoh, T.; Tajima, A.; Izumimoto, N.; Suzuki, T.; Saitoh, A.; Narita, M.; Kamei, J.; Tseng, L. F.; Mizoguchi, H.; Nagase, H.: TRK-820, a selective -opioid agonist, produces potent antinociception in cynomolgus monkeys. Jpn. J. Pharmacol. 2001, 85, 282290. URL http://dx.doi.org/10.1254/jjp.85.282 [129] Mori, T.; Nomura, M.; Nagase, H.; Narita, M.; Suzuki, T.: Eects of a newly synthesized -opioid receptor agonist, TRK-820, on the discriminative stimulus and rewarding eects of cocaine in rats. Psychopharmacology 2002, 161, 1722. URL http://dx.doi.org/10.1007/s00213-002-1028-z [130] Tortella, F. C.; Rose, J.; Robles, L.; Moreton, J. E.; Hughes, J.; Hunter, J. C.: EEG Spectral Analysis of the Neuroprotective Kappa Opioids Enadoline and PD117302. J. Pharmacol. Exp. Ther. 1997, 282, 286 293. URL http://jpet.aspetjournals.org/cgi/content/abstract/282/1/286 [131] Butelman, E. R.; Ko, M. C.; Traynor, J. R.; Vivian, J. A.; Kreek, M. J.; Woods, J. H.: GR89,696: a potent -opioid agonist with subtype selectivity in rhesus monkeys. J. Pharmacol. Exp. Ther. 2001, 298, 1049 1059. And references therein. URL http://jpet.aspetjournals.org/cgi/content/full/298/3/1049 [132] Connor, M.; Kitchen, I.: Has the sun set on 3 -opioid receptors? Br. J. Pharmacol. 2006, 147, 349350. URL http://dx.doi.org/10.1038/sj.bjp.0706603 [133] Schultes, R.; Hofmann, A.: The Botany and Chemistry of Hallucinogens. Charles C. Thomas, Springeld, IL, 2nd ed. 1980. See esp. p. 26 re structural types. Note that the phenylalkylamine skeletal types contain far more active compounds than the others. URL http://www.amazon.com/gp/reader/0398038635 [134] Lloyd, E. J.; Andrews, P. R.: A common structural model for central nervous system drugs and their receptors. J. Med. Chem. 1986, 29, 453 462. Figures of 82% and 58% taken from the appendix. Other gures calculated based on table IV. URL http://dx.doi.org/10.1021/jm00154a005
252
BIBLIOGRAPHY
[135] Chavkin, C.; Sud, S.; Jin, W.; Stewart, J.; Zjawiony, J. K.; Siebert, D. J.; Toth, B. A.; Hufeisen, S. J.; Roth, B. L.: Salvinorin A, an Active Component of the Hallucinogenic Sage Salvia divinorum Is a Highly Ecacious -Opioid Receptor Agonist: Structural and Functional Considerations. J. Pharmacol. Exp. Ther. 2004, 308, 11971203. URL http://dx.doi.org/10.1124/jpet.103.059394 [136] Harding, W. W.; Schmidt, M.; Tidgewell, K.; Kannan, P.; Holden, K. G.; Gilmour, B.; Navarro, H.; Rothman, R. B.; Prisinzano, T. E.: Synthetic Studies of Neoclerodane Diterpenes from Salvia divinorum: Semisynthesis of Salvinicins A and B and Other Chemical Transformations of Salvinorin A. J. Nat. Prod. 2006, 69, 107112. URL http://dx.doi.org/10.1021/np050398i [137] Lee, D. Y. W.; Ma, Z.; Liu-Chen, L.-Y.; Wang, Y.; Chen, Y.; Carlezon, W. A.; Cohen, B.: New neoclerodane diterpenoids isolated from the leaves of Salvia divinorum and their binding anities for human -opioid receptors. Bioorg. Med. Chem. 2005, 13, 56355639. URL http://dx.doi.org/10.1016/j.bmc.2005.05.054 [138] Nagase, H.; Hayakawa, J.; Kawamura, K.; Kawai, K.; Takezawa, Y.; Matsuura, H.; Tajima, C.; Endo, T.: Discovery of a structurally novel opioid -agonist derived from 4,5-epoxymorphinan. Chem. Pharm. Bull. 1998, 46, 366369. URL http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve& db=PubMed&dopt=Citation&list_uids=9501472 [139] Seki, T.; Awamura, S.; Kimura, C.; Ide, S.; Sakano, K.; Minami, M.; Nagase, H.; Satoh, M.: Pharmacological properties of TRK-820 on cloned -, - and -opioid receptors and nociceptin receptor. Eur. J. Pharmacol. 1999, 376, 159167. URL http://dx.doi.org/10.1016/S0014-2999(99)00369-6 [140] Zech, D. F.; Grond, S.; Lynch, J.; Hertel, D.; Lehmann, K. A.: Validation of World Health Organization Guidelines for cancer pain relief: a 10-year prospective study. Pain 1995, 63, 6576. URL http://dx.doi.org/10.1016/0304-3959(95)00017-M [141] Australian and New Zealand College of Anaesthetists: Acute Pain Management: Scientic Evidence. National Health and Medical Research Council, Canberra, 2nd ed. 2005. URL http://www.nhmrc.gov.au/publications/_les/cp104.pdf [142] Ref 15, pp. 103, 112. [143] Ref 15, p. 165. [144] Lewis, Sr, R. J.: Niacin (NCQ900): record number 16623. In Saxs Dangerous Properties of Industrial Materials, Wiley, New York, 10th ed.
BIBLIOGRAPHY 2000. URL http://www.knovel.com/knovel2/Toc.jsp?BookID=707
253
[145] ten Te, A. J.; Verwe, J.; Loos, W. J.; Sparreboom, A.: Pharmacological eects of formulation vehicles : implications for cancer chemotherapy. Clin. Pharmacokinet. 2003, 42, 665685. URL http://www.ingentaconnect.com/content/adis/cpk/2003/ 00000042/00000007/art00005 [146] Mowry, M.; Mosher, M.; Briner, W.: Acute physiologic and chronic histologic changes in rats and mice exposed to the unique hallucinogen salvinorin A. J. Psychoactive Drugs 2003, 35, 379 382. Manuscript available online. Accessed March 2006. URL http://www.sagewisdom.org/mowryetal.pdf [147] Boire, R. G.; Russo, E.; Fish, A. R.; Bowman, J.: Salvia divinorum: information concerning the plant and its active principle. Center for Cognitive Liberty and Ethics, Davis, CA, 2nd ed. 2002. URL http://www.cognitiveliberty.org/pdf/salvia_dea.pdf [148] Gottlieb, A.: Legal Highs. High Times/Level Press, San Francisco 1973. [149] Grubber, H.: Growing the Hallucinogens. High Times/Level Press, San Francisco 1973. [150] Record of the Reasons, 33rd Meeting, 20-22 November 2001. National Drugs and Poisons Schedule Committee, Therapeutic Goods Administration, Commonwealth Department of Health and Ageing, Canberra 2001. Accessed March 2006. URL http://tga.health.gov.au/ndpsc/record/rr200111.pdf [151] White, W. E.: All About Salvia Divinorum. v 1.0 ed. 1995. Posted to alt.drugs, 17th October 1995. Accessed March 2006. URL http://groups.google.com.au/group/alt.drugs/browse_thread/ thread/93775d77997065f9 [152] Moss, G. P.: Nomenclature of fused and bridged fused ring systems. Pure Appl. Chem. 1998, 70, 143216. Accessed March 2006. URL http://www.chem.qmul.ac.uk/iupac/fusedring/ [153] Munro, T.: The Prohibition of Diviners Sage 2002. Accessed March 2006. URL http://www.thomasmunro.com/salvia.htm [154] Leask, E.: Should Diviners Sage be banned? Australian Vital 2002, 1(3), 149151. [155] Record of the Reasons, 36th Meeting, 15-17 October 2002. National Drugs and Poisons Schedule Committee, Therapeutic Goods Administration, Commonwealth Department of Health and Ageing, Canberra
254
BIBLIOGRAPHY 2002. Accessed March 2006. URL http://tga.health.gov.au/ndpsc/record/rr200210.pdf
[156] alt.drugs (usenet group). Accessed March 2006. URL http://groups.google.com/group/alt.drugs?hl=en [157] Guidelines for the Classication of Publications. Oce of Film and Literature Classication, Canberra 2005. See p. 16. Accessed March 2006. URL http://www.oc.gov.au/resource.html?resource=63&lename=63. pdf [158] Andrews, K.: Salvinorin A (Question No. 1963), 11th August 2003. In Parliamentary Debates, House of Representatives, Commonwealth of Australia, Canberra, pp. 1814218143. Accessed March 2006. URL http://www.aph.gov.au/hansard/reps/dailys/dr110803.pdf [159] McEwen, J.: Letter to Lindsay Tanner MP, 8th December 2003. [160] Hartley, M.: Freedom of Information Request No: 117/04. Letter to Karl Hanes, 7th January 2005. [161] BEK nr 714 af 14/08/2003 (Gldende). Retsinformation, Legal Information Division, Ministry of Justice, Denmark. URL http://www.retsinfo.dk/deln/html/b2003/0071405.htm [162] Siebert, D. J.: The legal status of Salvia divinorum 2006. Accessed March 2006. URL http://www.sagewisdom.org/legalstatus.html [163] United States. Hallucinogen Control Act of 2002, HR 5607, 107th Congress, 2nd session, 10th October 2002. URL http://thomas.loc.gov/cgi-bin/query/z?c107:H.R.5607: [164] Gruber, J. W.: Quantication of Salvinorin A from Tissues of Salvia divinorum (Epling & Jtiva-M.). Masters thesis, Philadelphia College of Pharmacy and Science, Philadelphia, PA 1997. ProQuest document ID: 740502451. Publication Number: AAT 1385726. See pp. 46 - 50 (eect of temperature on recovery). URL http://wwwlib.umi.com/dissertations/fullcit/1385726 [165] Valds, III, L. J. J. Email, 22nd July 2005. [166] Sphere: Salvia Divinorum Extractions using Chilled Acetone. Accessed March 2006. URL http://photos.imageevent.com/sphere/ salviadivinorumextractiontech/Chilled_acetone_extraction.pdf [167] Ref 164, pp. 35-37 (isolation procedure). [168] Ref 164, pp. 52-54 (purity of 1a).
BIBLIOGRAPHY
255
[169] Tidgewell, K.; Harding, W. W.; Schmidt, M.; Holden, K. G.; Murry, D. J.; Prisinzano, T. E.: A facile method for the preparation of deuterium labeled salvinorin A: synthesis of [2,2,2-2 H3 ]-salvinorin A. Bioorg. Med. Chem. Lett. 2004, 14, 50995102. URL http://dx.doi.org/10.1016/j.bmcl.2004.07.081 [170] Davies, K. M., ed.: Plant pigments and their manipulation. Blackwell, Oxford 2004. See esp. pp. 1 - 22 (general introduction). [171] Goodwin, T. W., ed.: Chemistry and Biochemistry of Plant Pigments. Academic Press, London 1965. See vol.1 (general introduction), vol. 2 p.19 (partitioning chlorophylls). [172] Vogel, A. I.; Furniss, B. S.; Hannaford, A. J.; Rogers, V.; Smith, P. W. G.; Tatchell, A. R.: Vogels Textbook of Practical Organic Chemistry, Including Qualitative Organic Analysis. Longman, London, 4th ed. 1978. See p. 64 (Rh vs Pd catalysts), 109 - 110 (use of decolourising carbon), p. 493 (KOH/MeOH ester solvolysis). [173] Hassler, J. W.: Purication with Activated Carbon: industrial, commercial, environmental. Chemical Pub. Co., New York, 3rd ed. 1974. See esp. pp. 15 ,16 (general factors governing adsorption). [174] Sanghi, R.; Bhattacharya, B.: Review on decolorisation of aqueous dye solutions by low cost adsorbents. Color. Technol. 2002, 118, 256269. Accessed March 2006. URL http://www.ingentaconnect.com/content/sdc/ct/2002/00000118/ 00000005/art00009 [175] Belfort, G.; Altshuler, G. L.; Thallam, K. K.; Feerick, C. P.; Woodeld, K. L.: Selective adsorption of organic homologues onto activated carbon from dilute aqueous solutions: Solvophobic interaction approach: Part IV. Eect of simple structural modications with aliphatics. AIChE J. 1984, 30, 197207. URL http://dx.doi.org/10.1002/aic.690300205 [176] Vailaya, A.; Horvth, C.: Retention in reversed-phase chromatography: partition or adsorption? J. Chromatogr., A 1998, 829, 127. See esp. pp. 9, 11. URL http://dx.doi.org/10.1016/S0021-9673(98)00727-4 [177] Luehrs, D. C.; Hickey, J. P.; Nilsen, P. E.; Godbole, K. A.; Rogers, T. N.: Linear Solvation Energy Relationship of the Limiting Partition Coecient of Organic Solutes between Water and Activated Carbon. Environ. Sci. Technol. 1996, 30, 143152. See esp. Table 1, entries 129, 160, 165 and 174; 277 and 289; 348, 358, 361 and 366 (adsorption constants of benzene, toluene, p-xylene and o-dichlorobenzene from 3 studies - their refs. 14,17 and 27). URL http://dx.doi.org/10.1021/es950200o
256
BIBLIOGRAPHY
[178] Chiou, C. C. T.; Manes, M.: Application of the Polanyi Adsorption Potential Theory to Adsorption from Solution on Activated Carbon. IV. Steric Factors, as Illustrated by the Adsorption of Planar and Octahedral Metal Acetylacetonates. J. Phys. Chem. 1973, 77, 809813. URL http://dx.doi.org/10.1021/j100625a015 [179] Dbrowski, A.; Podkocielny, P.; Hubicki, Z.; Barczak, M.: Adsorption of phenolic compounds by activated carbon - a critical review. Chemosphere 2005, 58, 10491070. URL http://dx.doi.org/10.1016/j.chemosphere.2004.09.067 [180] Gordon, A. J.; Ford, R. A.: The chemists companion: a handbook of practical data, techniques, and references. Wiley, New York 1972. See pp. 371 - 376 (chromatography on activated carbon). [181] Wu, Z.; Xie, L.; Xia, G.; Zhang, J.; Nie, Y.; Hu, J.; Wang, S.; Zhang, R.: A new tetrodotoxin-producing actinomycete, Nocardiopsis dassonvillei, isolated from the ovaries of puer sh Fugu rubripes. Toxicon 2005, 45, 851859. URL http://dx.doi.org/10.1016/j.toxicon.2005.02.005 [182] Naundorf, A.; Ajisaka, K.: Purication of -N -acetyl-galactosaminidase from Aspergillus niger and its use in the synthesis of GalNAc--(1O)serine. Enzyme Microb. Technol. 1999, 25, 483488. URL http://dx.doi.org/10.1016/S0141-0229(99)00073-3 [183] Aksu, Z.; Tun, O.: Application of biosorption for penicillin G removal: comparison with activated carbon. Process Biochem. 2005, 40, 831847. URL http://dx.doi.org/10.1016/j.procbio.2004.02.014 [184] Li, S. Z.; Li, X. Y.; Cui, Z. F.; Wang, D. Z.: Application of ultraltration to improve the extraction of antibiotics. Sep. Purif. Technol. 2004, 34, 115123. URL http://dx.doi.org/10.1016/S1383-5866(03)00185-0 [185] Scrivens, W. A.; Bedworth, P. V.; Tour, J. M.: Purication of Gram Quantities of C60 . A New Inexpensive and Facile Method. J. Am. Chem. Soc. 1992, 114, 7917 7919. URL http://dx.doi.org/10.1021/ja00046a051 [186] Scrivens, W. A.; Cassell, A. M.; North, B. L.; Tour, J. M.: Single Column Purication of Gram Quantities of C70 . J. Am. Chem. Soc. 1994, 116, 69396940. URL http://dx.doi.org/10.1021/ja00094a060 [187] Athanasiadou, M.; Jensen, S.; Klasson Wehler, E.: Preparative Fractionation of a Commercial PCB Product. Chemosphere 1991, 23, 957970. URL http://dx.doi.org/10.1016/0045-6535(91)90123-U
BIBLIOGRAPHY
257
[188] Koan, A.; Petrk, J.; Chovancov, J.; Drobn, B.: Method for the group separation of non-ortho-, mono-ortho- and multi-orthosubstituted polychlorinated biphenyls and polychlorinated dibenzop-dioxins/polychlorinated dibenzofurans using activated carbon chromatography. J. Chromatogr., A 1994, 665, 139153. URL http://dx.doi.org/10.1016/0021-9673(94)87042-X [189] Sericano, J. L.; El-Husseini, A. M.; Wade, T. L.: Isolation of planar polychlorinated biphenyls by carbon column chromatography. Chemosphere 1991, 23, 915924. URL http://dx.doi.org/10.1016/0045-6535(91)90096-V [190] Zupani-Kralj, L.; Jan, J.; Marsel, J.: Fractionation of chloroorganic compounds on a carbon cartridge. Chemosphere 1991, 23, 841843. URL http://dx.doi.org/10.1016/0045-6535(91)90089-V [191] Lombardo, M. E.; Viscelli, T. A.; Mittelman, A.; Hudson, P. B.: The Removal of Non-steroidal Pigments of Urinary Extracts by Adsorption on Charcoal. J. Biol. Chem. 1955, 212, 353360. URL http://intl.jbc.org/cgi/reprint/212/1/353 [192] Murai, M.; Takenaka, T.; Nishibe, S.: Iridoids from Plantago major. Nat. Med. 1996, 50, 306. [193] MacRae, H. F.; McKinley, W. P.: Chromatographic Identication of Some Organophosphate Insecticides in the Presence of Plant Extracts. J. Agric. Food Chem. 1963, 11, 174178. URL http://dx.doi.org/10.1021/jf60126a023 [194] Leonard, J.; Lygo, B.; Procter, G.: Advanced Practical Organic Chemistry. Blackie Academic & Professional, London, 2nd ed. 1995. See pp. 208 - 214. [195] Ribeiro, M. H. L.; Lourenco, P. A. S.; Monteiro, J. P.; Ferreira-Dias, S.: Kinetics of selective adsorption of impurities from a crude vegetable oil in hexane to activated earths and carbons. Eur. Food Res. Technol. 2001, 213, 132138. URL http://dx.doi.org/10.1007/s002170100347 [196] Bigham, A. K.; Munro, T. A.; Rizzacasa, M. A.; Robins-Browne, R. M.: Divinatorins A-C, New Neoclerodane Diterpenoids from the Controlled Sage Salvia divinorum. J. Nat. Prod. 2003, 66, 12421244. URL http://dx.doi.org/10.1021/np030313i [197] Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.: Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Adv. Drug Delivery Rev. 1997, 23, 325. URL http://dx.doi.org/10.1016/S0169-409X(96)00423-1
258
BIBLIOGRAPHY
[198] Kavvadias, D.; Monschein, V.; Sand, P.; Riederer, P.; Schreier, P.: Constituents of sage (Salvia ocinalis) with in vitro anity to human brain benzodiazepine receptor. Planta Med. 2003, 69, 113117. URL http://dx.doi.org/10.1055/s-2003-37712 [199] Huen, M. S. Y.; Hui, K.-M.; Leung, J. W. C.; Sigel, E.; Baur, R.; Wong, J. T.-F.; Xue, H.: Naturally occurring 2-hydroxyl-substituted avonoids as high-anity benzodiazepine site ligands. Biochem. Pharmacol. 2003, 66, 23972407. URL http://dx.doi.org/10.1016/j.bcp.2003.08.016 [200] Salgueiro, J. B.; Ardenghi, P.; Dias, M.; Ferreira, M. B. C.; Izquierdo, I.; Medina, J. H.: Anxiolytic natural and synthetic avonoid ligands of the central benzodiazepine receptor have no eect on memory tasks in rats. Pharmacol. Biochem. Behav. 1997, 58, 887891. URL http://dx.doi.org/10.1016/S0091-3057(97)00054-3 [201] Zanoli, P.; Avallone, R.; Baraldi, M.: Behavioral characterization of the avonoids apigenin and chrysin. Fitoterapia 2000, 71, S117S123. URL http://dx.doi.org/10.1016/S0367-326X(00)00186-6 [202] Gonzlez, A. G.; Aguiar, Z. E.; Luis, J. G.; Ravelo, A. G.; Vzquez, J. T.; Domnguez, X. A.: Flavonoids from Salvia texana. Phytochemistry 1989, 28, 28712872. URL http://dx.doi.org/10.1016/S0031-9422(00)98114-7 [203] Valds, III, L. J. J. Email, 1st Oct. 2002. [204] Medana, C. Email, 2nd Nov. 2005. [205] Goutman, J. D.; Waxemberg, M. D.; Donate-Oliver, F.; Pomata, P. E.; Calvo, D. J.: Flavonoid modulation of ionic currents mediated by GABAA and GABAC receptors. Eur. J. Pharmacol. 2003, 461, 7987. URL http://dx.doi.org/10.1016/S0014-2999(03)01309-8 [206] Lutterodt, G. D.: Inhibition of gastrointestinal release of acetylcholine by quercetin as a possible mode of action of Psidium guajava leaf extracts in the treatment of acute diarrhoeal disease. J. Ethnopharmacol. 1989, 25, 235247. URL http://dx.doi.org/10.1016/0378-8741(89)90030-5 [207] Capasso, R.; Borrelli, F.; Capasso, F.; Siebert, D. J.; Stewart, D. J.; Zjawiony, J. K.; Izzo, A. A.: The hallucinogenic herb Salvia divinorum and its active ingredient salvinorin A inhibit enteric cholinergic transmission in the guinea-pig ileum. Neurogastroenterol. Motil. 2006, 18, 6975. URL http://dx.doi.org/10.1111/j.1365-2982.2005.00725.x [208] Singh, A.; Naidu, P. S.; Kulkarni, S. K.: Quercetin, a bioavonoid, reverses development of tolerance and dependence to morphine. Drug
BIBLIOGRAPHY Dev. Res. 2002, 57, 167172. URL http://dx.doi.org/10.1002/ddr.10119
259
[209] Ma, M.-H.; Lin, C.-I.: Adsorption kinetics of -carotene from soy oil using regenerated clay. Sep. Purif. Technol. 2004, 39, 201209. URL http://dx.doi.org/10.1016/j.seppur.2003.12.007 [210] Tanin, S.; Gurgey, I.: Bleaching of cottonseed and sunower oils by active carbons obtained from rice hulls. Chim. Acta Turc. 1988, 16, 209219. [211] Ortega, A. Email, 15th Oct. 2005. [212] Delgado, G.; Romo de Vivar, A.; Ortega, A.; Crdenas, J.; Schlemper, E. O.: Diterpenoids from Viguiera insignis. Phytochemistry 1983, 22, 12271230. URL http://dx.doi.org/10.1016/0031-9422(83)80227-1 [213] Jimenez, M.; Ortega, A.; Navarro, A.; Maldonado, E.; Van Calsteren, M. R.; Jankowski, C. K.: Reaction of the Molluscicide Glaucolide B with Bentonite. J. Nat. Prod. 1995, 58, 424427. URL http://dx.doi.org/10.1021/np50117a012 [214] Ruvalcaba, R. M.; Arroyo Razo, G. A.; Carrillo, G. P.; Reyes, F. D.; Ortiz, A. C.; Toledano, C. A.; Salazar, M. S.: Preparative heterocyclic chemistry using tonsil, a bentonitic clay: 1981 to 2003. Trends Heterocycl. Chem. 2003, 9, 195235. [215] Munro, T. A.; Rizzacasa, M. A.: Salvinorins D-F, New Neoclerodane Diterpenoids from Salvia divinorum, and an Improved Method for the Isolation of Salvinorin A. J. Nat. Prod. 2003, 66, 703705. URL http://dx.doi.org/10.1021/np0205699 [216] Ohsaki, A.; Yan, L. T.; Ito, S.; Edatsugi, H.; Iwata, D.; Komoda, Y.: The isolation and in vivo potent antitumor activity of clerodane diterpenoid from the oleoresin of the Brazilian medicinal plant, Copaifera langsdori Desfon. Bioorg. Med. Chem. Lett. 1994, 4, 28892892. URL http://dx.doi.org/10.1016/S0960-894X(01)80834-9 [217] McChesney, J. D.; Clark, A. M.; Silveira, E. R.: Antimicrobial diterpenes of Croton sonderianus, 1. Hardwickic and 3,4-secotrachylobanoic acids. J. Nat. Prod. 1991, 54, 16251633. URL http://dx.doi.org/10.1021/np50078a021 [218] Costa, M.; Tanaka, C. M. A.; Imamura, P. M.; Marsaioli, A. J.: Isolation and synthesis of a new clerodane from Echinodorus grandiorus. Phytochemistry 1999, 50, 117122. URL http://dx.doi.org/10.1016/S0031-9422(98)00464-6 [219] Spectral Database for Organic Compounds (SDBS), National Institute of Advanced Industrial Science and Technology (AIST), Japan. (E)-Phytol
260
BIBLIOGRAPHY = compound 7578 (accessed Oct. 2005). URL http://www.aist.go.jp/RIODB/SDBS/cgi-bin/cre_index.cgi
[220] Boxer, S. G.; Closs, G. L.; Katz, J. J.: Eect of magnesium coordination on the 13 C and 15 N magnetic resonance spectra of chlorophyll a. The Relative energies of nitrogen n* states as deduced from a complete assignment of chemical shifts. J. Am. Chem. Soc. 1974, 96, 70587066. URL http://dx.doi.org/10.1021/ja00829a038 [221] Huang, K.-F.; Hsu, C.-J.: Constituents of stem bark of Erythrina arborescens. J. Chin. Med. (Taiwan) 2001, 12, 61 67. (accessed Oct. 2005). URL http://www.nricm.edu.tw/jcm/012/12-1-06.pdf [222] Altman, L. J.; Kowerski, R. C.; Laungani, D. R.: Studies in terpene biosynthesis. Synthesis and resolution of presqualene and prephytoene alcohols. J. Am. Chem. Soc. 1978, 100, 61746182. URL http://dx.doi.org/10.1021/ja00487a037 [223] Rogers, D. H.; Yi, E. C.; Poulter, C. D.: Enantioselective Synthesis of (+)-Presqualene Diphosphate. J. Org. Chem. 1995, 60, 941945. URL http://dx.doi.org/10.1021/jo00109a026 [224] Giner, J.-L.; Berkowitz, J. D.; Andersson, T.: Nonpolar Components of the Latex of Euphorbia peplus. J. Nat. Prod. 2000, 63, 267269. URL http://dx.doi.org/10.1021/np990081g [225] Misra, R.; Pandey, R. C.; Dev, S.: The chemistry of the oleo resin from hardwickia pinnata: a series of new diterpenoids. Tetrahedron Lett. 1964, 5, 37513759. URL http://dx.doi.org/10.1016/S0040-4039(01)89373-4 [226] Misra, R.; Pandey, R. C.; Dev, S.: Absolute stereochemistry of hardwickiic acid and its congeners. Tetrahedron Lett. 1968, 9, 26812684. URL http://dx.doi.org/10.1016/S0040-4039(00)89672-0 [227] Misra, R.; Pandey, R. C.; Dev, S.: Higher isoprenoidsVIII: Diterpenoids from the oleoresin of hardwickia pinnata part 1: hardwickiic acid. Tetrahedron 1979, 35, 23012310. URL http://dx.doi.org/10.1016/0040-4020(79)80125-8 [228] Chaichantipyuth, C.; Muangsin, N.; Chaichit, N.; Roengsumran, S.; Petsom, A.; Watanabe, T.; Ishikawa, T.: Crystal structure of (-)hardwickiic acid, C19 H27 OCOOH. Z. Kristallogr. - New Cryst. Struct. 2004, 219, 111113. (accessed Oct 2005). URL http://www.oldenbourg.de/verlag/zkristallogr/mn-ncsc0402. htm#p111
BIBLIOGRAPHY
261
[229] Ortega, A.; Crdenas, J.; Gage, D. A.; Maldonado, E.: Abietane and clerodane diterpenes from Salvia regla. Phytochemistry 1995, 39, 931 933. URL http://dx.doi.org/10.1016/0031-9422(95)00083-J [230] Bandara, B. M. R.; Wimalasiri, W. R.; Bandara, K. A. N. P.: Isolation and insecticidal activity of (-)-hardwickiic acid from Croton aromaticus. Planta Med. 1987, 53, 575. [231] Cocker, W.; Moore, A. L.; Pratt, A. C.: Dextrorotatory hardwickiic acid, an extractive of Copaifera ocinalis. Tetrahedron Lett. 1965, 6, 19831985. URL http://dx.doi.org/10.1016/S0040-4039(01)83897-1 [232] Cascon, V.; Gilbert, B.: Characterization of the chemical composition of oleoresins of Copaifera guianensis Desf., Copaifera duckei Dwyer and Copaifera multuga Hayne. Phytochemistry 2000, 55, 773778. URL http://dx.doi.org/10.1016/S0031-9422(00)00284-3 [233] Komiya, T.; Kyohkon, M.; Ohwaki, S.; Eto, J.; Katsuzaki, H.; Imai, K.; Kataoka, T.; Yoshioka, K.; Ishii, Y.; Hibasami, H.: Phytol induces programmed cell death in human lymphoid leukemia Molt 4B cells. Int. J. Mol. Med. 1999, 4, 377380. [234] Tomita, Y.: Immunological role of vitamin A and its related substances in prevention of cancer. Nutr. Cancer 1983, 5, 187194. [235] Khler, I.; Jenett-Siems, K.; Kraft, C.; Siems, K.; Abbiw, D.; Bienzle, U.; Eich, E.: Herbal remedies traditionally used against malaria. Part 7. Herbal remedies traditionally used against malaria in Ghana: Bioassay-guided fractionation of Microglossa pyrifolia (Asteraceae). Z. Naturforsch., C: J. Biosci. 2002, 57, 10221027. URL http://www.znaturforsch.com/sc/57c/s57c1022.pdf [236] Rajab, M. S.; Cantrell, C. L.; Franzblau, S. G.; Fisher, N. H.: Antimycobacterial activity of (E)-phytol and derivatives. A preliminary structure-activity study. Planta Med. 1998, 64, 24. [237] Saludes, J. P.; Garson, M. J.; Franzblau, S. G.; Aguinaldo, A. M.: Antitubercular constituents from the hexane fraction of Morinda citrifolia Linn. (Rubiaceae). Phytother. Res. 2002, 16, 683685. URL http://dx.doi.org/10.1002/ptr.1003 [238] Arnhold, T.; Elmazar, M. M. A.; Nau, H.: Prevention of Vitamin A Teratogenesis by Phytol or Phytanic Acid Results from Reduced Metabolism of Retinol to the Teratogenic Metabolite, All-trans-retinoic Acid. Toxicol. Sci. 2002, 66, 274282. URL http://dx.doi.org/10.1093/toxsci/66.2.274
262
BIBLIOGRAPHY
[239] Topu, G.: Bioactive Triterpenoids from Salvia Species. J. Nat. Prod. 2006, 69, 482487. URL http://dx.doi.org/10.1021/np0600402 [240] Ovesn, Z.; Vachlkov, A.; Horvthov, K.; Tthov, D.: Pentacyclic triterpenoic acids: new chemoprotective compounds. Neoplasma 2004, 51, 327333. URL http://www.neoplasma.sk/abstrakt.asp?idn=645 [241] Liu, J.: Pharmacology of oleanolic acid and ursolic acid. J. Ethnopharmacol. 1995, 49, 5768. URL http://dx.doi.org/10.1016/0378-8741(95)01310-5 [242] Ma, C.-m.; Nakamura, N.; Hattori, M.; Kakuda, H.; Qiao, J.-c.; Yu, H.l.: Inhibitory eects on HIV-1 protease of constituents from the wood of Xanthoceras sorbifolia. J. Nat. Prod. 2000, 63, 238242. URL http://dx.doi.org/10.1021/np9902441 [243] Watanabe, M.; Kobayashi, Y.; Ogihara, J.; Kato, J.; Oishi, K.: HIV-1 reverse transcriptase-inhibitory compound in Salvia ocinalis. Food Sci. Technol. Res. 2000, 6, 216220. [244] Ikeda, T.; Yokomizo, K.; Okawa, M.; Tsuchihashi, R.; Kinjo, J.; Nohara, T.; Uyeda, M.: Anti-herpes virus type 1 activity of oleanane-type triterpenoids. Biol. Pharm. Bull. 2005, 28, 17791781. URL http://dx.doi.org/10.1248/bpb.28.1779 [245] Woldemichael, G. M.; Singh, M. P.; Maiese, W. M.; Timmermann, B. N.: Constituents of antibacterial extract of Caesalpinia paraguariensis Burk. Z. Naturforsch., C: J. Biosci. 2003, 58, 7075. URL http://www.znaturforsch.com/s58c/s58c0070.pdf [246] Weimann, C.; Goransson, U.; Pongprayoon-Claeson, U.; Claeson, P.; Bohlin, L.; Rimpler, H.; Heinrich, M.: Spasmolytic eects of Baccharis conferta and some of its constituents. J. Pharm. Pharmacol. 2002, 54, 99104. URL http://www.aapspharmaceutica.com/search/view.asp?ID=5187 [247] Wachter, G. A.; Valcic, S.; Flagg, M. L.; Franzblau, S. G.; Montenegro, G.; Suarez, E.; Timmermann, B. N.: Antitubercular activity of pentacyclic triterpenoids from plants of Argentina and Chile. Phytomedicine 1999, 6, 341345. [248] Tan, N.; Kaloga, M.; Radtke, O. A.; Kiderlen, A. F.; ksz, S.; Ulubelen, A.; Kolodziej, H.: Abietane diterpenoids and triterpenoic acids from Salvia cilicica and their antileishmanial activities. Phytochemistry 2002, 61, 881884. URL http://dx.doi.org/10.1016/S0031-9422(02)00361-8
BIBLIOGRAPHY
263
[249] Mallavadhani, U. V.; Mahapatra, A.; Raja, S. S.; Manjula, C.: Antifeedant activity of some pentacyclic triterpene acids and their fatty acid ester analogs. J. Agric. Food Chem. 2003, 51, 19521955. URL http://dx.doi.org/10.1021/jf020691d [250] Blagg, B. S. J.; Jarstfer, M. B.; Rogers, D. H.; Poulter, C. D.: Recombinant Squalene Synthase. A Mechanism for the Rearrangement of Presqualene Diphosphate to Squalene. J. Am. Chem. Soc. 2002, 124, 88468853. URL http://dx.doi.org/10.1021/ja020411a [251] Faure, S.; Connolly, J. D.; Fakunle, C. O.; Piva, O.: Structure and Synthesis of Anhydrobisfarnesol from Euphorbia lateriora and Asymmetric Synthesis of (R)-Sesquilavandulol. Tetrahedron 2000, 56, 96479653. URL http://dx.doi.org/10.1016/S0040-4020(00)00913-3 [252] Connolly, J. D.: Natural products from around the world. Pure Appl. Chem. 2001, 73, 567571. URL http://www.iupac.org/publications/pac/2001/pdf/7303x0567.pdf [253] Connolly, J. D.: Natural products from around the world. Rev. Latinoamer. Qum. 1997, 25, 7785. [254] Van Tamelen, E. E.; Leopold, E. J.: Mechanism of presqualene pyrophosphate-squalene biosynthesis. II. Synthesis of bifarnesol. Tetrahedron Lett. 1985, 26, 33033306. URL http://dx.doi.org/10.1016/S0040-4039(00)98283-2 [255] Valds, III, L. J. J.: Divinorin C, a new neoclerodane diterpene from a bioactive TLC fraction of Salvia divinorum. Entheogen Rev. 2000, 9, 141. Reprinted in Ref. 89. URL http://www.sagewisdom.org/divinorinc.html [256] Medana, C. Emails, 10th February, 4th and 5th March 2005. [257] PC Model v5.04 for Mac; Serena Software: Bloomington, IN, 1994. URL http://www.serenasoft.com [258] Grof, S.: Stanislav Grof interviews Dr. Albert Hofmann. MAPS Bulletin 2001, 11, 2235. See p. 29. URL http://www.maps.org/news-letters/v11n2/11222gro.pdf [259] Lee-Stecum, P. Email, 7th May 2003. [260] Brown, L.: The Stereocontrolled Synthesis of Optically Active Vitamin E Side Chains. II. Benzoyl Triate and its Application in the Determination of the Absolute Conguration of Divinorin A and B, and Terrecyclic acid. Ph.D. thesis, University of Michigan, Ann Arbor, MI 1984. ProQuest Publication Number: AAT 8422201 (Document ID: 748992941). See pp. 78 80, 196 200 (hydride reductions). URL http://wwwlib.umi.com/dissertations/fullcit/8422201
264 [261] Ref 260, pp. 84, 85 (compound 123). [262] Ref 260, p. 76 (direct diacetylation).
BIBLIOGRAPHY
[263] Ref 260, pp. 72 75 (BH reductions, epimerizations and proposed mech4 anism). [264] Goddard, R.; Akhtar, F.: The structure of jateorin. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1986, C42, 12171220. URL http://dx.doi.org/10.1107/S0108270186092831 [265] Swaminathan, K.; Sinha, U. C.; Ramakumar, S.: Structure of columbin, a diterpenoid furanolactone from Tinospora cordifolia Miers. Acta Crystallogr., Sect. C: Cryst. Struct. Commun. 1989, C45, 300303. URL http://dx.doi.org/10.1107/S0108270188010583 [266] Yonemitsu, M.; Fukuda, N.; Kimura, T.; Komori, T.; Lindner, H. J.; Habermehl, G.: Crystal structure and NMR spectrometric analysis of palmarin. Liebigs Ann. Chem. 1989, pp. 485487. [267] Cheung, K. K.; Melville, D.; Overton, K. H.; Robertson, J. M.; Sim, G. A.: The stereochemistry of isocolumbin: x-ray analysis of the 1-piodophenyl-3-phenylpyrazoline adduct of isocolumbin. J. Chem. Soc., B 1966, pp. 853861. URL http://dx.doi.org/10.1039/J29660000853 [268] Overton, K. H.; Weir, N. G.; Wylie, A.: The stereochemistry of the colombo root bitter principles. J. Chem. Soc., C 1966, pp. 14821490. See isocolumbin, dihydrocolumbin and dihydroisocolumbin (hydrogenations); table 2 (furanolactone infrared absorptions). URL http://dx.doi.org/10.1039/J39660001482 [269] Yonemitsu, M.; Fukuda, N.; Kimura, T.; Komori, T.: Studies on the constituents of Jateorhiza palmata Miers (colombo root). I. Separation and structure of a new furanoid diterpene glucoside (palmatoside A). Liebigs Ann. Chem. 1986, pp. 13271333. See compounds 3, 4 (nujol mulls) and 6 (mull vs CHCl3 ). [270] Munro, T. A.; Rizzacasa, M. A.; Roth, B. L.; Toth, B. A.; Yan, F.: Studies toward the Pharmacophore of Salvinorin A, a Potent -Opioid Receptor Agonist. J. Med. Chem. 2005, 48, 345348. URL http://dx.doi.org/10.1021/jm049438q [271] Lee, D. Y. W.; Karnati, V. V. R.; He, M.; Liu-Chen, L.-Y.; Kondareti, L.; Ma, Z.; Wang, Y.; Chen, Y.; Bguin, C.; Carlezon, W. A.; Cohen, B.: Synthesis and in vitro pharmacological studies of new C(2) modied salvinorin A analogues. Bioorg. Med. Chem. Lett. 2005, 15, 37443747. Epimerization is discussed on p. 3746 c.1. URL http://dx.doi.org/10.1016/j.bmcl.2005.05.048
BIBLIOGRAPHY [272] Ref 260, p. 83 (deacetylations with KCN); cf. p. 202 (yield). [273] Valds, III, L. J. J. Email, 28th Apr. 2003.
265
[274] Khong, P. W.; Lewis, K. G.: Acetyl migration in urs-12-ene-3,24-diol. Aust. J. Chem. 1975, 28, 201206. See also references 1 - 6 therein. URL http://dx.doi.org/10.1071/CH9750201 [275] Smith, M. B.; March, J.: Marchs advanced organic chemistry: reactions, mechanisms, and structure. Wiley, New York, 5th ed. 2001. See Table 8.1, pp. 330 - 331 (pKa of conjugate acids), Section 12-3 pp. 773 - 775 (keto/enol tautomerism). [276] Barton, D. H. R.; Overton, K. H.; Wylie, A.: Diterpenoid bitter principles. IV. Investigations on the constitution of palmarin. J. Chem. Soc. 1962, pp. 48094815. See hydrogenations of palmarin and isojateorin. URL http://dx.doi.org/10.1039/JR9620004809 [277] Balasubramanian, S. K.; Barton, D. H. R.; Jackman, L. M.: Diterpenoid bitter principles. V. Constitution of palmarin and its congeners. J. Chem. Soc. 1962, pp. 48164820. URL http://dx.doi.org/10.1039/JR9620004816 [278] Ref 260, pp. 90 92 (deuteration). The NMR data given are consistent with this structure. The 13 C peak at 171.6 is incorrectly assigned to the lactone, however, on the basis of an apparent reduction in intensity relative to 1a. The apparent reduction was probably due to an inadequate signal-to-noise ratio; ve peaks were not observed. The claimed yield of 89% is implausible given the epimerization we and Valds observed with KCN. [279] Strazzolini, P.; Giumanini, A. G.; Cauci, S.: Acetic formic anhydride: a review. Tetrahedron 1990, 46, 10811118. URL http://dx.doi.org/10.1016/S0040-4020(01)86676-X [280] Reber, F.; Lardon, A.; Reichstein, T.: Teilsynthese von 11-Epicorticosteron. Bestandteile der Nebennierenrinde und verwandte Stoe, 86. Mitteilung. Helv. Chim. Acta 1954, 37, 4558. See compounds XII and XVIII. URL http://dx.doi.org/10.1002/hlca.19540370107 [281] Oliveto, E. P.; Gerold, C.; Rausser, R.; Hershberg, E. B.: 11-Oxygenated Steroids. XII. The Preparation of 17-Hydroxycorticosterone 21-Acetate (Kendalls Compound F Acetate) via 11-Formates. J. Am. Chem. Soc. 1955, 77, 3564 3567. See compound V. URL http://dx.doi.org/10.1021/ja01618a042 [282] Ref 260, pp. 84 87 (benzoylation).
266
BIBLIOGRAPHY
[283] Danklmaier, J.; Hoenig, H.: Regioselektive benzylierungen an Hydroxyalkyl-3-morpholinonen. Liebigs Ann. Chem. 1989, pp. 665669. [284] Eckenberg, P.; Groth, U.; Huhn, T.; Richter, N.; Schmeck, C.: A useful application of benzyl trichloroacetimidate for the benzylation of alcohols. Tetrahedron 1993, 49, 16191624. URL http://dx.doi.org/10.1016/S0040-4020(01)80349-5 [285] Tanabe, M.; Peters, R. H.: (R,S)-Mevalonolactone-2-13 C. Org. Synth. 1990, 60, 92. Also Coll. Vol. 7, p. 386. See note 4. URL http://www.orgsyn.org/orgsyn/pdfs/CV7P0386.pdf [286] Bouzide, A.; Sauve, G.: Highly selective silver (I) oxide mediated monoprotection of symmetrical diols. Tetrahedron Lett. 1997, 38, 59455948. URL http://dx.doi.org/10.1016/S0040-4039(97)01328-2 [287] Charette, A. B.: Tetra-n-butylammonium iodide. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rt018 [288] Kraus, G. A.; Frazier, K. A.; Roth, B. D.; Taschner, M. J.; Neuenschwander, K.: Conversion of lactones into ethers. J. Org. Chem. 1981, 46, 24172419. URL http://dx.doi.org/10.1021/jo00324a050 [289] Lundt, I.: Oxidation, reduction and deoxygenation. In Glycoscience: chemistry and chemical biology, B. O. Fraser-Reid; K. Tatsuta; J. Thiem, eds., Springer, New York, vol. 1, pp. 501531 2001. [290] Hansen, M. C.; Verdaguer, X.; Buchwald, S. L.: Convenient Two-Step Conversion of Lactones into Cyclic Ethers. J. Org. Chem. 1998, 63, 23602361. URL http://dx.doi.org/10.1021/jo9716082 [291] Kawamura, K.; Hinou, H.; Matsuo, G.; Nakata, T.: Ecient strategy for convergent synthesis of trans-fused polycyclic ethers based on an intramolecular SmI2 -promoted cyclization of iodo ester. Tetrahedron Lett. 2003, 44, 52595261. See compound 11. URL http://dx.doi.org/10.1016/S0040-4039(03)01277-2 [292] Miki, Y.; Ohta, M.; Hachiken, H.; Takemura, S.: A simple synthesis of naphtho[1,8-bc]pyran. Synthesis 1990, p. 312. URL http://dx.doi.org/10.1055/s-1990-26860 [293] Kourafalos, V. N.; Marakos, P.; Pouli, N.; Townsend, L. B.: The Synthesis of 4-Deazaformycin A. J. Org. Chem. 2003, 68, 64666469. See compound 7. URL http://dx.doi.org/10.1021/jo026715x
BIBLIOGRAPHY
267
[294] Barbier, M.; Devys, M.; Parisot, D.: Synthesis of anhydronectriachrysone, an extensively conjugated -pyrone. Synth. Commun. 1993, 23, 14811488. [295] Ikawa, T.; Sajiki, H.; Hirota, K.: Unexpected deprotection of silyl and THP ethers induced by serious disparity in the quality of Pd/C catalysts and elucidation of the mechanism. Tetrahedron 2004, 60, 61896195. URL http://dx.doi.org/10.1016/j.tet.2004.05.040 [296] Sajiki, H.; Hattori, K.; Hirota, K.: Highly chemoselective hydrogenation with retention of the epoxide function using a heterogeneous Pd/Cethylenediamine catalyst and THF. Chem. Eur. J. 2000, 6, 22002204. See also references therein. URL http://dx.doi.org/10.1002/1521-3765(20000616)6:12<2200:: AID-CHEM2200>3.0.CO;2-3 [297] Rylander, P. N.: Hydrogenation methods. Academic Press, London 1985. See pp. 10 (use of base), 157+ (catalyst eects). [298] Bruno, M.; Rosselli, S.; Pibiri, I.; Piozzi, F.; Simmonds, M.: Hydrogenation derivatives of neo-clerodanes. Heterocycles 2000, 55, 599612. See hydrogenations of montanin C (4) and teucrin A (6). Site accessed February 2006. URL https://www2.heterocycles.jp/FMPro?-db=journalred.fp5& -format=/w2/charge.html&-lay=data&-recid=35852&-Token. 3=COM-99-8777&-Script=citingAndDuplicate&-Find [299] Housley, J. R.; King, F. E.; King, T. J.; Taylor, P. R.: Chemistry of hardwood extractives. XXXIV. Constituents of Guarea species. J. Chem. Soc. 1962, pp. 50955104. See dihydrogedunene hydrogenations. URL http://dx.doi.org/10.1039/JR9620005095 [300] Malakov, P. Y.; Papanov, G. Y.; Mollov, N. M.; Spassov, S. L.: Montanin-C, a New Furanoid Diterpene from Teucrium montanum L. Z. Naturforsch., B: Anorg. Chem., Org. Chem. 1978, 33B, 789791. [301] Brieskorn, C. H.; Pfeuer, T.: Labiatenbitterstoe: Pikropolin und hnliche Diterpenoide aus Poleigamander. Chem. Ber. 1967, 100, 1998 2010. See picropolin acetate. URL http://dx.doi.org/10.1002/cber.19671000629 [302] Adesogan, E. K.: The structure of penduliaworosin, a new furanoid diterpene from Croton penduliorus. J. Chem. Soc., Perkin Trans. 1 1981, pp. 11511153. URL http://dx.doi.org/10.1039/P19810001151 [303] Bui, A. M.; Parello, J.; Potier, P.; Janot, M. M.: Sur la structure du collybolide, nouvelle substance sesquiterpnique extraite du Collybia maculata Alb. et Sch. ex Fries (Basidiomyctes). C. R. Acad. Sci., C. 1970, 270, 10221026. See 2 and 2.
268
BIBLIOGRAPHY
[304] Chan, W. R.; Taylor, D. R.; Willis, C. R.: Terpenoids from the Euphorbiaceae. I. The structure of crotonin, a norditerpene from Croton lucidus. J. Chem. Soc., C 1968, pp. 27812785. URL http://dx.doi.org/10.1039/J39680002781 [305] Chatterjee, A.; Kundu, A. B.; Chakrabortty, T.; Chandrasekharan, S.: Extractives of Aphanamixis polystachya wall (Parker). Structures and stereochemistry of aphanamixin and aphanamixinin. Tetrahedron 1970, 26, 18591867. See hydrogenations of XXIX. URL http://dx.doi.org/10.1016/S0040-4020(01)92762-0 [306] Emerson, O. H.: Bitter principles in citrus. III. Some reactions of limonin. J. Am. Chem. Soc. 1952, 74, 688693. URL http://dx.doi.org/10.1021/ja01123a031 [307] Ruberto, G.; Renda, A.; Tringali, C.; Napoli, E. M.; Simmonds, M. S. J.: Citrus limonoids and their semisynthetic derivatives as antifeedant agents against Spodoptera frugiperda larvae. A structure-activity relationship study. J. Agric. Food Chem. 2002, 50, 67666774. URL http://dx.doi.org/10.1021/jf020607u [308] De Alvarenga, M. A.; Gottlieb, H. E.; Gottlieb, O. R.; Magalhes, M. T.; Da Silva, V. O.: Diasin, a diterpene from Croton diasii. Phytochemistry 1978, 17, 17731776. URL http://dx.doi.org/10.1016/S0031-9422(00)88692-6 [309] Rodriguez-Hahn, L.; OReilly, R.; Esquivel, B.; Maldonado, E.; Ortega, A.; Cardenas, J.; Toscano, R. A.; Chan, T.-M.: Tilifodiolide, tetraline-type diterpenoid of clerodanic origin from Salvia tiliaefolia. J. Org. Chem. 1990, 55, 35223525. URL http://dx.doi.org/10.1021/jo00298a026 [310] Hori, T.; Kiang, A. K.; Nakanishi, K.; Sasaki, S.; Woods, M. C.: Structures of braurin and a minor product from Fibraurea chloroleuca. Tetrahedron 1967, 23, 26492656. URL http://dx.doi.org/10.1016/0040-4020(67)85129-9 [311] Melera, A.; Schaner, K.; Arigoni, D.; Jeger, O.: Zur Konstitution des Limonins I. ber den Verlauf der alkalischen Hydrolyse von Limonin und Limonol. Helv. Chim. Acta 1957, 40, 14201437. See hydrogenations of I and II, pp. 1431 - 1432. URL http://dx.doi.org/10.1002/hlca.19570400529 [312] Rosenfeld, R. S.; Hofmann, K.: The chemistry of limonin. J. Am. Chem. Soc. 1951, 73, 24912493. URL http://dx.doi.org/10.1021/ja01150a024 [313] Ida, Y.; Kubo, S.; Fujita, M.; Komori, T.; Kawasaki, T.: Struktur der Diosbulbine-D, -E, -F, -G und H. Justus Liebigs Ann. Chem. 1978, pp. 818833. See hydrogenation of compounds D and F.
BIBLIOGRAPHY
269
[314] Popa, D. P.; Reinbold, A. M.: Structure of teucrin A. Chem. Nat. Comp. 1975, 9, 2730. Translation of Khim. Prir. Soedin., 1973, 9, 31-35. URL http://dx.doi.org/10.1007/BF00580883 [315] Pailer, M.; Schaden, G.; Spiteller, G.; Frenzl, W.: Die Konstitution des Fraxinellons aus Dictamnus albus L. Monatsh. Chem. 1965, 96, 1324 1346. URL http://dx.doi.org/10.1007/BF00904284 [316] Arndt, R. R.; Baarschers, W. H.: Structure of phragmalin. Meliacin with a norbornane part skeleton. Tetrahedron 1972, 28, 23332340. URL http://dx.doi.org/10.1016/S0040-4020(01)93576-8 [317] Kouzi, S. A.; McMurtry, R. J.; Nelson, S. D.: Hepatotoxicity of Germander (Teucrium chamaedrys L.) and One of Its Constituent Neoclerodane Diterpenes Teucrin A in the Mouse. Chem. Res. Toxicol. 1994, 7, 850 856. URL http://dx.doi.org/10.1021/tx00042a020 [318] MacKinnon, S.; Durst, T.; Arnason, J. T.; Angerhofer, C.; Pezzuto, J.; Sanchez-Vindas, P. E.; Poveda, L. J.; Gbeassor, M.: Antimalarial Activity of Tropical Meliaceae Extracts and Gedunin Derivatives. J. Nat. Prod. 1997, 60, 336341. See compound 10. URL http://dx.doi.org/10.1021/np9605394 [319] Siegel, S.: Rhodium on Alumina. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rr003 [320] Kabalka, G. W.; Wang, L.; Pagni, R. M.: Potassium uoride doped alumina: an eective reagent for ester hydrolysis under solvent free conditions. Green Chem. 2001, 3, 261262. URL http://dx.doi.org/10.1039/b106423c [321] Ref 260, p. 81 (Li/NH3 reduction); cf. p. 201 (structure). The [M+H]+ value given in the mass spectral data (265) is evidently a typographical error, being smaller than several daughter ions. The correct value is 365. [322] Presser, A.; Hfner, A.: Trimethylsilyldiazomethane - A Mild and Ecient Reagent for the Methylation of Carboxylic Acids and Alcohols in Natural Products. Monatsh. Chem. 2004, 135, 10151022. URL http://dx.doi.org/10.1007/s00706-004-0188-4 [323] Ref 60, p. 153 (cyclic -diones). [324] Raymond, S.: The Structure of Methyl Acetylacrylate. J. Am. Chem. Soc. 1950, 72, 43044306. URL http://dx.doi.org/10.1021/ja01165a537
270
BIBLIOGRAPHY
[325] Caton, M. P. L.; Darnbrough, G.; Parker, T.: Prostaglandins VI: BaseCatalyzed Autoxidation of -Hydroxycyclopentanones and the Synthesis of 9,10-Diketoprostanoic Acids. Tetrahedron Lett. 1980, 21, 16851686. URL http://dx.doi.org/10.1016/S0040-4039(00)77786-0 [326] Sundt, E.; Jeger, O.; Prelog, V.: Conversion of Cevagenin into a Diosphenol. Chem. Ind. 1953, pp. 13651366. [327] Lavie, D.; Shlomo, S.: The Constituents of Ecballium elaterium L. II. -Elaterin. J. Am. Chem. Soc. 1958, 80, 707710. URL http://dx.doi.org/10.1021/ja01536a046 [328] Lack, R. E.; Ridley, A. B.: Autoxidation of 2-Hydroxy-5-Cholestan3-One in Methanolic Potassium Hydroxide. J. Chem. Soc., C 1968, pp. 30173020. URL http://dx.doi.org/10.1039/J39680003017 [329] Weissberger, A.; Schwarze, W.: ber die Autoxydation des 1-Oxy-tetralons. Justus Liebigs Ann. Chem. 1931, 487, 5361. [330] Slavikova, B.; Kasal, A.; Budesinsky, M.: Autoxidation vs hydrolysis in 16-acyloxy steroids. Collect. Czech. Chem. Commun. 1999, 64, 1125 1134. URL http://dx.doi.org/10.1135/cccc19991125 [331] Dauben, W. G.; Boswell, G. A.; Templeton, W.: Base-catalyzed autoxidation of cyclic 1,4-diketones to enediones. J. Org. Chem. 1960, 25, 18531855. URL http://dx.doi.org/10.1021/jo01081a006 [332] Ref 260, p. 82 (PCC oxidation). [333] Gonzlez, A. G.; Aguiar, Z. E.; Luis, J. G.; Ravelo, A. G.: New diterpenes from Salvia texana. Chemical and biogenetic aspects. Tetrahedron 1989, 45, 52035214. See compounds 13a, b. URL http://dx.doi.org/10.1016/S0040-4020(01)81097-8 [334] Ganguly, A. K.; Govindachari, T. R.; Mohamed, P. A.: Oxidation of ring A in lupeol. Tetrahedron 1966, 22, 35973599. URL http://dx.doi.org/10.1016/S0040-4020(01)92548-7 [335] Marchand, A. P.; Reddy, G. M.: Ultrasound-promoted sodium borohydride reduction of pentacyclo[5.4.0.02,6 .03,10 .05,9 ]undecane-8,11-dione (PCUD-8,11-dione) and of 4,4-dimethoxy-2,3,5,6-tetrachloro-PCUD8,11-dione. Org. Prep. Proced. Int. 1990, 22, 528531. 59 was sonicated at 35 kHz, 320 W output. [336] Gemal, A. L.; Luche, J. L.: Lanthanoids in organic synthesis. 6. Reduction of -enones by sodium borohydride in the presence of lanthanoid chlorides: synthetic and mechanistic aspects. J. Am. Chem. Soc. 1981,
BIBLIOGRAPHY 103, 54545459. On decomposition, see p. 5456 c. 2, esp. g. 3. URL http://dx.doi.org/10.1021/ja00408a029
271
[337] Marchand, A. P.; LaRoe, W. D.; Sharma, G. V. M.; Suri, S. C.; Reddy, D. S.: Facile stereoselective reductions of enediones and cage diketones using NaBH4 -CeCl3 . J. Org. Chem. 1986, 51, 16221625. See attempted reduction of 12 and 16. URL http://dx.doi.org/10.1021/jo00359a054 [338] Dimitrov, V.; Kostova, K.; Genov, M.: Anhydrous cerium(III) chloride eect of the drying process on activity and eciency. Tetrahedron Lett. 1996, 37, 67876790. URL http://dx.doi.org/10.1016/S0040-4039(96)01479-7 [339] Zsak, J.; Lszl, T.: Complex compounds of some diosphenolic type carotenoids. Astacene complexes. Spectrophotometric study of the system iron(III)-astacene. Rev. Roum. Chim. 1981, 26, 237243. [340] Rubin, Y.; Knobler, C. B.; Diederich, F.: Precursors to the Cyclo[n]carbons: From 3,4-Dialkynyl-3-cyclobutene-1,2-diones and 3,4Dialkynyl-3-cyclobutene-1,2-diols to Cyclobutenodehydroannulenes and Higher Oxides of Carbon. J. Am. Chem. Soc. 1990, 112, 16071617. See Scheme V. URL http://dx.doi.org/10.1021/ja00160a047 [341] McMurry, J. E.: Ester cleavages via SN 2-type dealkylation. Org. React. 1976, 24, 187224. URL http://dx.doi.org/10.1002/0471264180.or024.02 [342] Salomon, C. J.; Mata, E. G.; Mascaretti, O. A.: Recent developments in chemical deprotection of ester functional groups. Tetrahedron 1993, 49, 36913734. URL http://dx.doi.org/10.1016/S0040-4020(01)90225-X [343] Mller, P.; Siegfried, B.: SN 2 reactions with carboxylic esters. Selective cleavage of methyl esters. Helv. Chim. Acta 1974, 57, 987994. URL http://dx.doi.org/10.1002/hlca.19740570404 [344] Elsinger, F.; Schreiber, J.; Eschenmoser, A.: Notiz ber die Selektivitt der Spaltung von Carbonsuremethylestern mit Lithiumjodid. Helv. Chim. Acta 1960, 43, 113118. URL http://dx.doi.org/10.1002/hlca.19600430116 [345] Dean, P. D. G.: Halogenolysis of methyl glycyrrhetate with lithium iodide-dimethylformamide. J. Chem. Soc. 1965, p. 6655. URL http://dx.doi.org/10.1039/JR9650006655 [346] Mller, P.; Siegfried, B.: Decarboxylation of -keto esters in hexamethylphosphoric triamide. Tetrahedron Lett. 1973, pp. 35653568. URL http://dx.doi.org/10.1016/S0040-4039(01)86971-9
272
BIBLIOGRAPHY
[347] Dykstra, R. R.: Hexamethylphosphoric Triamide. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rh020 [348] Elsinger, F.: 2-Benzylcyclopentanone. Org. Synth. 1965, 45, 7. Also Coll. Vol. 5, p. 76. URL http://www.orgsyn.org/orgsyn/pdfs/CV5P0076.pdf [349] McMurry, J. E.; Wong, G. B.: Improved method for the cleavage of methyl esters. Synth. Commun. 1972, 2, 389394. [350] Beck, A. K.; Seebach, D.: N,N -Dimethylpropyleneurea. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rd366 [351] Sharma, L.; Nayak, M. K.; Chakraborti, A. K.: A mild and chemoselective method for ester O-alkyl cleavage using in situ generated potassium thiophenoxide from catalytic quantities of base. Tetrahedron 1999, 55, 95959600. URL http://dx.doi.org/10.1016/S0040-4020(99)00505-0 [352] Misner, J. W.; Kennedy, J. H.; Biggs, W. S.: Integration of a highly selective demethylation of a quaternized ergoline into a one-pot synthesis of pergolide. Org. Process Res. Dev. 1997, 1, 7780. URL http://dx.doi.org/10.1021/op9600015 [353] Bartlett, P. A.; Johnson, W. S.: Improved reagent for the O-alkyl cleavage of methyl esters by nucleophilic displacement. Tetrahedron Lett. 1970, 11, 44594462. URL http://dx.doi.org/10.1016/S0040-4039(01)83950-2 [354] Kelly, T. R.; Dali, H. M.; Tsang, W. G.: Lithium thiomethoxide: a convenient mercaptide reagent. Tetrahedron Lett. 1977, 18, 38593860. URL http://dx.doi.org/10.1016/S0040-4039(01)83373-6 [355] Feutrill, G. I.; Mirrington, R. N.: Reactions with thioethoxide ion in dimethylformamide. II. O-Alkyl cleavage of methylene ethers, methyl esters, and aryloxyacetates. Aust. J. Chem. 1972, 25, 17311735. URL http://dx.doi.org/10.1071/CH9721731 [356] Vaughan, W. R.; Baumann, J. B.: Reactions of Alkyl Carboxylic Esters with Mercaptides. J. Org. Chem. 1962, 27, 739744. URL http://dx.doi.org/10.1021/jo01050a010 [357] Ireland, R. E.; Thompson, W. J.: An approach to the total synthesis of chlorothricolide: the synthesis of the top half. J. Org. Chem. 1979, 44, 30413052. See p. 3048 compound 13 (preparation of LiSPr). URL http://dx.doi.org/10.1021/jo01331a017
BIBLIOGRAPHY
273
[358] Nicolaou, K. C.; Estrada, A. A.; Zak, M.; Lee, S. H.; Sana, B. S.: A Mild and Selective Method for the Hydrolysis of Esters with Trimethyltin Hydroxide. Angew. Chem. Int. Ed. 2005, 44, 13781382. URL http://dx.doi.org/10.1002/anie.200462207 [359] Ishida, N.; Akaike, M.; Tsutsumi, S.; Kanai, H.; Masui, A.; Sadamatsu, M.; Kuroda, Y.; Watanabe, Y.; McEwen, B. S.; Kato, N.: Trimethyltin syndrome as a hippocampal degeneration model: temporal changes and neurochemical features of seizure susceptibility and learning impairment. Neuroscience 1997, 81, 11831191. URL http://dx.doi.org/10.1016/S0306-4522(97)00220-0 [360] Mascaretti, O. A.; Furlan, R. L. E.: Esterications, transesterications, and deesterications mediated by organotin oxides, hydroxides, and alkoxides. Aldrichimica Acta 1997, 30, 5568. [361] Lee, D. Y.; He, M.; Kondaveti, L.; Liu-Chen, L. Y.; Ma, Z.; Wang, Y.; Chen, Y.; Li, J. G.; Bguin, C.; Carlezon, W. A.; Cohen, B.: Synthesis and in vitro pharmacological studies of C(4) modied salvinorin A analogues. Bioorg. Med. Chem. Lett. 2005, 15, 41694173. URL http://dx.doi.org/10.1016/j.bmcl.2005.06.092 [362] Oikawa, M.; Wada, A.; Okazaki, F.; Kusumoto, S.: Acidic, Selective Monoacylation of vic-Diols. J. Org. Chem. 1996, 61, 44694471. See compound 38. URL http://dx.doi.org/10.1021/jo960081a [363] Burchat, A. F.; Chong, J. M.; Nielsen, N.: Titration of alkyllithiums with a simple reagent to a blue endpoint. J. Organomet. Chem. 1997, 542, 281283. URL http://dx.doi.org/10.1016/S0022-328X(97)00143-5 [364] Yoon, N. M.; Pak, C. S.; Brown, H. C.; Krishnamurthy, S.; Stocky, T. P.: Selective reductions. XIX. Rapid reaction of carboxylic acids with borane-tetrahydrofuran. Remarkably convenient procedure for the selective conversion of carboxylic acids to the corresponding alcohols in the presence of other functional groups. J. Org. Chem. 1973, 38, 27862792. URL http://dx.doi.org/10.1021/jo00956a011 [365] Lane, C. F.: Reduction of organic compounds with diborane. Chem. Rev. 1976, 76, 773 799. URL http://dx.doi.org/10.1021/cr60304a005 [366] Corey, E. J.; Sachdev, H. S.: 2,4-Dinitrobenzenesulfonylhydrazine, a useful reagent for the Eschenmoser , cleavage of ,-epoxy ketones. Conformational control of halolactonization. J. Org. Chem. 1975, 40, 579 581. See compound IV. URL http://dx.doi.org/10.1021/jo00893a008
274
BIBLIOGRAPHY
[367] McCorkindale, N. J.; Roy, T. P.; Hutchinson, S. A.: Isolation and synthesis of 3-chlorogentisyl alcohola metabolite of Penicillium canadense. Tetrahedron 1972, 28, 11071111. See compound 5. URL http://dx.doi.org/10.1016/0040-4020(72)80170-4 [368] Hagishita, S.; Kuriyama, K.: Optical activity of C2 symmetrical 9,10dihydro-9,10-ethanoanthracenes. Tetrahedron 1972, 28, 14351467. See compound 5. BH3 was generated in situ as per reference 5 therein. Stoichiometry inferred from limiting reagent NaBH4 . URL http://dx.doi.org/10.1016/0040-4020(72)88028-1 [369] Brown, H. C.; Heim, P.; Yoon, N. M.: Selective reductions. XV. Reaction of diborane in tetrahydrofuran with selected organic compounds containing representative functional groups. J. Am. Chem. Soc. 1970, 92, 1637 1646. URL http://dx.doi.org/10.1021/ja00709a037 [370] Doxsee, K. M.; Feigel, M.; Stewart, K. D.; Canary, J. W.; Knobler, C. B.; Cram, D. J.: Host-guest complexation. 42. Preorganization strongly enhances the tendancy of hemispherands to form hemispheraplexes. J. Am. Chem. Soc. 1987, 109, 3098 3107. See compound 24. URL http://dx.doi.org/10.1021/ja00244a037 [371] Artz, S. P.; Cram, D. J.: Host-guest complexation. 28. Hemispherands with four self-organizing units. J. Am. Chem. Soc. 1984, pp. 21602171. See compounds 32 and 37. URL http://dx.doi.org/10.1021/ja00319a042 [372] Dias, J. R.; Pettit, G. R.: Reduction of -lactones and hindered esters with diborane. J. Org. Chem. 1971, 36, 34853489. See compounds 7 and 9. URL http://dx.doi.org/10.1021/jo00822a003 [373] Kende, A. S.; Fludzinski, P.: Ethyl 4-hydroxycrotonate. Org. Synth. 1986, 64, 104. Also Coll. Vol. 7, p. 221. URL http://www.orgsyn.org/orgsyn/pdfs/CV7P0221.pdf [374] Brown, H. C.; Stocky, T. P.: Selective reductions. 24. Acyloxyboranes in the controlled reaction of carboxylic acids with borane-tetrahydrofuran. Acyloxyboranes as intermediates in the fast reduction of carboxylic acids by borane-tetrahydrofuran. J. Am. Chem. Soc. 1977, 99, 8218 8226. URL http://dx.doi.org/10.1021/ja00467a016 [375] Lee, K. C.: Methyltriphenylphosphonium Bromide. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rm273
BIBLIOGRAPHY
275
[376] Pine, S. H.; Shen, G. S.; Hoang, H.: Ketone methylenation using the Tebbe and Wittig reagents - a comparison. Synthesis 1991, pp. 165 167. URL http://dx.doi.org/10.1055/s-1991-26406 [377] Pine, S. H.: Carbonyl methylenation and alkylidenation using titaniumbased reagents. Org. React. 1993, 43, 191. URL http://dx.doi.org/10.1002/0471264180.or043.01 [378] Lombardo, L.: Methylenation of carbonyl compounds with zincdibromomethane-titanium tetrachloride. Applications to gibberellins. Tetrahedron Lett. 1982, 23, 42934296. Note: due to the thick consistency of this reagent, transfer was very dicult using 0.5 mm I.D. (21 gauge) needles, and a larger gauge would be advisable. URL http://dx.doi.org/10.1016/S0040-4039(00)88728-6 [379] Lombardo, L.: Methylenation of carbonyl compounds: (+)-3-methylenecis-p-menthane. Org. Synth. 1987, 65, 81. Also Coll. Vol. 8, p. 386. URL http://www.orgsyn.org/orgsyn/pdfs/CV8P0386.pdf [380] Vedejs, E.: Clemmensen reduction of ketones in anhydrous organic solvents. Org. React. 1975, 22, 401422. See table p. 418 (eliminations). URL http://dx.doi.org/10.1002/0471264180.or022.03 [381] Yamamura, S.; Toda, M.; Hirata, Y.: Modied Clemmensen Reduction: Cholestane. Org. Synth. 1973, 53, 289. See discussion re acetoxycholestanone. URL http://www.orgsyn.org/orgsyn/pdfs/CV6P0289.pdf [382] Chamberlin, A. R.; Sheppeck, II, J. E.: p-Toluenesulfonylhydrazide. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rt137 [383] Caglioti, L.: Reduction of ketones by use of the tosylhydrazone derivatives: androstan-17-ol. Org. Synth. 1972, 52, 122. URL http://www.orgsyn.org/orgsyn/pdfs/CV6P0062.pdf [384] Friedman, L.; Litle, R. L.; Reichle, W. R.: p-Toluenesulfonylhydrazide. Org. Synth. 1960, 40, 93. Also Coll. Vol. 5, p. 1055. See note 6 re recrystallisation. URL http://www.orgsyn.org/orgsyn/pdfs/CV5P1055.pdf [385] Bandgar, B. P.; Sadavarte, V. S.; Uppalla, L. S.; Govande, R.: Chemoselective preparation of oximes, semicarbazones, and tosylhydrazones without catalyst and solvent. Monatsh. Chem. 2001, 132, 403406. URL http://dx.doi.org/10.1007/s007060170126
276
BIBLIOGRAPHY
[386] Jeelnik, M.; Varma, R. S.; Polanc, S.; Koevar, M.: Solid-state synthesis of heterocyclic hydrazones using microwaves under catalyst-free conditions. Green Chem. 2002, 4, 3538. URL http://dx.doi.org/10.1039/b108029f [387] Hajipour, A. R.; Mallakpour, S. E.; Imanzadeh, G.: A Rapid and Convenient Synthesis of Oximes in Dry Media under Microwave Irradiation. J. Chem. Res. 1999, pp. 228229. URL http://dx.doi.org/10.1039/a806359a [388] Bertz, S. H.; Dabbag, G.: Improved preparations of some arenesulfonylhydrazones. J. Org. Chem. 1983, 48, 116119. URL http://dx.doi.org/10.1021/jo00149a022 [389] Iida, T.; Tamura, T.; Matsumoto, T.; Chang, F.: Improved Conditions for Preparation and Reductive Cleavage of Steroidal Ketone Tosylhydrazones. Synthesis 1984, pp. 957959. URL http://dx.doi.org/10.1055/s-1984-31036 [390] Jarikote, D. V.; Deshmukh, R. R.; Rajagopal, R.; Lahoti, R. J.; Daniel, T.; Srinivasan, K. V.: Ultrasound promoted facile synthesis of arylhydrazones at ambient conditions. Ultrason. Sonochem. 2003, 10, 4548. URL http://dx.doi.org/10.1016/S1350-4177(02)00100-1 [391] Vinczer, P.; Novak, L.; Szantay, C.: A mild method for the formation of tosylhydrazone derivatives from keto esters. Synth. Commun. 1984, 14, 281288. [392] Hertz, H. S.; Coxon, B.; Siedle, A. R.: Disproportionation and pyrolysis of p-toluenesulfonylhydrazine. J. Org. Chem. 1977, 42, 25082509. URL http://dx.doi.org/10.1021/jo00434a036 [393] Pouchert, C. J.; Behnke, J.: The Aldrich library of 13 C and 1 H FT-NMR spectra, vol. 2. Aldrich Chemical Co., Milwaukee 1993. See p. 1634A. [394] El-Sayed, R. A.: Review on the chemistry of sulfonohydrazides and sulfonoazides. Phosphorus, Sulfur Silicon Relat. Elem. 2004, 179, 237266. Compare references 1-138 (pre-1972) with 139-158 (self citations). URL http://dx.doi.org/10.1080/10426500490274673 [395] Sutherland, A. G.: One or More CH Bond(s) Formed by Substitution: Reduction of C-Halogen and C-Chalcogen Bonds. In Comprehensive Organic Functional Group Transformations, A. R. Katritzky; O. MethCohn; C. W. Rees, eds., Pergamon, Oxford, vol. 1, pp. 126, 1st ed. 1995. [396] Barton, D. H. R.; Ferreira, J. A.; Jaszberenyi, J. C.: Free radical deoxygenation of thiocarbonyl derivatives of alcohols. In Preparative Carbohydrate Chemistry, S. Hanessian, ed., Marcel Dekker, New York, pp. 151172 1997.
BIBLIOGRAPHY
277
[397] Salomon, C. J.; Danelon, G. O.; Mascaretti, O. A.: A Practical Method for the Disposal of Organotin Residues from Reaction Mixtures. J. Org. Chem. 2000, 65, 92209222. See also references therein. URL http://dx.doi.org/10.1021/jo000525+ [398] Barton, D. H. R.; Jang, D. O.; Jaszberenyi, J. C.: The invention of radical reactions. 32. Radical deoxygenations, dehalogenations, and deaminations with dialkyl phosphites and hypophosphorous acid as hydrogen sources. J. Org. Chem. 1993, 58, 68386842. URL http://dx.doi.org/10.1021/jo00076a054 [399] Takamatsu, S.; Katayama, S.; Hirose, N.; Naito, M.; Izawa, K.: Radical deoxygenation and dehalogenation of nucleoside derivatives with hypophosphorous acid and dialkyl phosphites. Tetrahedron Lett. 2001, 42, 76057608. URL http://dx.doi.org/10.1016/S0040-4039(01)01617-3 [400] Rho, H.-S.; Ko, B.-S.: Regioselective deoxygenation of the cyclic thionocarbonates of 2,3-dihydroxy esters with magnesium in methanol. Synth. Commun. 1999, 29, 28752880. [401] Brown, A. C.; Carpino, L. A.: Magnesium in methanol: substitute for sodium amalgam in desulfonylation reactions. J. Org. Chem. 1985, 50, 17491750. URL http://dx.doi.org/10.1021/jo00210a035 [402] Fukukawa, K.; Ueda, T.; Hirano, T.: Nucleosides and nucleotides. VL. Facile deoxygenation of neplanocin A and nucleosides by the use of tributyltin hydride. Chem. Pharm. Bull. 1983, 31, 18421847. [403] Patroni, J. J.; Stick, R. V.; Engelhardt, L. M.; White, A. H.: The deoxygenation of some carbohydrate diols via the derived cyclic thiocarbonate. Aust. J. Chem. 1986, 39, 699711. URL http://dx.doi.org/10.1071/CH9860699 [404] Rho, H.-S.: Tributyltin hydride-induced free radical deoxygenation of the cyclic thionocarbonates of threo-2,3-dihydroxy esters and ketones. Synth. Commun. 1997, 27, 38873893. [405] Kanemitsu, K.; Tsuda, Y.; Haque, M. E.; Tsubono, K.; Kikuchi, T.: Reaction of cyclic thioxocarbonates with tributyltin hydride. Chem. Pharm. Bull. 1987, 35, 38743879. [406] Tsuda, Y.; Kanemitsu, K.; Kakimoto, K.; Kikuchi, T.: Utilization of sugars in organic synthesis. XIX. Trialkyltin radical used to catalyze the O,S -rearrangement of cyclic thionocarbonates. A new entry to thio sugars. Chem. Pharm. Bull. 1987, 35, 21482150.
278
BIBLIOGRAPHY
[407] Patroni, J. J.; Stick, R. V.; Matthew, D.; Tilbrook, G.; Skelton, B. W.; White, A. H.: The synthesis and reactivity of cyclic thiocarbonates derived from some carbohydrate 1,2-diols. Aust. J. Chem. 1989, 42, 2127 2141. URL http://dx.doi.org/10.1071/CH9892127 [408] Szammer, J.; tvs, L.: A convenient preparation of tri-n-butyltin hydride. Chem. Ind. 1988, p. 764. [409] McAlonan, H.; Stevenson, P. J.: Convenient One-Pot Synthesis of Hexan-butylditin from Bis(tri-n-butyltin) Oxide. Organometallics 1995, 14, 40214022. URL http://dx.doi.org/10.1021/om00008a058 [410] Giese, B.; Grninger, K. S.: 1,3,4,6-Tetra-O-acetyl-2-deoxy--Dglucopyranose. Org. Synth. 1990, 69, 6671. Also Coll. Vol. 8, p. 583. URL http://www.orgsyn.org/orgsyn/pdfs/CV8P0583.pdf [411] Berge, J. M.; Roberts, S. M.: Recommended work-up procedure for reductions employing tri-n-butyltin hydride. Synthesis 1979, pp. 471 472. URL http://dx.doi.org/10.1055/s-1979-28726 [412] Farina, V.: A simple chromatographic technique for the purication of organic stannanes. J. Org. Chem. 1991, 56, 49854987. URL http://dx.doi.org/10.1021/jo00016a037 [413] Crich, D.; Sun, S.: A Practical Method for the Removal of Organotin Residues from Reaction Mixtures. J. Org. Chem. 1996, 61, 72007201. URL http://dx.doi.org/10.1021/jo960751c [414] Rajan Babu, T. B. V.; Bulman Page, P. C.; Buckley, B. R.: Tri-nbutylstannane. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rt181 [415] Sano, H.: Bis(tri-n-butyltin) Oxide. In Encyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York 1995. URL http://dx.doi.org/10.1002/047084289X.rb195 [416] Jang, D. O.; Cho, D. H.; Barton, D. H. R.: Radical deoxygenation of alcohols via their S -methyl dithiocarbonate derivatives with di-nbutylphosphine oxide as hydrogen-atom donor. Synlett 1998, pp. 3940. URL http://dx.doi.org/10.1055/s-1998-1582 [417] Jang, D. O.; Cho, D. H.: Radical deoxygenation of alcohols and vicinal diols with N -ethylpiperidine hypophosphite in water. Tetrahedron Lett. 2002, 43, 59215924. URL http://dx.doi.org/10.1016/S0040-4039(02)01247-9
BIBLIOGRAPHY
279
[418] Jang, D. O.; Song, S. H.: Facile synthesis of enantiopure (R)-malates. Tetrahedron Lett. 2000, 41, 247249. URL http://dx.doi.org/10.1016/S0040-4039(99)02057-2 [419] Cho, D. H.; Song, S. H.; Jang, D. O.: A method for preparation of unnatural (R)-malic acid derivatives with phenylsilanes. Synth. Commun. 2003, 33, 515519. URL http://dx.doi.org/10.1081/SCC-120015803 [420] Klein Gebbinck, E. A.; Jansen, B. J. M.; de Groot, A.: Insect antifeedant activity of clerodane diterpenes and related model compounds. Phytochemistry 2002, 61, 737770. See table 6 (H. armigera). URL http://dx.doi.org/10.1016/S0031-9422(02)00174-7 [421] Enriz, R. D.; Baldoni, H. A.; Zamora, M. A.; Juregui, E. A.; Sosa, M. E.; Tonn, C. E.; Luco, J. M.; Gordaliza, M.: Structure-Antifeedant Activity Relationship of Clerodane Diterpenoids. Comparative Study with Withanolides and Azadirachtin. J. Agric. Food Chem. 2000, 48, 13841392. URL http://dx.doi.org/10.1021/jf990006b [422] Simmonds, M. S. J.; Blaney, W. M.; Esquivel, B.; Rodriguez-Hahn, L.: Eect of clerodane-type diterpenoids isolated from Salvia on the feeding behavior of Spodoptera littoralis. Pestic. Sci. 1996, 47, 1723. URL http://dx.doi.org/10.1002/(SICI)1096-9063(199605)47:1<17:: AID-PS378>3.0.CO;2-I [423] Simmonds, M. S. J.; Blaney, W. M.; Fellows, L. E.: Behavioral and electrophysiological study of antifeedant mechanisms associated with polyhydroxy alkaloids. J. Chem. Ecol. 1990, 16, 31673196. See p. 3171 (choice assay). URL http://dx.doi.org/10.1007/BF00979618 [424] Novac, O.; Guenier, A. S.; Pelletier, J.: Inhibitors of protein synthesis identied by a high throughput multiplexed translation screen. Nucleic Acids Res. 2004, 32, 902915. See p. 904 under dual-luciferase reporter assay.. URL http://dx.doi.org/10.1093/nar/gkh235 [425] Reference Method for Broth Dilution Antifungal Susceptibility Testing of Yeasts; Approved Standard M27-A2. National Committee for Clinical Laboratory Standards, Wayne, PA, 2nd ed. 2002. [426] Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria that Grow Aerobically; Approved Standard M7-A6. National Committee for Clinical Laboratory Standards, Wayne, PA, 6th ed. 2003. [427] Performance Standards for Antimicrobial Disk Susceptibility Tests; Approved Standard M2-A8. National Committee for Clinical Laboratory Standards, Wayne, PA, 8th ed. 2003.
280 [428] Photos by Andrea Bigham, University of Melbourne.
BIBLIOGRAPHY
[429] Rovinsky, S. A.; Cizadlo, G. R.: Salvia divinorum Epling et Jtiva-M. (Labiatae): An Ethnopharmacological Investigation. The McNair Scholarly Review 1998, 3, 142156. College of St Scholastica (Duluth, MN). Unocial transcription available online. Accessed March 2006. URL http://www.sagewisdom.org/rovinsky.html [430] Peterson, P. K.; Gekker, G.; Lokensgard, J. R.; Bidlack, J. M.; Chang, A. C.; Fang, X. G.; Portoghese, P. S.: -opioid receptor agonist suppression of HIV-1 expression in CD4+ lymphocytes. Biochem. Pharmacol. 2001, 61, 11451151. See also references therein (other cell types). URL http://dx.doi.org/10.1016/S0006-2952(01)00574-3 [431] Lokensgard, J. R.; Gekker, G.; Peterson, P. K.: -opioid receptor agonist inhibition of HIV-1 envelope glycoprotein-mediated membrane fusion and CXCR4 expression on CD4+ lymphocytes. Biochem. Pharmacol. 2002, 63, 10371041. URL http://dx.doi.org/10.1016/S0006-2952(02)00875-4 [432] Reeves, J. D.; Piefer, A. J.: Emerging drug targets for antiretroviral therapy. Drugs 2005, 65, 17471766. URL http://www.ingentaconnect.com/content/adis/dgs/2005/ 00000065/00000013/art00002 [433] Gekker, G.; Lokensgard, J. R.; Peterson, P. K.: Naltrexone potentiates anti-HIV-1 activity of antiretroviral drugs in CD4+ lymphocyte cultures. Drug Alcohol Depend. 2001, 64, 257263. URL http://dx.doi.org/10.1016/S0376-8716(01)00140-5 [434] Sheng, W. S.; Hu, S. X.; Lokensgard, J. R.; Peterson, P. K.: U50,488 inhibits HIV-1 Tat-induced monocyte chemoattractant protein-1 (CCL2) production by human astrocytes. Biochem. Pharmacol. 2003, 65, 914. URL http://dx.doi.org/10.1016/S0006-2952(02)01480-6 [435] Chao, C. C.; Hu, S.; Gekker, G.; Lokensgard, J. R.; Heyes, M. P.; Peterson, P. K.: U50,488 protection against HIV-1-related neurotoxicity: involvement of quinolinic acid suppression. Neuropharmacology 2000, 39, 150160. URL http://dx.doi.org/10.1016/S0028-3908(99)00063-5 [436] OBrien, M. E.; Clark, R. A.; Besch, C. L.; Myers, L.; Kissinger, P.: Patterns and correlates of discontinuation of the initial HAART regimen in an urban outpatient cohort. J. Acquir. Immune Dec. Syndr. 2003, 34, 407414. URL http://www.jaids.org/pt/re/jaids/abstract. 00126334-200312010-00008.htm
BIBLIOGRAPHY
281
[437] Blancquaert, J. P.; Lefebvre, R. A.; Willems, J. L.: Emetic and antiemetic eects of opioids in the dog. Eur. J. Pharmacol. 1986, 128, 143150. URL http://dx.doi.org/10.1016/0014-2999(86)90760-0 [438] Kohl, R. L.; MacDonald, S.: New pharmacologic approaches to the prevention of space/motion sickness. J. Clin. Pharmacol. 1991, 31, 934946. URL http://jcp.sagepub.com/cgi/content/abstract/31/10/934 [439] Lewin, S. R.; Vesanen, M.; Kostrikis, L.; Hurley, A.; Duran, M.; Zhang, L.; Ho, D. D.; Markowitz, M.: Use of real-time PCR and molecular beacons to detect virus replication in human immunodeciency virus type 1-infected individuals on prolonged eective antiretroviral therapy. J. Virol. 1999, 73, 60996103. URL http://jvi.asm.org/cgi/reprint/73/7/6099 [440] Solomon, A.; Lane, N.; Wightman, F.; Gorry, P. R.; Lewin, S. R.: Enhanced replicative capacity and pathogenicity of HIV-1 isolated from individuals infected with drug-resistant virus and declining CD4+ T-cell counts. J. Acquir. Immune Dec. Syndr. 2005, 40, 140148. URL http://www.jaids.com/pt/re/jaids/abstract. 00126334-200510010-00004.htm [441] Project Report A694-38: Anti-HIV Activity and Cytotoxicity of One NIAID Compound (9966) Tested in PBMC Cell-Based Assays. Southern Research Institute, Frederick, Maryland 2004. Available from the author. [442] Chao, C. C.; Gekker, G.; Hu, S.; Sheng, W. S.; Shark, K. B.; Bu, D. F.; Archer, S.; Bidlack, J. M.; Peterson, P. K.: opioid receptors in human microglia downregulate human immunodeciency virus 1 expression. Proc. Natl. Acad. Sci. U. S. A. 1996, 93, 80518056. URL http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve& db=PubMed&dopt=Citation&list_uids=8755601 [443] Chao, C. C.; Gekker, G.; Hu, S. X.; Kravitz, F.; Peterson, P. K.: -opioid potentiation of tumor necrosis factor--induced anti-HIV-1 activity in acutely infected human brain cell cultures. Biochem. Pharmacol. 1998, 56, 397404. URL http://dx.doi.org/10.1016/S0006-2952(98)00161-0 [444] Chou, J. Z.; Chait, B. T.; Wang, R.; Kreek, M. J.: Dierential biotransformation of dynorphin A (1-17) and dynorphin A (1-13) peptides in human blood, ex vivo. Peptides 1996, 17, 983990. URL http://dx.doi.org/10.1016/0196-9781(96)00154-4 [445] Childers, S. R.; Xiao, R.; Vogt, L.; Sim, L. J.: Lack of evidence of 2 selective activation of G-proteins: opioid receptor stimulation of [35 S] GTPS binding in guinea pig brain. Biochem. Pharmacol. 1998, 56,
282
BIBLIOGRAPHY 113120. URL http://dx.doi.org/10.1016/S0006-2952(98)00123-3
[446] Bosworth, D. T.: Salvia, AIDS, and the healing journey 2005. URL http://www.sagewisdom.org/aidshealing.html [447] Holbeck, S. L.: Update on NCI in vitro drug screen utilities. Eur. J. Cancer 2004, 40, 785793. URL http://dx.doi.org/10.1016/j.ejca.2003.11.022 [448] Johnson, J. I.; Decker, S.; Zaharevitz, D.; Rubinstein, L. V.; Venditti, J. M.; Schepartz, S.; Kalyandrug, S.; Christian, M.; Arbuck, S.; Hollingshead, M.; Sausville, E. A.: Relationships between drug activity in NCI preclinical in vitro and in vivo models and early clinical trials. Br. J. Cancer 2001, 84, 14241431. URL http://dx.doi.org/10.1054/bjoc.2001.1796 [449] US National Cancer Institute, Developmental Therapeutics Program Data Search. Salvinorin A = NSC 604585. URL http://dtp.nci.nih.gov/docs/dtp_search.html [450] In Vivo Cancer Models 1976-1982. National Institutes of Health, Washington, DC 1984. NIH Publication # 84-2635. See p. 27 (P388 protocol). URL http://dtp.nci.nih.gov/branches/btb/pdf/In_Vivo_Cancer_ Models.pdf [451] Schultz, R. J. Drug Synthesis and Chemistry Branch, National Cancer Institute. Email, 5th March 2006. [452] Roth, B. L. R. Email, 23rd August 2005. [453] Yan, F.; Mosier, P. D.; Westkaemper, R. B.; Stewart, J.; Zjawiony, J. K.; Vortherms, T. A.; Sheer, D. J.; Roth, B. L.: Identication of the Molecular Mechanisms by Which the Diterpenoid Salvinorin A Binds to -Opioid Receptors. Biochemistry 2005, 44, 86438651. URL http://dx.doi.org/10.1021/bi050490d [454] Rovner, S. L.: Chemical Naming Method Unveiled. Chem. Eng. News 2005, 83, 3940. (accessed Jan. 2006). URL http://pubs.acs.org/isubscribe/journals/cen/83/i34/html/ 8334sci1.html [455] IUPAC International Chemical Identier (accessed Oct. 2005). URL http://www.iupac.org/inchi [456] Hoye, T. R.; Hanson, P. R.; Vyvyan, J. R.: A Practical Guide to FirstOrder Multiplet Analysis in 1 H NMR Spectroscopy. J. Org. Chem. 1994, 59, 40964103. URL http://dx.doi.org/10.1021/jo00094a018
BIBLIOGRAPHY
283
[457] Hoye, T. R.; Zhao, H.: A Method for Easily Determining Coupling Constant Values: An Addendum to A Practical Guide to First-Order Multiplet Analysis in 1 H NMR Spectroscopy. J. Org. Chem. 2002, 67, 4014 4016. URL http://dx.doi.org/10.1021/jo001139v [458] Gottlieb, H. E.; Kotlyar, V.; Nudelman, A.: NMR Chemical Shifts of Common Laboratory Solvents as Trace Impurities. J. Org. Chem. 1997, 62, 75127515. URL http://dx.doi.org/10.1021/jo971176v [459] Glennon, R. A.; Lee, M.; Rangisetty, J. B.; Dukat, M.; Roth, B. L.; Savage, J. E.; McBride, A.; Rauser, L.; Hufeisen, S.; Lee, D. K.: 2Substituted tryptamines: agents with selectivity for 5-HT6 serotonin receptors. J. Med. Chem. 2000, 43, 10111018. URL http://dx.doi.org/10.1021/jm990550b [460] Pauwels, P. J.; Colpaert, F. C.: Heterogeneous ligand-mediated Ca++ responses at wt and mutant 2A -adrenoceptors suggest multiple ligand activation binding sites at the 2A -adrenoceptor. Neuropharmacology 2000, 39, 21012111. URL http://dx.doi.org/10.1016/S0028-3908(00)00040-X

Thomas Anthony Munro - The Chemistry of Salvia Divinorum

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thomas Anthony Munro - The Chemistry of Salvia Divinorum

Uploaded by

Copyright:

Available Formats

The Chemistry of Salvia divinorum

Submitted in total fullment of the requirements of the degree of Doctor of Philosophy

Department of Chemistry The University of Melbourne

OR3 R1 28a 28b 28c R2 R3

Thomas Anthony Munro

12 1.4.1 1.4.2 1.4.3 1.4.4

CONTENTS Animal Testing. . . . . . . . . . . . . . . . . . . . . . . . 47 Human Testing. . . . . . . . . . . . . . . . . . . . . . . . 51 In vitro Testing. . . . . . . . . . . . . . . . . . . . . . . . 53 Mechanism: Opioids. . . . . . . . . . . . . . . . . . . . 54

CONTENTS 3 Synthesis. 3.1 3.2

Mechanism. . . . . . . . . . . . . . . . . . . . . . . . . . 126 Attempted Deacetylation under Acidic Conditions. . . . 128

5.2 Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 5.2.1 Extraction of Commercial S. divinorum. . . . . . . . . . 185

CONTENTS 5.2.2 5.2.3 5.2.4 5.2.5 5.2.6 5.2.7 5.2.8 5.2.9

15 Extraction of Australian S. divinorum. . . . . . . . . . . 187 Salvinorin A (1a). . . . . . . . . . . . . . . . . . . . . . 188

Salvinorin B (1b). . . . . . . . . . . . . . . . . . . . . . . 189 Salvinorin C (1c). . . . . . . . . . . . . . . . . . . . . . 189

5.3.10 8-epi-Salvinorin A (37a). . . . . . . . . . . . . . . . . . . 208

5.3.24 1-Deoxysalvinorin A (81a). . . . . . . . . . . . . . . . . . 233 Bibliography 239

1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

[distances of analyte (a) and solvent front (f ) from

LIST OF TABLES TLC Thin layer chromatography

Figure 1.1: Flowering specimen of S. divinorum.1

100 200 300 km

Figure 1.9: Biosynthetic precursors of diterpenoids.

Scheme 1.1: Cyclization of 7 to the trans-neoclerodane skeleton(stereoview).

Scheme 1.2: Cyclization of 7 to the cis-neoclerodane skeleton.

Figure 1.10: Peltate glandular trichome (SEM stereoview).39

Above: immature leaf (1 mm wide); below: mature leaf (10 cm wide).

Figure 1.11: Underside of S. divinorum leaves (SEM stereoview).39

Cats and Rats.

Rhesus Monkeys (Postscript).

Structure, Potency and Selectivity.

1.6. SOCIAL IMPACT.

time 4d 100 min

2.1. ISOLATION PROCEDURE.

Problems Caused by Pigments.

Developed in 50% EtOAc/petrol, visualised in vanillin/H2 SO4 .

2.1. ISOLATION PROCEDURE.

Chemistry of Plant Pigments.

Use of Activated Carbon.

2.1. ISOLATION PROCEDURE.

Use during Recrystallisation.

2.1. ISOLATION PROCEDURE. 2.1.3.3 Use in Chromatography.

2.1. ISOLATION PROCEDURE.

Application to S. divinorum Extract.

2.1. ISOLATION PROCEDURE.

General Applicability of the Method.

2.1. ISOLATION PROCEDURE.

Other Decolourisation Adsorbents.

Figure 2.6: SEM image of salvinorin A crystals (blade morphology).39

2.1. ISOLATION PROCEDURE.

Figure 2.7: Other salvinorin A crystal morphologies (stereoview).39

2.1. ISOLATION PROCEDURE. 2.1.4.2 Chromatography.

OR3 R1 28a 28b 28c 29a OH OH H H R2 H OH OAc H R3 H Me H H

20 28a 10 10 20 30 40 50 60 70 80 hRf (70% Et2O/petrol)

2.1. ISOLATION PROCEDURE.

Column on silica 5-50% acetone/CH2Cl2

Series B 150 mg Column 70-90% Et2O/petrol

Column 50-80% Et2O/petrol 1c (219 mg) 1a (2.9 g)

Triturate (Et2O) 1d (75 mg)

28a (36 mg)

Combine Column 60-100% Et2O/petrol

HPLC 60% EtOAc/petrol 1f (1 mg)

Figure 2.10: Isolation of terpenoids from commercial S. divinorum.