The Chemistry of Salvia divinorum
Thomas Anthony Munro
Submitted in total fulfilment of the requirements
of the degree of Doctor of Philosophy
April 2006
Department of Chemistry
The University of Melbourne
2
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 κ opioid receptor (KOR),
and is the first non-nitrogenous opioid. The plant is used recreationally as
a hallucinogen, but is unpopular due to its dysphoric effects. 1a has been
prohibited in Australia under an invalid systematic name.
An early report of psychoactive alkaloids in S. divinorum proved to be irre-
producible. Similarly, tests in mice suggesting the presence of psychoactive
compounds other than 1a were confounded and therefore unreliable.
O
O
O
H HOR1
R2
O O
R2
O
H HR1
O OR3
1a
R1
AcHH
R2
OHOAcH
1d1e1f
28a28b28c
R1
OH OH H
R2
HOHOAc
R3
HMeH
22 88
O
O
O
H HOO
O
O O
In this work, an improved isolation method for 1a was developed, using fil-
tration through activated carbon to decolourise the crude extract. Six new
diterpenoids were isolated: salvinorins D–F (1d–1f) and divinatorins A–C
(28a–28c). Five known terpenoids not previously reported from this species
were also isolated.
3
4
The structure–activity relationships of 1a were evaluated via selective mod-
ifications of each functional group. Useful synthetic methods are reviewed,
including the first 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 functionali-
ties are not. Other compounds from S. divinorum did not bind to the KOR,
suggesting that 1a is the plant’s active principle.
The structure of the 8-epimer of 1a, reported previously without supporting
evidence, was firmly established. This epimerisation proved to be a general
phenomenon among salvinorins and related furanolactones, occurring via eno-
lisation 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 in-
sects, bacteria, fungi, HIV, tumour cell lines and protein synthesis.
59
O
O
O
HOH
O
O O
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
5
6
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 (Fred-
erick, 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. Na-
tional 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 D–F, New Neoclerodane
Diterpenoids from Salvia divinorum, and an Improved Method for the
Isolation of Salvinorin A. J. Nat. Prod. 2003, 66, 703–705. http://dx.
doi.org/10.1021/np0205699
• 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, 1242–1244. 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 κ Opioid
Receptor Agonist. J. Med. Chem. 2005, 48, 345–348. http://dx.doi.
org/10.1021/jm049438q
• Munro, T. A.; Goetchius, G. W.; Roth, B. L.; Vortherms, T. A.; Rizza-
casa, M. A. Autoxidation of Salvinorin A under Basic Conditions. J. Org.
Chem. 2005, 70, 10,057–10,061. 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. I’m glad it paid
off. Thanks to the man with the coolest name in chemistry, Leander Jerome
Julian Valdés III, for generously sharing ideas and advice. Daniel Siebert for
helpful advice on lots of stuff, 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 world’s 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 Hoye’s
NMR papers. The man with the second-coolest name in chemistry, Carlos
Rodríguez, for translating Díaz. Vic Iwanov for letting me use the safe, and
filling 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 dad’s better than yours. Thanks mum
9
10
for the sacrifices 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 five days before leaving for my postdoc!
Contents
Abstract 3
Declaration 5
Preface 7
Acknowledgments 9
List of Figures 17
List of Schemes 23
List of Tables 25
Acronyms 27
1 Introduction. 31
1.1 Botany. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
1.2 Ethnopharmacology. . . . . . . . . . . . . . . . . . . . . . . . . 33
1.3 Chemistry. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.3.1 Terpenoids. . . . . . . . . . . . . . . . . . . . . . . . . . 36
1.3.2 Alleged alkaloids. . . . . . . . . . . . . . . . . . . . . . . 43
1.4 Pharmacology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
11
12 CONTENTS
1.4.1 Animal Testing. . . . . . . . . . . . . . . . . . . . . . . . 47
1.4.2 Human Testing. . . . . . . . . . . . . . . . . . . . . . . . 51
1.4.3 In vitro Testing. . . . . . . . . . . . . . . . . . . . . . . . 53
1.4.4 Mechanism: κ Opioids. . . . . . . . . . . . . . . . . . . . 54
1.5 Toxicology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
1.6 Social impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.6.1 Recreational Use. . . . . . . . . . . . . . . . . . . . . . . 62
1.6.2 Legal Status. . . . . . . . . . . . . . . . . . . . . . . . . 63
1.7 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
2 Isolation. 67
2.1 Isolation Procedure. . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.1.1 Extraction Conditions. . . . . . . . . . . . . . . . . . . . 67
2.1.2 Problems Caused by Pigments. . . . . . . . . . . . . . . 69
2.1.3 Use of Activated Carbon. . . . . . . . . . . . . . . . . . . 72
2.1.4 Separation of Terpenoids. . . . . . . . . . . . . . . . . . 82
2.2 Structure Elucidation. . . . . . . . . . . . . . . . . . . . . . . . 90
2.2.1 Revised NMR Assignments for Salvinorin A (1a). . . . . 90
2.2.2 Revised NMR Assignments for Salvinorin C (1c). . . . . 92
2.2.3 Other Known Diterpenoids. . . . . . . . . . . . . . . . . 92
2.2.4 Known Triterpenoids. . . . . . . . . . . . . . . . . . . . . 96
2.2.5 Salvinorins D-F (1d-1f). . . . . . . . . . . . . . . . . . . 98
2.2.6 Divinatorins A-C (28a-28c). . . . . . . . . . . . . . . . . 110
2.2.7 Subsequent isolations. . . . . . . . . . . . . . . . . . . . 118
CONTENTS 13
3 Synthesis. 119
3.1 Known derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.2 Epimerisation at C-8 under Basic Conditions. . . . . . . . . . . 120
3.2.1 Previous Reports. . . . . . . . . . . . . . . . . . . . . . . 120
3.2.2 8-epi-Salvinorins A and B (37a and 37b). . . . . . . . . . 121
3.2.3 Control of Epimerisation and Separation of Epimers. . . 122
3.2.4 8-epi-Salvinorin C (37c) and Related Compounds. . . . . 124
3.2.5 Chromatographic Identification of Epimers. . . . . . . . 125
3.2.6 Mechanism. . . . . . . . . . . . . . . . . . . . . . . . . . 126
3.2.7 Attempted Deacetylation under Acidic Conditions. . . . 128
3.3 Simple Derivatives. . . . . . . . . . . . . . . . . . . . . . . . . . 128
3.3.1 Esters (46 and 47). . . . . . . . . . . . . . . . . . . . . . 128
3.3.2 Attempted Benzyl Ether Formation (48). . . . . . . . . . 129
3.3.3 17-Deoxy Compounds (49 and 50). . . . . . . . . . . . . 130
3.3.4 Tetrahydrosalvinorin A (51). . . . . . . . . . . . . . . . . 131
3.3.5 (+)-Hardwickiic Acid (ent-29a). . . . . . . . . . . . . . . 134
3.4 Modification of the Methyl Ester. . . . . . . . . . . . . . . . . . 134
3.4.1 Relevant Results from Previous Work. . . . . . . . . . . 134
3.4.2 Treatment of Salvinorin A with KOH in MeOH. . . . . . 135
3.4.3 O-Demethylsalvinorin A (67a). . . . . . . . . . . . . . . 144
3.4.4 O-Demethyl-18-deoxysalvinorin A (77). . . . . . . . . . . 150
3.5 Modification of the Ketone. . . . . . . . . . . . . . . . . . . . . 153
3.5.1 Attempted Methylenation. . . . . . . . . . . . . . . . . . 153
3.5.2 Attempted Direct Deoxygenation. . . . . . . . . . . . . . 154
3.5.3 Indirect Deoxygenation. . . . . . . . . . . . . . . . . . . 155
14 CONTENTS
4 Bioassays. 161
4.1 Insect Antifeedant Activity. . . . . . . . . . . . . . . . . . . . . 161
4.2 Eukaryotic Protein Synthesis Inhibition. . . . . . . . . . . . . . 163
4.3 Antimicrobial Activity. . . . . . . . . . . . . . . . . . . . . . . . 164
4.3.1 Bacteria and Fungi. . . . . . . . . . . . . . . . . . . . . . 164
4.3.2 HIV-1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
4.4 NCI Anticancer Screen. . . . . . . . . . . . . . . . . . . . . . . . 168
4.5 Activity at the κ Opioid Receptor. . . . . . . . . . . . . . . . . 170
4.5.1 Other Salvinorins and Divinatorins. . . . . . . . . . . . . 171
4.5.2 Modification of the Ketone. . . . . . . . . . . . . . . . . 173
4.5.3 Modification of the Acetoxy Group. . . . . . . . . . . . . 174
4.5.4 Modification of the Methyl Ester. . . . . . . . . . . . . . 175
4.5.5 Modification of the Lactone. . . . . . . . . . . . . . . . . 176
4.5.6 Modification of the Furan Ring. . . . . . . . . . . . . . . 176
4.5.7 Incorporation into a Revised Binding Model. . . . . . . . 177
4.5.8 Subsequent Results. . . . . . . . . . . . . . . . . . . . . . 179
5 Experimental. 181
5.1 General Conditions . . . . . . . . . . . . . . . . . . . . . . . . . 181
5.1.1 Instruments and Procedures. . . . . . . . . . . . . . . . . 181
5.1.2 Reagents. . . . . . . . . . . . . . . . . . . . . . . . . . . 183
5.1.3 Plant Materials. . . . . . . . . . . . . . . . . . . . . . . 183
5.1.4 Assays. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
5.2 Isolation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
5.2.1 Extraction of Commercial S. divinorum. . . . . . . . . . 185
CONTENTS 15
5.2.2 Extraction of Australian S. divinorum. . . . . . . . . . . 187
5.2.3 Salvinorin A (1a). . . . . . . . . . . . . . . . . . . . . . 188
5.2.4 Salvinorin B (1b). . . . . . . . . . . . . . . . . . . . . . . 189
5.2.5 Salvinorin C (1c). . . . . . . . . . . . . . . . . . . . . . 189
5.2.6 Salvinorin D (1d). . . . . . . . . . . . . . . . . . . . . . . 190
5.2.7 Salvinorin E (1e). . . . . . . . . . . . . . . . . . . . . . . 191
5.2.8 Salvinorin F (1f). . . . . . . . . . . . . . . . . . . . . . . 192
5.2.9 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 Salvinorin C (1c) via acetylation of salvinorin D (1d). . . 199
5.3.2 Salvinorins D (1d) and E (1e) via acetylation of 1h. . . . 200
5.3.3 Salvinorins C (1c) and E (1e) via acetylation of 1h. . . . 200
5.3.4 Dideacetylsalvinorin C (1h) from 1c. . . . . . . . . . . . 201
5.3.5 (+)-Hardwickiic acid (ent-29a). . . . . . . . . . . . . . . 202
5.3.6 Salvinorin A lactol (35). . . . . . . . . . . . . . . . . . . 203
5.3.7 (4R)-3,4-Dihydrosalvinorin C (36c). . . . . . . . . . . . . 205
5.3.8 (4R)-3,4-Dihydrosalvinorin E (36e). . . . . . . . . . . . . 206
5.3.9 (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 Flowering specimen of S. divinorum.1 . . . . . . . . . . . . . . . 31
1.2 Location of Oaxaca and the Sierra Mazateca within Mexico. . . 32
1.3 Young Salvia divinorum plant, Oaxaca.8 . . . . . . . . . . . . . 33
1.4 S. divinorum flower in bud (stereoview).1 . . . . . . . . . . . . . 33
1.5 Ortega et al’s X-ray structure of 1a (stereoview). . . . . . . . . 37
1.6 Salvinorins B (1b) and C (1c). . . . . . . . . . . . . . . . . . . 38
1.7 Known terpenoids. . . . . . . . . . . . . . . . . . . . . . . . . . 38
1.8 Clerodin (5) and standard clerodane numbering. . . . . . . . . . 39
1.9 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 field assay). . 48
1.13 Morphine (11) and derivatives. . . . . . . . . . . . . . . . . . . 54
1.14 Arylacetamide κ opioids. . . . . . . . . . . . . . . . . . . . . . . 56
1.15 Structurally diverse κ opioids. . . . . . . . . . . . . . . . . . . . 59
1.16 Relevant IUPAC fused-ring numbering schemes. . . . . . . . . . 64
2.1 Siebert’s TLC analysis of crude CHCl3 extracts.40 . . . . . . . . 69
2.2 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 1H 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 1H 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 1H 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 1H 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 1H 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 1H 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 A–C (28a-28c) and hardwickiic acid (29a). . . . . 110
2.39 1H 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 1H 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 1H 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/H2SO4. . . . . . . . . 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 µM). 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 affinity and potency of salvinorin A. . . . . . . . . 171
4.9 KOR binding affinities of salvinorins and divinatorins. . . . . . . 172
4.10 KOR activity after ketone modifications. . . . . . . . . . . . . . 174
4.11 KOR activity after acetoxy group modifications. . . . . . . . . . 174
4.12 KOR activity after methyl ester modifications. . . . . . . . . . . 175
4.13 KOR activity after lactone modifications. . . . . . . . . . . . . . 176
LIST OF FIGURES 21
4.14 KOR activity after furan modifications. . . . . . . . . . . . . . . 176
4.15 Westkaemper’s original binding model. . . . . . . . . . . . . . . 177
4.16 Westkaemper’s revised binding model (stereoview). . . . . . . . 178
4.17 KOR binding affinities and potencies of recent derivatives. . . . 179
22 LIST OF FIGURES
List of Schemes
1.1 Cyclization of 7 to the trans-neoclerodane skeleton(stereoview). 41
1.2 Cyclization of 7 to the cis-neoclerodane skeleton. . . . . . . . . 41
2.1 Interconversion of 1c-1h. . . . . . . . . . . . . . . . . . . . . . . 104
3.1 Preparation of known derivatives. . . . . . . . . . . . . . . . . . 119
3.2 Formation of 8-epi-salvinorins A and B. . . . . . . . . . . . . . 121
3.3 Synthesis of 37c-37h. . . . . . . . . . . . . . . . . . . . . . . . 125
3.4 Koreeda et al’s proposed mechanism of epimerisation. . . . . . . 127
3.5 Epimerisation of related natural products with base. . . . . . . . 127
3.6 Brown’s deuteration of 1b. . . . . . . . . . . . . . . . . . . . . . 128
3.7 Deoxygenation of lactol 35. . . . . . . . . . . . . . . . . . . . . 130
3.8 Hydrogenation of 1a and other furanolactones. . . . . . . . . . . 131
3.9 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 BAl2 and BAc2 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
1.1 Relative potency of 1a at cloned κ opioid receptors. . . . . . . . 60
1.2 Alternate systematic names for 1a. . . . . . . . . . . . . . . . . 64
2.1 Effects of solvent and temperature on recovery of 1a. . . . . . . 68
2.2 Data and sources used to identify known compounds. . . . . . . 94
3.1 Coupling constants (Hz) at H-8 for 1a, 1b and 8-epimers. . . . . 122
3.2 Some previously reported furanolactone hydrogenations. . . . . 133
3.3 Summary of results - nucleophilic cleavage of 1a methyl ester. . 148
3.4 Unsuccessful treatment of 1a with excess tosylhydrazide. . . . . 155
4.1 Antifeedant test results. . . . . . . . . . . . . . . . . . . . . . . 162
4.2 KOR radioligand and functional assay results. . . . . . . . . . . 172
5.1 Yields and TLC data (hRf) of isolated compounds. . . . . . . . 187
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 (2−4JHH)
DEPT Distortionless enhancement by polarization transfer NMR (13C mul-
tiplicity)
DMAP 4-Dimethylaminopyridine
DMF Dimethylformamide
DMPU 1,3-dimethyltetrahydropyrimidin-2-one
EC50 Concentration causing 50% of maximal efficacy
GC/MS Gas chromatography/mass spectrometry
HMBC Heteronuclear multiple bond correlation 2D NMR (2−3JCH )
HMQC Heteronuclear multiple quantum coherence 2D NMR (1JCH)
HMPA Hexamethylphosphoric triamide (a.k.a. HMPT)
HPLC High performance liquid chromatography
HRESIMS High resolution electrospray ionisation mass spectrometry
27
28 LIST OF TABLES
hRf = Rf × 100
HSQC Heteronuclear single quantum coherence 2D NMR (1JCH)
IC50 Concentration causing 50% inhibition
InChI IUPAC International Chemical Identifier
K i Receptor binding affinity constant
KOR κ (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 effect (through–space signal enhancement)
NOESY Nuclear Overhauser enhancement spectroscopy (through–space 2D
NMR)
PCB Polychlorinated biphenyl
PCR Polymerase chain reaction
Rf Retardation factor af
[distances of analyte (a) and solvent front (f) from
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
LIST OF TABLES 29
TLC Thin layer chromatography
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 CHAPTER 1. INTRODUCTION.
Mexico
N
Oaxaca
Sierra Mazateca
0 300 km200100
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 & Játiva.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 flowers are distinctive (Figure 1.1), the plant’s 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 flower (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 Was-
son and Hofmann were the first to obtain a flowering specimen, which is essen-
tial for species identification. 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 mis-
named “Wasson and Hofmann” strain.12
Figure 1.4: S. divinorum flower in bud (stereoview).1
There has been no report of growing the plant from seed: propagation is ex-
clusively vegetative.6, 13 Moreover, there are no published first-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 CHAPTER 1. INTRODUCTION.
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 patient’s 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 effects, larger doses of the infusion are used, or the leaves themselves
are slowly chewed and eaten. Eating the leaves is very difficult, due to their
sickeningly bitter taste, and often induces vomiting. First-hand accounts of
the effects 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 first 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 flowers. 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 ška María pastora (the leaves of Mary shepherdess), appar-
ently 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 in-
ferior 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, drink-
ing an infusion has been shown to be very inefficient; much stronger effects
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 Wasson’s suggestion19 that S. divinorum may represent
the divinatory plant pipiltzintzintli cultivated by the Aztecs.7 The Aztecs pre-
pared 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 pipiltz-
intzintli make no mention of seeds, unlike other Aztec visionary plants; and S.
divinorum is effectively seedless, unlike other contemporary Mexican visionary
plants. Here, then, we have a plant without an indigenous name, and an in-
digenous 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. Díaz has pointed out that male and female gen-
36 CHAPTER 1. INTRODUCTION.
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 affect the plant’s
potency, and the addition of other plant parts, while superfluous, would not
affect the infusion. Moreover, Díaz’s preferred candidate, Cannabis sativa, can-
not be correct, since it was introduced after European settlement.7 Valdés also
rejects Wasson’s hypothesis, preferring Beltrán’s proposal that pipiltzintzintli
was a synonym for the morning glory, ololiuhqui.13 Ott has shown, however,
that these were explicitly described as different plants, and that S. divinorum
is thus
... the only Mexican entheogenic plant which fits 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. Valdés notes early anecdotal reports of use by the
Cuicatec and Otomí tribes, whose lands adjoin the Sierra Mazateca.21, 9
1.3 Chemistry.
1.3.1 Terpenoids.
1.3.1.1 Isolation.
The first compound isolated from S. divinorum was 1a, a clerodane diter-
penoid discovered by Ortega et al in 1982 and named salvinorin.22 The ter-
minology of these terpenoids will be discussed in the next section. Extraction
of the dried leaves in refluxing CHCl3, chromatography on “Tonsil” activated
clay and crystallisation from MeOH gave 1a in unstated yield. The structure
1.3. CHEMISTRY. 37
1a
O
O
O
H HOO
O
O O
Figure 1.5: Ortega et al’s X-ray structure of 1a (stereoview).
was elucidated spectroscopically, and confirmed by X-ray crystallography.22
Absolute stereochemistry was tentatively assigned by comparing the circular
dichroism spectrum with known compounds.
Subsequently, Valdés isolated the same compound. Unaware of Ortega’s 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 pub-
lished, this oversight was corrected, and the compounds were named salvi-
norins A and B.23 Valdés’s isolation procedure was more complex. The Et2O
extract was dissolved in MeOH and washed with hexanes. Repeated chro-
matography 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 confirmed by
X-ray crystallography. The fit of the model was in excellent agreement with
the diffraction data (R-factor = 8.7 %), though not quite as good as Ortega
et al’s (5.2 %).
38 CHAPTER 1. INTRODUCTION.
1b
O
O
O
H HOHO
O O 1c
O
O
O
H HOO
O
O
O O
Figure 1.6: Salvinorins B (1b) and C (1c).
Ortega et al’s model also has the advantage of including the positions of hydro-
gen atoms. Unfortunately, the absolute stereochemistry of Valdés et al’s model
is incorrect, although it was correctly assigned in the paper, again on the basis
of circular dichroism. This stereochemistry has since been definitively con-
firmed 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
Valdés et al subsequently isolated salvinorin C (1c).27 Repeated chromatog-
raphy of the Et2O extract followed by HPLC gave 1c in 78 mg/kg yield.
Spectroscopic structure elucidation was supported by partial synthesis of ana-
logues.
4
HOH
H
H
3
HO OO
H
2
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 Valdés,
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 nor-
triterpenoid stigmasterol (3) and the diterpenoid neophytadiene (4).29
1.3.1.2 Terminology.
5
H
OOAc
O
O
H H
H
OAc
22
3344
55
101011
66
77
8899 1717
12121111
1313
1616
1515
1414
1818
1919
2020
Figure 1.8: Clerodin (5) and standard clerodane numbering.
Clerodane diterpenoids30 are named after clerodin (5). Conclusively establish-
ing 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 com-
pounds 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 configuration 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 isopen-
tenyl diphosphate (6).33 Assembly of four isoprene units gives geranylger-
anyl diphosphate (7), the final common biosynthetic intermediate of all diter-
40 CHAPTER 1. INTRODUCTION.
OPP
7
P
O
O OHOH
OP
O
OH6
= OPP
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 in-
termediate 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 cis-
clerodanes, 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 confirms 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 as-
sumption.37, 38 An alternate pathway, via 1-deoxyxylulose 5-phosphate, has
been established. Indeed, the mevalonic acid pathway makes a negligible con-
tribution to diterpenoid biosynthesis. Nonetheless, as reflected 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
H+
++
7
14
1819
1819
15
14
15
7
99
H+
OPP
++
18 1918 19
88
CH2OPP
CH2OPPCH2OPP
CH2OPP
OPP
Scheme 1.1: Cyclization of 7 to the trans-neoclerodane skeleton(stereoview).
55
101088
99
1818 1919
H 19191818
H
+++
8 10
H
H+
H
10
7
OPP OPP OPP OPP
Scheme 1.2: Cyclization of 7 to the cis-neoclerodane skeleton.
42 CHAPTER 1. INTRODUCTION.
0 20 µm0 20 µm
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 (Fig-
ure 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 finding is significant for future isolation work. Siebert found that dipping
fresh leaves in CHCl3 (30 seconds × 3) gave nearly complete recovery of salvi-
norins.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
0 200 µm
0 200 µm 0 200 µm
0 200 µm
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 Alleged alkaloids.
1.3.2.1 Summary of the Original Report.
The first investigation into the chemistry of S. divinorum was reported by
José Luis Díaz in 1975.41 The report is in Spanish, but key sections have been
translated by Valdés, who also added flowcharts clarifying Díaz’s procedures.42
The report was not peer reviewed, and the workers involved were not named.
Valdés, who later visited the lab, reports that the chemistry was performed
44 CHAPTER 1. INTRODUCTION.
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 specified: “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 deficiencies, the procedures and results will not be repro-
duced here. The interested reader will find Valdés’s translation helpful.42 One
important omission should be noted: the isolation procedure Valdés describes
as method 2 was performed twice: on the first 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 briefly summarised as follows: certain fractions of the
plant extract, soluble in aqueous acid but insoluble in aqueous base, some-
times 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 Dragendorff’s reagent,44 a standard alkaloid
visualisation reagent, in both standard and modified (Lüdy-Tenger)45, 46 forms.
On this basis, Díaz 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 difficult to identify the substance(s) respon-
sible for these interesting effects. There exists a great variability
or instability in the constituents of S. divinorum, which has im-
peded the consistent reproduction of the mental or behavioural al-
terations, preventing the identification of the active fraction. Some
1.3. CHEMISTRY. 45
initial observations indicated the presence of nitrogenous compounds,
possibly amino acids or amines, although they now appear to be of
no pharmacological interest.47
These remarks, while vague, are not consistent with the original report. Pre-
sumably 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 Dragendorff’s reagent,51 but
far more tested negative. Furthermore, Dragendorff’s gives false positives with
numerous non-nitrogenous compounds.44
The American species S. reflexa gave false positives to several alkaloid test
reagents.52 The compound responsible proved to be choline [Me3NEtOH]+.
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 Dragendorff’s-positive compounds,44 and thus
none of these compounds can account for Díaz’s results. Díaz41 cites a sec-
ondary 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 chemotaxo-
nomic 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 CHAPTER 1. INTRODUCTION.
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 Díaz’s findings, however. Valdés
found no compound in the crude extract which gave a positive reaction to Dra-
gendorff’s or other alkaloid-specific test reagents.56, 23, 57 He has hypothesised
that Díaz’s group actually used mislabelled Erlich’s reagent, which reacts with
alkaloids but also furanolactones, and thus mistook the salvinorins for alka-
loids.57
This hypothesis, however, is inconsistent with the published claims; all defi-
nitely active fractions were soluble in aqueous acid, unlike 1a. Thus, if the ac-
tive fractions contained the salvinorins then not only the visualisation reagent,
but the fractions themselves, must have been misidentified. Moreover, the ac-
tivity reported by Díaz was dramatically different from that later observed
in cats by Valdés. Díaz reported “intense attention”, “reactions of fear and
attack” and “fury”, lasting 10 minutes after intravenous injection.41, 42 The ef-
fects observed by Valdés 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 effects lasted over 24 hours rather than 10 minutes.58
Chemical investigations by several groups have now yielded a total of 20 ter-
penoids, 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, identifiable 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, Díaz’s tests in cats yielded no useful results. The next
worker to study the plant’s pharmacology was Valdés. 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 “precluded”61
(i.e. forbidden by his supervisor).
For the purposes of bioassay-guided fractionation, Valdés therefore had to
develop an animal assay. This presented an enormous challenge. Previous
attempts to identify the active principles of hallucinogenic plants using ani-
mal assays, even when sustained and well funded, had failed.7 In each case
the active principle was later identified, 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 effect other than
“disquiet”.10 To these must of course be added Díaz’s work with S. divinorum.
Evidently, and unsurprisingly, it is practically impossible to tell if an animal
is hallucinating. In Valdés’s preliminary trials, standard hallucinogen assays
in rats and cats were not sensitive to the effects of the extract.63 However, the
cats exhibited impaired motor coordination, which reminded him of the im-
paired physical coordination and slurred speech caused by the infusion.16 The
duration of action was much longer, however (24 hours). Given the difficulty
of detecting psychological states in animals, Valdés decided to choose an assay
sensitive to this physical effect.
48 CHAPTER 1. INTRODUCTION.
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 insufficiently so. Next, a
modified65 version of the open field 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 effects, with a clear dose-response relation-
ship. Assay-guided fractionation led to the identification of 1a as the active
principle. No activity was seen with 1b, the only other pure compound iso-
lated. However, various mixed fractions gave puzzling results.
lines crossed
1a
0
50
100
150
200
250
300
dose (mg/kg)
lin
es
cro
ssed
rearings
impure 1a (~10% 1c) mixtures
0
10
20
30
40
50
60
dose (mg/kg)
reari
ng
s
immobility
0
2
4
6
8
10
12
14
16
dose (mg/kg)log scalelog scalelog scale
10 20 30 40 50 10 20 30 40 50 100 10 20 30 40 50 100100
tim
e im
mo
bile (
min
)
Figure 1.12: Activity of pure 1a versus mixtures in mice (open field assay).
Using the open field assay, impure 1a (later found to contain ∼10% 1c) was
found to be “significantly more potent”27 than the pure compound.65 It was
therefore long suspected13, 27, 67 that 1c was also psychoactive.
The open field 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 Valdés’s 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 efficacies 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 efficacy at high doses, however, these
1.4. PHARMACOLOGY. 49
fractions were not in fact more potent: the lines of best fit suggest that pure
1a had comparable efficacy at intermediate doses, and higher efficacy 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 specific to the open field assay: the same pattern was
evident69 in the inverted screen assay.
The confounding factor was probably differences 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 effective for some
hydrophobic drugs,71 and 1a is indeed hydrophobic. However, to be effectively
delivered in an emulsion, drugs must also be highly lipophilic:
Generally the most difficult 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 Valdés 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 amor-
phous solid even when pure. Hence, all four of the impure fractions shown
in Figure 1.12 were amorphous. This would result in differences 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 emul-
sion, the amorphous solids could form a dispersion of microscopic particles,
50 CHAPTER 1. INTRODUCTION.
with a higher surface area than macroscopic crystals. This in turn would in-
crease 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 influence absorption.
Many instances have been reported of enhanced absorption of active com-
pounds from a crude extract relative to the pure compound.76 For instance,
some terpenoids are known to act as permeation enhancers,77 increasing trans-
dermal absorption of co-administered drugs up to 90×. Thus, inactive com-
pounds 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/H2O), 1 mg/kg of 1a
reduced locomotor activity significantly over 30 minutes.78 Reinvestigation
by Valdés et al confirmed this, showing that 0.5 mg/kg caused maximal im-
pairment in the inverted screen test;79 quadrupling the dose did not increase
efficacy. By contrast, in the original tests using the emulsion, maximal im-
pairment 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 effect of 1a; performance on
the inverted screen test remained impaired after 30 minutes.15 By comparison,
the effects of the solution peaked within 5 minutes, and were undetectable by
15 minutes.79 Consistent with this, a recent study found a strong analgesic
effect 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
effect even at 40 mg/kg.81 Another study reported analgesic effects at 0.6
mg/kg, but unfortunately contained no detail on administration or timing.82
In summary, the open field results may have been confounded by differences
in bioavailability. The vehicle used gave very poor absorption, and superior
absorption of amorphous solids is typical. Thus, the differences 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 µg/kg) produced only “slight
overt behavioural effects”,83 described as “sedation-like” in both studies, con-
firming the difficulty of establishing animal models of psychoactivity.
1.4.2 Human Testing.
As mentioned above, Valdés’s open field assay results indicated that 1a was
the active principle of S. divinorum. Further testing with other compounds
in the open field indicated that 1a was approximately equal in potency to
mescaline.13, 14 However, this conclusion remained tentative. As Valdés had
written earlier,
... there is no definite 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 finally resolved by Siebert in 1994.18 He reported that
1a was hallucinogenic when vaporised and inhaled. Activity was percepti-
ble with doses as low as 200 µg. The effects commenced within seconds, and
the strongest effects lasted 5 – 10 minutes. No effects were detectable after
30 minutes, suggesting a half-life of under 10 minutes. Siebert also explored
the effect of the route of administration. Swallowing encapsulated 1a in very
large doses (10 mg) produced no effect. This also held for the plant: an in-
fusion prepared from ten fresh leaves, when swallowed, produced no effect in
any subject. The same amount, when held in the mouth for 10 minutes and
spat out, produced definite effects in all subjects. Thus, absorption through
52 CHAPTER 1. INTRODUCTION.
the oral mucosa is clearly far more efficient than through the gastrointestinal
system. Nonetheless, an ethanolic solution of 1a gave inconsistent effects sub-
lingually. The onset of effects is slower by the sublingual route, and the effects
last longer.
Siebert’s results were soon confirmed by others.7, 85 Ott and Gartz86 also re-
ported that sublingual application of 1a was effective in acetone or Me2SO,
with potency comparable to inhalation. Thus, far from being equipotent with
mescaline as the open field assay indicated, 1a is in fact ∼1000× more po-
tent.14, 7 Indeed, it is the most potent naturally-occurring hallucinogen yet
isolated.7 The rapid metabolism of 1a has been confirmed 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 effects
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 expe-
riences 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. Siebert’s initial
reports of these unique themes have again been confirmed by others.85, 89
Further research has revealed no evidence of active compounds other than 1a.
Siebert confirmed that 1b is inactive.40 He also found that self-administration
of 3 mg90 of 1c, sublingually in acetone, had no noticeable effect;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 definitive proof awaited human testing, Valdés nonethe-
less correctly identified 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.
Valdés’s supervisor had thus forbidden the only demonstrably effective tech-
nique available. Moreover, in this case even human testing had failed. And it
had been performed, independently, by arguably the two greatest authorities
in the field, Alexander Shulgin91 and Albert Hofmann92, 10 (twice in the latter
case).93 Consider also the prevailing consensus when Valdés began work: the
active principle, whose effects 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 al-
kaloid(s), which was also false. Sadly, despite overcoming these myths, Valdés
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 µM.
Subsequently, the compound was screened by Roth et al against a much larger
battery of targets, the National Institute of Mental Health’s Psychoactive Drug
Screening Program. This revealed that 1a bound with high affinity to the κ
opioid receptor (K i = 4 nM).67 Functional testing showed that it activated the
receptor with full efficacy and high potency (EC50 =1 nM). Thus, 1a is a potent
full agonist at the κ opioid receptor. This result has since been replicated
by other groups.82, 81, 95 Further confirmation has come from in vivo testing.
The effects of 1a in mice are blocked by selective κ opioid antagonists;81, 78, 80
rhesus monkeys trained to discriminate κ opioids recognised 1a as such, and
54 CHAPTER 1. INTRODUCTION.
the effects were blocked by a nonselective opioid antagonist.83
No binding was apparent in vitro to any of 48 other CNS targets at 10 µM; 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 affinity testing cannot yet be
performed.96 However, in vivo confirmation of the selectivity of 1a is available.
Siebert found that the effects of 1a were blocked by naloxone, a nonselective
opioid antagonist.97 This makes it highly unlikely that a nonopioid mechanism
contributes independently to the compound’s effects.
1.4.4 Mechanism: κ Opioids.
1.4.4.1 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 Sertürner in 1806; he also showed the com-
pound to be a potent analgesic.98
OHO OH
N
11
H
HO
N
OOHO
N
12
OH
HO
N
13
O
14
H
Figure 1.13: Morphine (11) and derivatives.
Morphine has had an immense impact on science. It was the first drug – the
first known pharmacologically active compound.99 It was the first 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 morphine’s discovery predated
the concept of molecular structure by several decades. Indeed it predated the
publication of Dalton’s atomic hypothesis, which was regarded with skepticism
for decades.100
Morphine’s very strong analgesic effect is accompanied by serious side effects:
nausea, respiratory depression and constipation.101 Intense euphoria also oc-
curs in some, making the drug highly addictive. The pursuit of an analgesic
lacking these side effects gave rise to medicinal chemistry. The first 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 Simplified structures
were found to retain activity: morphinans such as butorphanol (12), ben-
zomorphans such as ketazocine (13) and remarkably even phenylpiperidines
such as pethidine (meperidine, 14).
Some of these compounds (such as 14) closely mimic morphine’s actions.
Some, however, proved to be antagonists, which have proven immensely valu-
able in reversing opioid overdose. Other derivatives caused quite distinct be-
havioural effects, but were not antagonists and did not exhibit cross-tolerance.
The study of these differences led to the discovery of opioid receptor subtypes.
The µ and κ subtypes were named after the archetypes morphine (11) and
ketazocine (13), while δ 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 officially sanctioned names, OP1−3,102, 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 effects are
reversed by the antagonist naloxone.101 This terminology lends itself readily to
56 CHAPTER 1. INTRODUCTION.
selective compounds; thus, 1a will be referred to below as a “κ opioid.” This
is synonymous with the common but unwieldy tautology “κ opioid receptor
agonist.”
1.4.4.2 Development.
R O
N
O
N
Cl O
16
1718
O
N
N
Cl
Cl
15
Cl
Figure 1.14: Arylacetamide κ opioids.
U50,488 (15) was devised as a structurally simplified morphine derivative.104, 105
In animal tests, U50,488 was an effective analgesic, whose effects 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 non-µ opioid, 15 was soon found to be the first selective agonist at the κ
opioid receptor.
These remarkable findings 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 (CI-
977, 18). Testing of these compounds confirmed the findings with 15: at last it
appeared that the long-awaited nonaddictive opioid analgesics had been found.
1.4.4.3 CNS Effects.
Addiction can be studied using animal models. The µ 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,
κ 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, κ opioids
act as a punishment, where µ opioids act as a reward. This is a vast field of
research in its own right, and a thorough review is available.106 The mechanism
of these effects, 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 κ opioids was initially regarded as desirable. Human
tests soon revealed a problem: rather than euphoria, κ opioids produced dys-
phoria. This lowering of mood was accompanied by “psychotomimetic” effects:
confusion, hallucinations and depersonalisation.
Enadoline (18), for instance, caused side effects such as somnolence, hallucina-
tions, anxiety, depersonalisation, confusion and abnormal thinking; the sever-
ity of these effects 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 floor, and felt his body was blending into them109 (cf. 1a).18 Spi-
radoline (17) caused altered perceptions, impaired coordination and slurred
speech; subjects reported being more irritable, anxious and sad.110 Responses
to other κ opioids range from “personality disorders and mild confusion”111
to “depersonalisation”, “dreamlike” experiences and “episodes of unmotivated
and uncontrolled laughter”112 (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 effects?” They consistently
answered yes after administration of a µ opioid, but no after enadoline (18).109
The opposite was true when asked about “bad effects.” Similar results were
reported in tests of nonselective κ opioids such as 13113 and 12.114 Subjects
with extensive histories of illicit drug use reported dysphoria, hallucinations,
58 CHAPTER 1. INTRODUCTION.
paranoia and confusion after the κ opioids, but euphoria after morphine. Thus,
both drug-naïve subjects and experienced illicit drug users overwhelmingly find
κ opioids dysphoric and aversive.
1.4.4.4 Therapeutic Potential.
Clinical development of these compounds as analgesics was eventually aban-
doned due to these psychological effects.115, 105 Some κ opioids which do not
cross the blood-brain barrier, and are therefore not psychoactive, remain in de-
velopment 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 κ opi-
oids.115, 118, 119, 120 None of these therapies has reached clinical use.
There has been a report of an antidepressant effect 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 κ opioid had no effect.123 However, other studies reported
a prodepressant effect,124 one of which used 1a.107 This conclusion, which
is consistent with the aversive and dysphoric effects discussed above, is also
more consistent with prior results and recent theoretical insights.107, 125 Thus,
it seems that κ opioids exacerbate depression in animal models. Surprisingly,
however, there have been reports of antidepressant effects 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 µ opioids is followed by prolonged
dysphoria, the converse may be the case with κ opioids.
1.4.4.5 Claimed Subtypes.
There have been claims that certain κ opioids lack dysphoric effects. TRK
820 (19, Figure 1.15 on the facing page) allegedly produces “moderate be-
havioural/psychological side effects, 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 differences between κ opioids is that there are
subtypes of κ opioid receptor.131 Differences between κ opioids might reflect
different 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 different receptors.131, 132
1.4.4.6 Structure, Potency and Selectivity.
S
OHN
N
NF
20
OHO N
N
OH
O
O
19H HO
N
13
O
O
N
O
N
O 18
phenylalkylamine moiety
Figure 1.15: Structurally diverse κ opioids.
The major structural classes of κ opioid are peptides such as dynorphin A, mor-
phine derivatives such as 19, benzomorphans such as 13 and arylacetamides
such as 18 (Figure 1.15). The first major departure from these categories was
the benzodiazepine tifluadom (20).105 The discovery of such compounds, and
60 CHAPTER 1. INTRODUCTION.
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 alka-
loids. 1a was the first non-nitrogenous opioid reported. Indeed, it was the
first opioid lacking the phenylethylamine moiety or its propyl homologue (Fig-
ure 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 phenylalky-
lamine moiety, versus 8% of non-psychoactive compounds.134 Conversely, 58%
of phenylalkylamines were psychoactive, versus 3% of other compounds. Be-
ing non-nitrogenous and lacking a benzene ring, 1a is thus not merely unique
among opioids, but extremely unusual among psychoactive compounds in gen-
eral.
EC50 (nM) Ref1a 15 16 191 1.2 677 24 13 135
4.5 4.5 9545 207 1363.1 3.9 1374.6 2.2 0.025 81
1.1 0.0048 13816 0.15 139
Table 1.1: Relative potency of 1a at cloned κ opioid receptors.
The binding affinity 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
κ opioid to date appears to be TRK 820 (19). However, few groups have
studied this compound, and their binding affinity data are wildly discordant
(K i = 75 pM81 vs 3.5 nM).139 The compound also has lower µ/κ 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 affinity at the
µ opioid receptor (K i > 10 µM).
In summary, κ opioids have not lived up to initial expectations as analgesics.
1.5. TOXICOLOGY. 61
Ultimately, the enormous effort to improve upon morphine has failed. Two
hundred years after its discovery, morphine remains “the opioid of first choice”
for severe pain under World Health Organization guidelines.140 There is no
superior analgesic:
Morphine remains the most widely used opioid for the manage-
ment 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. Valdés
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. Subcu-
taneous 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 ex-
tracts,143 and veterinary examinations suggesting this was the cause,58 Valdés
prepared a polyphenol-free extract by partitioning between CH2Cl2/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 methano-
lic 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 CHAPTER 1. INTRODUCTION.
approximately equal to the toxicity of vitamin B3.144 Pure 1a caused no fatal-
ities 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 mislead-
ingly low impression of acute toxicity. On the other hand, the undissolved
material at the injection site may have had toxic effects through another mech-
anism, 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 effects.145 These data therefore cannot be used to
make reliable comparisons with other substances.
Mowry et al later published a study specifically devoted to the toxicology of
1a.146 Injection of 6.4 mg/kg/day for two weeks caused no fatalities; post-
mortem examination of organs and tissues showed no apparent changes. Un-
fortunately, since the drug was administered intraperitoneally as a “fine sus-
pension,” and behavioural effects were not described, the same issue of bioavail-
ability arises as with Valdés’s 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 effects or hospitalisation related to S. divinorum or salvinorin
A.147
1.6 Social impact.
1.6.1 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 Díaz also reported this at the time, the latter was in fact
referring to a different plant.41
The plant nonetheless remained extremely obscure until Siebert’s findings were
popularised in 1996 by Turner.85 This immediately overturned the plant’s
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 [...] be-
yond those experienced with any other psychedelic ...85
The plant’s 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 find 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 Legal Status.
1.6.2.1 Australia.
In 2002, Australia became the first country to prohibit S. divinorum and salvi-
norin A. The committee responsible, the National Drugs and Poisons Schedule
Committee (NDPSC), justified 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 CHAPTER 1. INTRODUCTION.
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.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
8-METHOXYCARBONYL- 7-carboxylic acid, methyl ester4A,8A-DIMETHYL- 6a,10b-dimethyl-
6-ACETOXY 9-(acetyloxy)-5-KETO 4,10-dioxo-
3-(4-FURANYL) 2-(3-furanyl)3,4,4B,7,9,10,10A-SEPTAHYDRO dodecahydro-2,1-NAPHTHO[4,3-E]PYRONE 2H -Naphtho[2,1-c]pyran
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.
11
2233
44
4a4a
4b4b
5566
7788
8a8a99
1010
10a10a
phenanthrene
444a4a10b10b
1122
O33
55
666a6a
7788
991010
10a10a
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. 65
I tried to get an IUPAC name out of this; unfortunately, I gagged
when trying to decide whether it was a naphthopyrone, or a modi-
fied phenanthrene, or whether it should start from cyclohexane in-
stead of aromatic rings, or whatever. The fact that phenanthrene
is numbered funny didn’t 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 haven’t done o-chem since the Reagan adminis-
tration.151
This explains the incorrect numbering of 1a: White named it as a naph-
thopyran 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 sub-
stituents in nonalphabetical order, in conformance with their own unspecified
“naming conventions.” The new name is therefore also incorrect.
The NDPSC is Australia’s peak body for the regulation of chemicals, whose
chief task is to list controlled chemicals by name. It is surprising and dis-
turbing, therefore, that no member of this body understands basic chemical
nomenclature. White’s posting also raises another serious issue. The news-
group alt.drugs156 is a forum for users, dealers and manufacturers of illicit
drugs to share ideas and advice. Publications which “promote, incite or in-
struct 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 CHAPTER 1. INTRODUCTION.
A disturbing question thus remains unanswered: how did obvious errors in a
prohibited publication find 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 White’s 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 κ opioid. How-
ever, virtually nothing was known of its structure-activity relationships, and
only three other compounds had been isolated from S. divinorum. The suspi-
cion that other psychoactive compounds were present persisted.
Chapter 2
Isolation.
2.1 Isolation Procedure.
2.1.1 Extraction Conditions.
The isolation methods reviewed above suffer from a number of drawbacks.
Both Valdés et al23 and Ortega et al22 employed refluxing solvents for the ini-
tial 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 refluxing for two hours, in either CHCl3 or MeOH
(see Table 2.1 on the next page). In optimising the room-temperature ex-
traction, 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 di-
lute. Furthermore, the decline is faster at reflux, when solubility is greater.
A more plausible explanation of the decline is decomposition, with the rate
proportional to temperature. This conjecture was confirmed 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 be-
came apparent by TLC (70% Et2O/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, Valdés reports that he recovered 1a
unchanged after refluxing in MeOH or MeOH/H2O for 2 weeks.165 It may be
that decomposition only occurs in the presence of oxygen.
recovery (mg/g)T time CHCl3 MeOHrt 4 d 1.75 1.56
reflux 100 min 1.41 0.44
Table 2.1: Effects 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 affordable solvent meeting both
of these criteria. Dried, powdered leaves were steeped for 30–60 minutes in
room-temperature acetone (× 3). TLC analysis (5% MeOH/CH2Cl2) showed
that the second extract was much lower in 1a than the first; 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 confirmed: 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 Problems Caused by Pigments.
2.1.2.1 Interference with Previous Isolations.
1a1b
youngest oldest
Developed in 50% EtOAc/petrol, visualised in vanillin/H2SO4.
Figure 2.1: Siebert’s TLC analysis of crude CHCl3 extracts.40
The extract from the above procedure was a blackish-green tar. The numer-
ous pigments present in this extract rendered TLC analysis difficult. Relative
to the colourless terpenoids of interest, the pigments were present in larger
quantities and stained more vividly, obscuring the terpenoid spots. The prob-
lem 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 figure; in this system it appears immediately above 1a,
which is itself nearly obscured in the youngest leaf extract.
Valdés et al had addressed this problem by partitioning the crude extract be-
tween aqueous MeOH and hexanes.23 Using his partitioning procedure proved
difficult: both phases were black, and could not be distinguished even by close
70 CHAPTER 2. ISOLATION.
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 difficulties with Valdés et al’s 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 re-
moved by rinsing with cold MeOH.167 It was presumably to remove these trace
pigments that Valdés 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 simplified Valdés et al’s 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 chromatog-
raphy. Tidgewell et al169 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 prob-
lems 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
2.1. ISOLATION PROCEDURE. 71
of this section, the pigments made the mere detection of other compounds
by TLC difficult. It was therefore critical to this work that the crude ex-
tract itself be decolourised, rather than individual compounds after isolation.
Valdés et al’s partitioning procedure was unsatisfactory, as was chromatog-
raphy 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 differences 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
O
N N
N N
Mg
O
O
O
21
22
OOH
HO O
OH
OOH
OHOH
O
OH
23
Figure 2.2: Representative major plant pigments.
The overwhelming majority of plant pigments fall into three categories: chloro-
phylls (eg. chlorophyll-a, 21, Figure 2.2), carotenoids (eg. β-carotene, 22) and
flavonoids (usually present as glycosides eg. quercitrin, 23).170, 171 In addition
72 CHAPTER 2. ISOLATION.
to these ubiquitous classes, there are many minor pigments confined to partic-
ular taxa, which have been reviewed elsewhere.170 The common factor among
these pigments is an extended π system, often but not always aromatic, which
results in intense absorption of visible light. As a result, pigments are gener-
ally 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
Valdés’s partitioning procedure, and polarity-based fractionations in general,
are of limited effectiveness 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 decolouris-
ing carbon.172 Activated carbon has been in widespread use, both scientifi-
cally and industrially, for centuries. The characteristics of compounds which
affect adsorption are well-established empirically.173 Within a homologous se-
ries of compounds, adsorption increases with length. In addition, aromatic
compounds are strongly adsorbed, especially polycyclic aromatic compounds.
Generally, increased polarity reduces adsorption, but this effect is smaller, and
easily overcome by size. Polyphenols, for instance, are strongly adsorbed de-
spite being freely soluble in water. Many other variables affect adsorption to
lesser extents.173 While these general trends are well-established, the underly-
ing mechanisms remain in dispute.
2.1.3.1 Mechanism of Action.
The surface chemistry of different 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 im-
portant common characteristic of all forms is their enormous surface area,
2.1. ISOLATION PROCEDURE. 73
estimated at up to 1500 m2/g in some cases,174 resulting from the material’s
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 sol-
vent cavity around the solute, and (favourable) van der Waals attraction be-
tween the solute and the carbon surface. As a result, “the theory predicts
the adsorbability to depend linearly on the nonpolar surface area of the adsor-
bate”.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 different 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 dis-
tance of separation between the surface of the activated carbon
and the center of the solute molecule, it is reasonable that a pla-
nar molecule would be adsorbed more strongly than an otherwise
similar globular molecule.177
That is, consistent with the solvophobic model, surface area available for con-
tact is the key factor in determining adsorbability. Interestingly, this effect was
found not only for aromatic compounds; planar olefins were just as strongly
adsorbed. A similar effect had been observed in earlier work,178 based on an-
other model, and used to predict that square planar organometallic complexes
would adsorb more strongly than octahedral complexes. This prediction was
confirmed experimentally.178 This also helps explain the strong adsorbability
of carotenoids such as 22, which although aliphatic can assume planar confor-
mations which present a much larger surface area to the carbon surface than
less conjugated hydrocarbons. Indeed, in this work non-conjugated terpenoids
74 CHAPTER 2. ISOLATION.
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 affect 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 remov-
ing coloured impurities from crude compounds. Typically, a small amount
of powdered activated carbon (1–2% w/w relative to the crude compound) is
added during recrystallisation; the decolourised solution is then filtered before
cooling.172 In many cases, however, this is not effective. Indeed, during the
preliminary work described above, flash column chromatography of the crude
extract gave an orange solid which was rich in 1a by 1H NMR. During re-
crystallisation from MeOH, the bright yellow colour of the solution was not
diminished by activated carbon, even at 10% w/w. Upon cooling, the result-
ing 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 effective, 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
“effectively a chromatographic procedure.”172
2.1. ISOLATION PROCEDURE. 75
2.1.3.3 Use in Chromatography.
Activated carbon is rarely employed as a stationary phase in chromatography.
It is not suited to general use; for most separations, other adsorbents offer
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 conven-
tional adsorbents is in the separation of fullerenes,185, 186 giving better reso-
lution and recovery than alumina, along with higher speed and lower cost.
The larger ellipsoidal fullerenes, above C70, are very difficult to desorb.186
Activated carbon has also proven valuable in separation of polychlorinated
biphenyls (PCBs).187, 188, 189 Commercial PCB mixtures are very complex, con-
taining hundreds of PCBs differing in the number and position of chlorine sub-
stituents.189 Many of these compounds differ very little in polarity and volatil-
ity, resulting in incomplete resolution by either HPLC or GC. This problem
can be resolved by multi-dimensional GC, but activated carbon permits sepa-
ration on a different 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 simplified relative to the initial mixture, and can
be further purified using conventional techniques.187, 188 This method also has
the attraction that degree of o-substitution is of great biological significance.
In addition to their toxicity, non-o-substituted PCBs such as 24 are extremely
teratogenic; this effect is reduced by one o-substituent, and abolished by more
than one.187
76 CHAPTER 2. ISOLATION.
ClCl
ClCl
Cl
Cl
Cl Cl
Cl
Cl
24 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 purification of individual compounds.
As with any adsorbent, eluent selection is critical to achieving effective sepa-
ration. The eluotropic series on activated carbon given by Gordon180 is:
H2O < MeOH < EtOH < acetone < iPrOH < Et2O < EtOAc < hexane �benzene.
In general, then, the larger and less polar the solvent, the greater its elut-
ing strength. That benzene is the strongest eluent in the series is unsurpris-
ing given the strong adsorption of aromatic compounds. In recent decades,
toluene has been used as a less toxic and equally effective substitute for ben-
zene.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 con-
stants,177 p-xylene would probably offer comparable eluting strength, with the
advantage of a lower boiling point (138 vs 178 ◦C). Another approach to elu-
tion 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 “fluidise” the activated carbon, using a
stream of nitrogen to remove the finest particles.187 Similarly, industrial-scale
water treatment usually employs granulated activated carbon.174 Using larger
2.1. ISOLATION PROCEDURE. 77
particles in this way, however, has the undesired effect of reducing available
surface area. A more common solution is therefore to use mixed adsorbents.180
The most common choice is diatomite filter 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 filter aid
was selected due to its low cost, safety and ready availability. The proce-
dure was performed as standard flash 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 Gor-
don’s eluotropic series,180 through Et2O, 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 sufficient 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 sub-
sequent 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 finer particles than flash
78 CHAPTER 2. ISOLATION.
chromatography grade silica; these can contaminate fractions and block
glass frits. A layer of filter aid below the carbon prevents this.
• As noted by others,185 the carbon/filter 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 difficult 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 filtration under vacuum proved more convenient and
practical than flash 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.
Separatory funnel
Vacuum adaptor
Glass fritFilter aid
Activated carbon/filter aidSand
Figure 2.4: Apparatus for Filtration through Activated Carbon.
While some separation between compounds in the decolourised fractions was
2.1. ISOLATION PROCEDURE. 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 filtration 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 filtration provides a much
greater reduction in mass and complexity of the extract than more common
partitioning procedures. Thus, as seen with PCBs, analysis is simplified by
utilising an extra “dimension” for separation (surface area/planarity), com-
pared to the typical reliance on polarity differences. Using the procedure de-
scribed 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 Lipinski’s 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 first from activated carbon
are more likely to be responsible for biological activity.
Nonetheless, as indicated previously, size is not the sole determinant of ad-
sorption. 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 flavonoids, some of which are known
to be psychoactive. Apigenin (26a), isolated from S. officinalis, binds weakly
to the benzodiazepine site of the GABAA receptor (IC50 = 30 µM).198 Other
flavones isolated from plants have shown much greater potency at this site;
27 binds with affinity comparable to diazepam (K i = 6 nM).199 Interestingly,
some flavonoids appear to share the anxiolytic properties of benzodiazepines,
80 CHAPTER 2. ISOLATION.
without the amnesic200 and sedating201 qualities.
There has been no published report of flavonoids from S. divinorum. However,
in unpublished work, Valdés isolated the known apigenin derivative 5-hydroxy-
7,4’-dimethoxyflavone (26b)202 from S. divinorum in 20 mg/kg yield.203 This
compound has been isolated from several Salvia species; identity was con-
firmed by comparison of 1H and 13C NMR, IR, MS(CI) and mp with literature
values.202 In other recent unpublished work, Claudio Medana detected a com-
pound whose ESIMS data were consistent with quercitrin (23) or a stereoiso-
mer 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 effects;206 if so, this might conceivably con-
tribute to the known effects of S. divinorum on gastric motility.207 However,
quercetin’s gastrointestinal effects remain in dispute.208
O
O
2'2'O
O
HO
HO
OHO
O
OR
RO
OH R
HMe
26a26b 27
OOH
HO O
OH
OOH
OHOH
O
OH
23
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 plant’s effects. The
binding affinity of 26a is more than three orders of magnitude below that of
1a.23 Moreover, Valdés isolated 1a in two orders of magnitude greater yield
than 26b.
In other plants, however, where flavonoids or other strongly adsorbed com-
pounds are responsible for activity, the above procedure would clearly be in-
adequate. Nonetheless, this obstacle is readily overcome. If bioassay-guided
fractionation showed loss of activity in the decolourised extract, further elu-
tion with stronger eluents as outlined in Section 2.1.3.3 could be used. The
2.1. ISOLATION PROCEDURE. 81
active fraction would then be greatly simplified 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 adsor-
bents 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 “ac-
tivated earths”) are widely used, for instance, to decolourise cooking oils.209
Although not as effective 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 confirmed that Tonsil
was selected in order to decolourise the crude extract.211 He reports that the
procedure was highly effective: 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 CHAPTER 2. ISOLATION.
0 100 µm
Figure 2.6: SEM image of salvinorin A crystals (blade morphology).39
2.1.4 Separation of Terpenoids.
2.1.4.1 Crystallisation of 1a.
TLC indicated that the major component of the decolourised extract (Sec-
tion 2.1.3.4 on page 77) was 1a, which was obtained as fine colourless needles
by crystallisation from MeOH. 1H NMR showed other trace components � 1%
each. Another crystallisation of the mother liquor gave additional 1a. The re-
sulting mother liquor was no longer predominantly 1a by TLC. In other tests,
EtOH was also an effective recrystallisation solvent (Lee et al subsequently
used acetone).137
Serendipitously, trituration in Et2O also proved effective for isolation of 1a
from the decolourised extract. In early tests, a chromatographic fraction rich in
1a was dissolved in Et2O for transfer; an amorphous, flaky precipitate formed
instantly. The filtrate proved, on further chromatography, to be devoid of
1a. The precipitate consisted of 1a which, despite its inferior appearance, was
highly pure by 1H NMR. Thus, trituration in cold Et2O represents a promising
alternative to recrystallisation from boiling alcohols – easier, safer and (more
significantly) much faster. Moreover, the risk of decomposition in heated solu-
tions (Section 2.1.1 on page 67) is avoided. This procedure, which was unfortu-
2.1. ISOLATION PROCEDURE. 83
nately given insufficient emphasis in our initial publication,215 clearly deserves
further exploration.
The fine needles of 1a were examined by scanning electron microscopy (SEM)
(Figure 2.6). Shorter prisms were also observed, as well as plates or flakes (Fig-
ure 2.7), and close examination of crystals showed small defective, semicrys-
talline particles adhering.
84 CHAPTER 2. ISOLATION.
0 200 µm
0 200 µm
0 20µm
0 20µm
0 20 µm 0 20 µm
Figure 2.7: Other salvinorin A crystal morphologies (stereoview).39
2.1. ISOLATION PROCEDURE. 85
2.1.4.2 Chromatography.
O
O
O
H HORO
O O
O
O
O
H HOR1
R2
O O
R2
O
H HR1
O OR3
R
AcH
1a1b
R1
AcAcHH
R2
OAcOHOAcH
1c1d1e1f
28a28b28c29a
R1
OH OH H H
R2
HOHOAcH
R3
HMeHH
H
O
OH
HOH
H 31
OH
30
OH
H
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, mak-
ing a precise count impossible. Valdés 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 CHAPTER 2. ISOLATION.
10 20 30 40 50 60 70 8010
20
30
40
50
60
70
80
hRf (70% Et2O/petrol)
hRf (
10%
ace
ton
e/C
H2C
l 2)
28a
1d
28b
1b
1f
1e
1a
1c
28c
32
29a
30
31
33
Figure 2.9: TLC data of isolated compounds.
resolved by the other. These differences in selectivity proved to be general
for petrol-based versus chlorinated solvent-based systems. After experimen-
tation with various petrol-based systems, Et2O gave similar resolution of 1a
and 1c to EtOAc, but better separation from other compounds. Similarly,
in CH2Cl2-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
1a–1c: salvinorins D–F (1d–1f) and divinatorins A–C (28a–28c, 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).
2.1. ISOLATION PROCEDURE. 87
Dried leaves (860 g)
Crude extract (30.5 g)
Powder, steep in acetone (× 3)
Decolourised extract (5.7 g)
Column on activated carbonacetone to petrol gradient
Recrystallise (MeOH, EtOH)1a (2.6 g)
Column on silica5-50% acetone/CH2Cl2
Series A656 mg
Series B150 mg
Series C359 mg
Series D77 mg
Column 50-80% Et2O/petrol
1c (219 mg) 29a (7 mg)
1a (2.9 g)
1b (13 mg)
Column 70-90% Et2O/petrol
28a (36 mg)
Columns
Triturate (Et2O)
1d (75 mg)
Combine
Column 60-100% Et2O/petrol
Fraction C155 mg
Fraction C2119 mg
Fraction C357 mg
28c (23 mg) 31 (3 mg)
Columns 28b (41 mg)
Column 60-100% Et2O/petrol
Columns
1d (114 mg)
Combine
1e (3 mg) 1f (1 mg)
HPLC60% EtOAc/petrol
Combine
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 differences between the two systems are
88 CHAPTER 2. ISOLATION.
Dried leaves (224 g)
Crude extract (7 g)
Powder, steep in acetone (× 3)
Vacuum filtration on activated carbon50-20% EtOAc/petrol
1a (126 mg)32 (23 mg) 30 (12 mg)33 (1 mg)
Series E97 mg
Series F279 mg
Columns on silica
Recrystallise (MeOH × 2)
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 Et2O/petrol, but trivial in acetone/CH2Cl2. 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 Et2O/petrol and acetone/CH2Cl2, re-
solved most compounds. In deciding which fractions to pool, TLC analysis
in each of the main solvent systems proved invaluable. For particular separa-
tions, 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 filtration
through activated carbon, but the severe tailing exhibited by all compounds
2.1. ISOLATION PROCEDURE. 89
prevented full separation. Since chromatography on silica was ultimately re-
quired, attempts to separate compounds on carbon are not recommended.
2.1.4.3 Crystallisation of new compounds.
As with 1a, trituration in Et2O proved effective for purification of 1d. At-
tempted dissolution of a fraction containing ≈ 33% 1d in boiling Et2O gave
white crystals of the pure compound. Also, as reported by Tidgewell et al,169
trituration in cold MeOH is effective for purification of 1b.
The ease of crystallising these compounds varies widely. Evaporation of a so-
lution of 1b gave crystals irrespective of the solvent. 1a and 1d gave crystals
from some solvents, but an amorphous resin from others (especially CH2Cl2and CHCl3). Although Valdés 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 oc-
casion, 28b gave crystals from cold CH2Cl2 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 reflect their actual pro-
portions 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 CHAPTER 2. ISOLATION.
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 1H
NMR, the final 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/CH2Cl2 gave an improvement, but full recovery was only achieved
using MeOH/CH2Cl2. 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 verified on the basis of 800 MHz 1H NMR (Figure 2.12 on the next page),
along with HSQC, HMBC, DEPT and nOe experiments. These new data
supported the original 1H and 13C assignments in all cases except H-6 and -7.
The HSQC spectrum showed cross peaks from C-6 (δ 38.1 ppm) to multiplets at
δ 1.57 & 1.78 ppm, and from C-7 (δ 18.1 ppm) to δ 1.63 & 2.15 ppm (Figure 2.13
on the facing page). The 800 MHz 1H NMR spectrum also gave well-resolved
first-order multiplets for H-6β, 7α and 11β, which overlap to form a single
multiplet (δ 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. 91
2.742.76 ppm
1.52.02.53.03.54.04.55.05.56.06.57.07.5 ppm
1.113
1.447
1.574
1.642
1.792
2.074
2.163
2.178
2.304
2.504
2.745
3.724
5.138
5.521
6.369
7.384
7.402
2.70
2.96
1.94
1.28
1.51
1.54
5.30
2.27
0.88
1.02
3.25
1.00
0.85
0.82
1.59
1.63
7.397.40 ppm
6.366.37
5.515.525.535.135.145.155.16 ppm 2.50 2.492.512.52
2.142.162.18
2.06
1.781.80 ppm
1.561.581.601.621.641.66 ppm
2.292.312.332.351a
O
O
O
H HO
O
O
O O
Figure 2.12: 1H NMR spectrum of 1a (800 MHz, CDCl3).
1.21.41.61.82.02.22.42.62.83.03.23.43.6 ppm
20
30
40
50
60
C-6
C-7
5.25.45.65.86.06.26.46.66.87.07.27.4
80
90
100
110
120
130
140
Figure 2.13: HSQC spectrum of 1a (800 MHz, CDCl3).
92 CHAPTER 2. ISOLATION.
171.5
171.1 171.1
51.9
35.4
38.1
42.0
171.5
169.9
30.9
20.5
64.075.0202.0
51.9
53.5
16.4
15.2
18.1
43.3
72.0
139.4
143.7
108.3
125.2
51.3
35.4
38.1
42.0
169.9
30.9
20.5
64.075.0202.0
53.5
16.4
15.2
18.1
43.3
72.0
139.4
143.7
108.3
125.2
51.3
5.14 ddt
3.72 s
2.74 dd1.58 td
1.64 tdd
1.79 dt
2.07 dd
2.17-2.14 m
2.16 s
2.31-2.28 m
2.18 br s
1.45 s
1.11 s
2.50 dd
1.57 ddd
5.52 ddd
6.37 dd
7.38 t
7.41 dt
5.14 ddt
3.72 s
2.74 dd1.58 td
1.64 tdd
1.79 dt
2.07 dd
2.17-2.14 m
2.16 s
2.31-2.28 m
2.18 br s
1.45 s
1.11 s
2.50 dd
1.57 ddd
5.52 ddd
6.37 dd
7.38 t
7.41 dt
δC
δH (REVISED ASSIGNMENT)δH
NOESY
δC
δH (REVISED ASSIGNMENT)δH
NOESY
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 verified on the
basis of 1H NMR (Figure 2.15 on the facing page), decoupling, HMQC, HMBC
and DEPT experiments. These data were consistent with the published struc-
ture, 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 confirmed 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-2-
O-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, five known compounds not previously reported from S.
divinorum were isolated. The data and sources used to identify them are given
2.2. STRUCTURE ELUCIDATION. 93
1c
2
4
117
O11
O
O
H HO
O
19
20
O
O
O O
4.7
4.65.6
5.6 5.55.8 5.74.26.2
2.1 1.13.17.18.12.2 1.2
ppm234567
0.670.70 0.88
0.781.00
1.973.00
1.031.00
5.062.95
1.144.43
1.17 4.54
062
.7
047.
3
22
1.2
730.
2
617.
1
612
.1
Figure 2.15: 1H NMR spectrum of 1c (400 MHz, CDCl3).
37.2
38.1
165.7
51.8
132.4
64.1 69.2
51.8
170.521.1
21.8
37.0
18.4
44.2
71.8
139.4
143.9
108.4
171.4
125.4
52.6
15.7
142.6
20.7169.8
37.2
38.1
165.7
51.8
132.4
64.1 69.2
51.8
170.521.1
21.8
37.0
18.4
44.2
71.8
139.4
143.9
108.4
171.4
125.4
52.6
15.7
142.6
20.7169.8
5.75 br d
2.12 s
5.54 dd
6.46 dd
3.74 s
1.71 s
1.21 s
2.59 dt
1.23 td
1.85-1.75 m
2.17-2.10 m
2.13 dd1.49 br s
2.48 dd
1.68 dd
5.53 dd
6.41 dd
7.42 t
7.45 m
2.04 s
5.75 br d
2.12 s
5.54 dd
6.46 dd
3.74 s
1.71 s
1.21 s
2.59 dt
1.23 td
1.85-1.75 m
2.17-2.10 m
2.13 dd1.49 br s
2.48 dd
1.68 dd
5.53 dd
6.41 dd
7.42 t
7.45 m
2.04 s
δC δH
HMBC
δC δH
HMBC
δC/H (REVISED ASSIGNMENT) δC/H (REVISED ASSIGNMENT)
Figure 2.16: Revised NMR assignments for 1c (stereoview).
in Table 2.2 on the following page.
94 CHAPTER 2. ISOLATION.
151617181920212223
δH(ppm)
1.3
1.2
1.4
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
040506070809001011021031041
52.1
02.1
51.1
03.1
53.1
04.1
54.1
05.1
55.1
06.1
56.1
07.1
57.1
5.8610.9615.9610.0715.0710.1715.1710.2715.271
55.5
05.5
06.5
56.5
07.5
57.5
08.5
δC(ppm)
a) HMQC b) HMBC
c) HMBCd) HMBC
168.5169.5170.5171.5
2.02
2.00
1.98
2.04
2.06
2.08
2.10
2.12
2.14
2.16
2.18
2.20
Figure 2.17: HMQC and HMBC spectra of 1c (400 MHz, CDCl3).
Compound 1H NMR 13C NMR FTIR [α]D HRESIMS EIMS TLC29a sa, d216, 217 cs
29b* sa, d218 d218 d218 d218, 217 d217 cs30 sp,219 d220 d219, 220 d219, 220
31 d221 d221 c32 d,222 sp223 d223 d222 d223
33 d224 d224 d224 d224
* 29a was methylated with CH2N2 for full characterisation.c consistent with calculated value.cs cospotted with authentic material in petrol- and CH2Cl2-based systems.d consistent with published data.sa superimposable with spectrum of authentic material.sp 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 first isolated from Hardwickia pinnata by Misra
et al,225 and the structure (including absolute stereochemistry) elucidated by
2.2. STRUCTURE ELUCIDATION. 95
O
H H
O OR
R
HMe
29a29b
Figure 2.18: Single-crystal X-ray structure of 29a (stereoview).
spectroscopic methods and extensive degradation studies.226, 227 Recently, X-
ray crystallography has confirmed 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. Interest-
ingly, its enantiomer (+)-hardwickiic acid (ent-29a) has been isolated from
Copaifera officinalis231 and related species.232 For definitive identification, au-
thentic 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 1H NMR spectrum, but
the opposite sign of optical rotation.
96 CHAPTER 2. ISOLATION.
OH30
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 benefits of green-yellow vegetables.234 30
also exhibits antiplasmodial,235 antimycobacterial236, 237 and antiteratogenic238
activities.
2.2.4 Known Triterpenoids.
2.2.4.1 Oleanolic Acid (31).
H
O
OH
HOH
H 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-inflammatory, hepatoprotective,
gastroprotective, cardiovascular, hypolipidemic and immunoregulatory effects.
Of more significance is anti-tumour activity, which appears to occur via several
independent mechanisms (inhibition of tumourigenesis, inhibition of tumour
promotion, and induction of tumour cell differentiation).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
2.2. STRUCTURE ELUCIDATION. 97
many other pathogens: herpes simplex virus,244 a range of bacteria245, 246 (in-
cluding mycobacteria)247 and Leishmania spp,248 as well as being an insect
antifeedant.249
2.2.4.2 Presqualene Alcohol (32) and Peplusol (33).
OH
H
32
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 pe-
plus by Giner et al.224 In a paper submitted a year later, Connolly et al
reported the same compound from E. lateriflora, naming it anhydrobisfar-
nesol.251 The earlier paper by Giner et al was mentioned, but only in a foot-
note; priority was claimed252 based on a preliminary communication.253 That
communication, however, contained no experimental or characterisation data,
and the absolute configuration was not determined even in the full paper.
Additionally, the ambiguous semisystematic name anhydrobisfarnesol leaves
regio- and stereochemistry unspecified. Thus, Giner et al clearly have priority,
and their name is used here. Connolly and co-workers confirmed 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 fifteen years earlier. Connolly has proposed 33 as a
biosynthetic precursor to 32 itself.252
98 CHAPTER 2. ISOLATION.
It is interesting to note that 33 and 32 are approximately equal in size (22-
carbon chain) to the carotenoid 22 (p.71, 24 carbons), but were easily eluted
from activated carbon in the above chromatography procedure. Since, un-
like 22, these compounds are not conjugated, this again confirms the strong
influence of planarity on adsorption.
2.2.5 Salvinorins D-F (1d-1f).
2.2.5.1 Salvinorins D and E (1d and 1e).
Inspection of the 1H NMR spectra of 1d-1f suggested that they were deriva-
tives of 1c. Compared to 1c, the 1H 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 (δ 2.01 in 1d; δ 1.94 in 1e) which exchanged with D2O. The presence of
a hydroxyl group was confirmed by IR spectroscopy (3475 cm−1 in 1d; 3510
cm−1 in 1e). HRESIMS established the molecular formula of both compounds
as C23H28O8, 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 upfield from δ 5.55 to δ 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 δ 5.76 to δ 4.46. Irra-
diation 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 ap-
parent 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
2.2. STRUCTURE ELUCIDATION. 99
1d
22
3344
55
11
O
O
O
H HOHO
O
O O
ppm234567
0.830.85 0.92
0.870.98
0.991.00
3.002.00
4.980.97
1.054.070.99
4.43
062
.7
837.
3
051.
2
686
.1
52
2.1
1.21.3 1.1
1.61.82.02.2 2.2
2.6 2.5
5.55.65.7
4.5 4.4
6.46.56.6
7.47.5
Figure 2.22: 1H NMR spectrum of 1d (400 MHz, CDCl3).
HMBC crosspeak with the acetate protons (see Figure 2.27 on page 102).
100 CHAPTER 2. ISOLATION.
5.70 dt
2.15 s
4.44 ddd
2.01 br d
6.54 dd
3.74 s
1.69 s
1.22 s
2.56 dt
1.20 td
1.78 dtd
2.17-2.09 m
2.13 dd1.42 br s
2.54 dd
1.64 ddd
5.53 dd
6.40 dd
7.42 t
7.44 s
5.70 dt
2.15 s
4.44 ddd
2.01 br d
6.54 dd
3.74 s
1.69 s
1.22 s
2.56 dt
1.20 td
1.78 dtd
2.17-2.09 m
2.13 dd1.42 br s
2.54 dd
1.64 ddd
5.53 dd
6.40 dd
7.42 t
7.44 s
δH δH
δC
HMBC
J (Hz)
2.4 Hz
1.3 Hz
δH δH
δC
HMBC
J (Hz)
2.4 Hz
1.3 Hz
37.0
37.6
166.2
51.7
135.7
68.6 66.5
52.5
171.621.2
21.6
37.0
18.3
43.9
71.6
139.4
143.8
108.4
171.6
125.4
51.8
15.6
141.2
37.0
141.2
37.6
166.2
51.7
135.7
68.6 66.5
52.5
171.621.2
21.6
37.0
18.3
43.9
71.6
139.4
143.8
108.4
171.6
125.4
51.8
15.6
Figure 2.23: NMR assignments and key 2D correlations for 1d (stereoview).
708090100110120130140
5.0
4.5
5.5
6.0
6.5
7.0
7.5
1520253035404550
1.6
1.4
1.2
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
δC(ppm) δC(ppm)
δH(ppm)
δH(ppm)
Figure 2.24: HMQC spectrum of 1d (400 MHz, CDCl3).
2.2. STRUCTURE ELUCIDATION. 101
37 363839404142434445
5.50
5.45
5.55
5.60
5.65
5.70
5.75
20406080100120140160
2.5
2.0
1.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
134135136137138139140141142143
4.36
4.40
4.44
4.48
4.52
4.56
δH
(ppm)
δC(ppm)
Doublets resulting from direct coupling (1JCH) have been removed digitally.
Figure 2.25: HMBC spectrum of 1d.
102 CHAPTER 2. ISOLATION.
1e
O
O
O
H HOH
O
O
O O
ppm2.02.53.03.54.04.55.05.56.06.57.0
0.840.91 0.93
0.921.03
1.051.06
3.022.08
5.190.870.97
3.121.18
3.16
1.20
1.53
062
.7
92
7.3
071
.2
355.
1
62
7.1
76
4.1
1.3 1.2 1.11.61.71.81.9
2.12.2
2.42.52.6
7.4
6.4
5.45.55.6
4.5 4.4
Figure 2.26: 1H NMR spectrum of 1e (400 MHz, CDCl3).
37.5
37.8
166.0
51.8
131.5
72.3 64.3
54.0
21.9
37.0
18.4
44.4
71.7
139.3
143.9
108.4
171.8
125.8
51.7
16.2
143.4
21.0169.8
37.5
37.8
166.0
51.8
131.5
72.3 64.3
54.0
21.9
37.0
18.4
44.4
71.7
139.3
143.9
108.4
171.8
125.8
51.7
16.2
143.4
21.0169.8
4.46 ddd5.40 dd
6.43 dd
3.73 s
1.72 s
1.47 s
2.52 ddd
1.19 td
1.84 dtd
2.18-2.07 m
2.18-2.07 m1.30 br s
2.46 dd
1.62 dd
5.60 ddd
6.41 dd
7.42 t
7.44 br s
2.17 s
1.94 dd
4.46 ddd5.40 dd
6.43 dd
3.73 s
1.72 s
1.47 s
2.52 ddd
1.19 td
1.84 dtd
2.18-2.07 m
2.18-2.07 m1.30 br s
2.46 dd
1.62 dd
5.60 ddd
6.41 dd
7.42 t
7.44 br s
2.17 s
1.94 dd
2.4 Hz1.6 Hz
2.4 Hz1.6 Hz
δH δH
δC
HMBC
J (Hz)
δC (REVISED ASSIGNMENT)
δH δH
δC
HMBC
J (Hz)
δC (REVISED ASSIGNMENT)
Figure 2.27: NMR assignments and key 2D correlations for 1e (stereoview).
2.2. STRUCTURE ELUCIDATION. 103
708090100110120130140
5.0
4.5
5.5
6.0
6.5
7.0
7.5
20 1525303540455055
1.6
1.5
1.4
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
δH
(ppm) δH
(ppm)
δC(ppm) δ
C(ppm)
Figure 2.28: HMQC spectrum of 1e.
110115125 120135 130145 140150160 155165
5.6
5.4
5.8
6.0
6.2
6.4
6.6
6.8
7.0
7.2
7.4
130 125135140145150155160165170
2.0
1.8
1.6
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
20 1525303540455055606570
1.5
1.4
1.3
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
δC(ppm)
δH
(ppm)
Doublets resulting from direct coupling (1JCH) have been removed digitally.
Figure 2.29: HMBC spectrum of 1e.
104 CHAPTER 2. ISOLATION.
2.2.5.2 Interconversion of 1c-1e via Diol 1h.
a
O
O
O
H HOR1
R2O
O O
a
a
b
R2 R1
1c Ac Ac
1d H Ac
1e Ac H
1h H H
a) Ac2O, pyridine, DMAP b) Na2CO3, CH2Cl2/MeOH (1:1), rt.
Scheme 2.1: Interconversion of 1c-1h.
ppm234567
0.780.78 0.83
0.820.95
1.011.03
3.002.06
1.952.14
1.153.16
3.18 0.991.31
062
.7
13
7.3
20
7.1
306.
1
67
4.1
1.101.151.20
1.551.601.651.701.751.801.851.90
4.24.34.4
(+D2O)
(+D2O)
(+D2O)
2.052.102.152.202.252.302.352.402.452.50
5.7 5.6
6.46.5
7.47.5
Figure 2.30: 1H NMR spectrum of 1h (400 MHz, CDCl3).
To confirm the structures of 1c-1e, diol 1h was prepared by deacetylation of
1c. Tidgewell et al’s standard conditions for deacetylation of 1a (Na2CO3 in
minimal MeOH)169 caused extensive epimerisation (see Section 3.2.4). How-
ever, epimerisation could be suppressed by addition of CH2Cl2, affording 1h
2.2. STRUCTURE ELUCIDATION. 105
2.40 -2.30 m
2.40 -2.30 m
37.5
37.5
166.6
51.7
135.3
69.6 65.6
54.1
22.1
37.0
18.4
44.4
71.8
139.3
143.9
108.4
172.0
125.8
51.8
16.3
142.2
37.5
37.5
166.6
51.7
135.3
69.6 65.6
54.1
22.1
37.0
18.4
44.4
71.8
139.3
143.9
108.4
172.0
125.8
51.8
16.3
142.2
4.32 br d4.28 dd
6.48 dd
3.73 s
1.70 s
1.47 s
2.50-2.45 m
1.16 tdd
1.82 dtd
2.09 dq
2.14-2.10 m1.22 d
2.49 dd
1.60 ddd
5.60 dd
6.40 dd
7.42 t
7.43 br s
2.40-2.30 m4.32 br d
4.28 dd
6.48 dd
3.73 s
1.70 s
1.47 s
2.50-2.45 m
1.16 tdd
1.82 dtd
2.09 dq
2.14-2.10 m1.22 d
2.49 dd
1.60 ddd
5.60 dd
6.40 dd
7.42 t
7.43 br s
2.40-2.30 m
δH δH
δC
HMBCNOESY
J (Hz)
2.4 Hz2.4 Hz
δH δH
δC
HMBCNOESY
J (Hz)
Figure 2.31: NMR assignments and key 2D correlations for 1h (stereoview).
30 10507090110130
2.5
2.0
1.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
140145150155160165170
2.0
1.8
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
a) HMQC b) HMBC
δC(ppm)
δH(ppm)
Figure 2.32: HMQC and HMBC spectra of 1h.
1.11.21.31.41.51.61.71.8
2.1
2.2
2.3
2.4
2.5
1.21.31.41.51.61.71.81.92.02.12.22.32.42.5 ppm
4.34.44.54.64.74.84.95.05.15.25.35.45.55.6
nOe effectdirect coupling
Figure 2.33: NOESY spectrum of 1h.
from 1c or 1d in up to 86% yield. K2CO3 was equally effective.
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 CHAPTER 2. ISOLATION.
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 D2O exchange. The relative stereochemistry was confirmed
using NOESY data (Figure 2.33 on the previous page). H-1 showed cross
peaks to H-2, -10 and -11α, while H-8 correlated to H-10 and -11β, and H-12
correlated to H-20 (Figure 2.31 on the preceding page). This confirms the con-
figurations 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 verified by acetylation, which proceeded
smoothly to give 1c, identical in all respects with the isolated material - TLC
cospot, 1H and 13C NMR, IR, and optical rotation (Scheme 2.1 on page 104).
Insufficient 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 acety-
lated. 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, confirming 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 insep-
arable 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, Valdés detected traces of what appeared to be 1d,
1e, and 1h in S. divinorum;255 while we subsequently confirmed 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 1c–1e. Alternatively, current iso-
lation 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. 107
MeOH/CH2Cl2 was required to achieve satisfactory recovery.
A sample of 1h was supplied to Claudio Medana for LCMS comparison with
the crude extract. This proved difficult, 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 CH2Cl2 gradient. Under these conditions, a crude
CH2Cl2 extract showed no detectable 1h.256
2.2.5.3 Salvinorin F (1f).
The molecular formula of compound 1f, C21H26O6, was established by HRES-
IMS. In contrast to 1d and 1e, 1H NMR (Figure 2.34 on the next page) showed
no acetyl peak, only one oxymethine signal, and a diastereotopic methylene
(δ 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-2α and H-2β were distin-
guished by their coupling constants: molecular modelling257 predicted J 1,2β
= 5.4 Hz (observed: 5.5 Hz) and J 1,2α = 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-2β also showed the
expected HMBC cross peaks to H-4 and -10 (Figure 2.37 on page 109).
108 CHAPTER 2. ISOLATION.
1f
22
3344
101011
O
O
O
H HOH
O O
ppm234567
0.820.85 0.97
0.901.01
1.063.02 1.01
2.08
1.05
2.161.023.07
1.11
3.191.04
1.31
1.28
06
2.7
027
.3
655.
1
707
.1
674
.1
1.3 1.2 1.1
1.61.8 1.72.2 2.1
2.4 2.32.6 2.5
4.5 4.44.65.55.65.7
6.46.56.66.7
7.47.5
Figure 2.34: 1H NMR spectrum of 1f (400 MHz, CDCl3).
2.2. STRUCTURE ELUCIDATION. 109
5.5 Hz
1.1 Hz
5.5 Hz
1.1 Hz
37.7
36.6
166.9
51.5
133.4
38.0 63.9
54.8
21.6
37.3
18.6
44.4
71.7
139.3
143.8
108.4
172.1
125.9
52.2
16.4
140.6
37.7
36.6
166.9
51.5
133.4
38.0 63.9
54.8
21.6
37.3
18.6
44.4
71.7
139.3
143.8
108.4
172.1
125.9
52.2
16.4
140.6
4.51 br dd2.60 ddd
6.67 ddd
3.72 s
1.71 s
1.48 s
2.53 dt
1.18 tdd
1.82 dtd
2.17-2.08 m
2.17-2.08 m1.25 br s
2.46 dd
1.62 ddd
5.60 ddd
6.41 dd
7.42 t
7.44 br s
1.29 dd
2.35 ddt
4.51 br dd2.60 ddd
6.67 ddd
3.72 s
1.71 s
1.48 s
2.53 dt
1.18 tdd
1.82 dtd
2.17-2.08 m
2.17-2.08 m1.25 br s
2.46 dd
1.62 ddd
5.60 ddd
6.41 dd
7.42 t
7.44 br s
1.29 dd
2.35 ddt
δH δH
δC
HMBC
J (Hz)
δH δH
δC
HMBC
J (Hz)
Figure 2.35: NMR assignments and key 2D correlations for 1f (stereoview).
708090100110120130140
5.0
4.5
5.5
6.0
6.5
7.0
7.5
20 1525303540455055
1.6
1.4
1.2
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
3.4
3.6
3.8
δH
(ppm)
δC(ppm)
Figure 2.36: HMQC spectrum of 1f.
110120130140150160
2.5
2.0
1.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
20 1525303540455055606570
1.4
1.3
1.2
1.1
1.5
1.6
1.7
1.8
1.9
2.0
2.1
2.2
2.3
2.4
2.5
δH
(ppm)
δC(ppm)
Doublets resulting from direct coupling (1JCH) have been removed digitally.
Figure 2.37: HMBC spectrum of 1f.
110 CHAPTER 2. ISOLATION.
2.2.6 Divinatorins A-C (28a-28c).
2.2.6.1 Structure Elucidation.
28a28b28b29a
R1
OH OH H H
R2
O
H HR1
18
R2
HOHOAcH
R3
HMeHHO OR3
Figure 2.38: Divinatorins A–C (28a-28c) and hardwickiic acid (29a).
28a
O
H H
O OH
OH
3
3.2
1.1
2.1
2.1
4.2
1.1
1.1
3.1
1
1
0.92
0.99
0.87
0.88
1234567 ppm
7.27.3
6.886.906.92
6.256.26
4.474.484.494.504.51
2.32.42.52.61.81.92.02.1
1.661.681.701.72
1.21.31.41.51.6
Figure 2.39: 1H NMR spectrum of 28a (400 MHz, CDCl3).
2.2. STRUCTURE ELUCIDATION. 111
171.8
140.8
142.8
49.037.1
39.7
39.1
38.1
38.6
37.427.4
21.4
19.8
18.2
64.715.7
138.4
125.2
136.2
110.9
171.8
140.8
142.8
49.037.1
39.7
39.1
38.1
38.6
37.427.4
21.4
19.8
18.2
64.715.7
138.4
125.2
136.2
110.9
1.67 ddd2.05 ddd
4.49 br d
2.56 ddd
6.90 ddd
1.64 s
1.15 s
1.47-1.42 m1.23-1.17 m
1.60-1.54 m
1.60-1.54 m
2.43-2.33 dt
1.45 br s
6.25 dt
7.36 t
7.20 td
0.84 d2.43-2.33 m
2.56 ddd
2.43-2.33 m
2.34 td
1.85 ddd1.67 ddd
2.05 ddd 2.34 td
1.85 ddd4.49 br d
6.90 ddd
1.64 s
1.15 s
1.47-1.42 m1.23-1.17 m
1.60-1.54 m
1.60-1.54 m
2.43-2.33 dt
1.45 br s
6.25 dt
7.36 t
7.20 td
0.84 d
4.9 Hz
2.7 Hz4.8 Hz
δH δH
δC
HMBCNOESY
J (Hz)
4.9 Hz
2.7 Hz4.8 Hz
δH δH
δC
HMBCNOESY
J (Hz)
Figure 2.40: NMR assignments and key 2D correlations for 28a (stereoview).
2030405060708090100110120130140 ppm
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
Figure 2.41: HMBC spectrum of 28a.
The structures of 28a-28c were elucidated mainly by NMR spectroscopy (1H,13C, DEPT, HMQC, HMBC, COSY, and NOESY). The 1H NMR spectra
suggested that they were derivatives of hardwickiic acid (29a). The molecular
formula of 28a (C20H28O4 by HRESIMS) differed from that of 29a by one
oxygen atom, suggesting the presence of a hydroxy group, which was confirmed
112 CHAPTER 2. ISOLATION.
0.81.01.21.41.61.82.02.22.4 ppm
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
2.1
2.2
2.3
2.4
nOe effectdirect coupling
1.11.21.31.41.51.61.71.81.92.02.12.22.32.42.52.6 ppm
4.47
4.48
4.49
4.50
4.51
4.52
Figure 2.42: NOESY spectrum of 28a.
by a strong IR absorption at 3392 cm−1. This was located at C-1 based on 1H
NMR: the oxymethine at δ 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
1α configuration was confirmed by an H-1 to H-11 crosspeak in the NOESY
spectrum (Figure 2.42). NOESY cross peaks also permitted stereospecific
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 confirm the relative
configuration of all stereogenic centres.
2.2. STRUCTURE ELUCIDATION. 113
4.1
0.92
4.6
2.9
3.1
1
3
0.93
0.93
3
1
0.89
0.81
0.92
0.76
0.771234567 ppm
7.27.3
6.646.666.68
6.246.26 4.45 3.853.4
2.32.42.5
1.81.92.02.1
1.551.60
28b
OH
O
H HOH
O O
Figure 2.43: 1H NMR spectrum of 28b (400 MHz, CDCl3).
167.3
141.4
142.8
51.3
44.848.7
39.1
38.8
38.0
38.0
37.121.9
21.4
20.9
18.2
64.3 63.9
138.4
124.9
133.2
110.8
167.3
141.4
142.8
51.3
44.848.7
39.1
38.8
38.0
38.0
37.121.9
21.4
20.9
18.2
64.3 63.9
138.4
124.9
133.2
110.8
3.84 dd4.46 dq4.46 dq
6.65 ddd
2.53 ddd
3.71 s
1.66 s
1.18 s
1.85 dq1.19-1.13 m
1.64-1.49 m
1.64-1.49 m
2.36 dt
1.49 br s
1.49 br s
1.45 br s
6.25 dd
7.35 t
7.20 dd
3.38 dd
2.42 dddd
1.90 ddd
2.31 ddt
1.77 ddd
3.84 dd
1.49 br s
3.38 dd
2.42 dddd
1.77 ddd2.08 dddd
6.65 ddd
2.53 ddd
3.71 s
1.66 s
1.18 s
1.85 dq1.19-1.13 m
1.64-1.49 m
1.64-1.49 m
2.36 dt
1.49 br s
1.45 br s
6.25 dd
7.35 t
7.20 dd
1.90 ddd
2.31 ddt
2.08 dddd
5.1 Hz
2.8 Hz4.9 Hz
δH δH
δC
HMBCNOESY
J (Hz)
5.1 Hz
2.8 Hz4.9 Hz
δH δH
δC
HMBCNOESY
J (Hz)
Figure 2.44: NMR assignments and key 2D correlations for 28b (stereoview).
The 1H NMR spectrum of 28b suggested a methyl ester (δ 3.71) with two hy-
droxy groups (δ 1.49, 2H, D2O-exchangeable). IR showed a strong absorption
at 3434 cm−1. The molecular formula, C21H30O5 (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 CHAPTER 2. ISOLATION.
2030405060708090100110120130 ppm
3.4
3.5
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
Figure 2.45: HMBC spectrum of 28b.
1.51.61.71.81.92.02.12.22.32.4
1.2
1.3
1.4
1.5
1.6
1.7
nOe effectdirect coupling
1.21.31.41.51.61.71.81.92.02.12.22.32.42.52.6 ppm
3.4
3.5
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
4.5
Figure 2.46: NOESY spectrum of 28b.
plings of the oxymethylene signals (δ 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, confirming the configuration at these
centres (Figure 2.46). Again, NOESY data permitted full assignment of the
diastereotopic H-11, -12 and -17 methylene protons.
The 1H NMR spectrum of 28c showed an acetyl methyl signal (δ 2.03). Com-
2.2. STRUCTURE ELUCIDATION. 115
28c
O
O
H H
O OH
O
2.9
1.2
3.8
3.3
5.4
3
2
2.1
0.99
1
1
0.87
1
0.81
0.8
1234567 ppm
38.0
72.1
30.2
7.27.3
6.876.886.896.906.91
6.286.29
4.25 3.753.80
2.22.32.42.5
1.11.2
1.41.51.61.71.8
Figure 2.47: 1H NMR spectrum of 28c (400 MHz, CDCl3).
171.9
171.2
141.2
142.8
40.946.8
38.4
38.9
27.4
35.2
37.422.3
20.5
21.0
171.2
21.0
19.0
18.3
17.066.1
138.5
110.9
140.3
125.2
171.9
141.2
142.8
40.946.8
38.4
38.9
27.4
35.2
37.422.3
20.5
19.0
18.3
17.066.1
138.5
110.9
140.3
125.2
4.26 dd
1.64 ddd
1.82-1.67 m1.82-1.67 m
6.89 dd
2.24-2.16 m2.24-2.16 m
2.03 s
1.27 s
0.83 s
1.82-1.67 m1.15 td
1.54-1.44 m
1.82-1.67 m
1.54-1.44 m2.53 dt
6.28 dd
7.35 t
7.22 dd
3.79 dd
4.26 dd
2.03 s
3.79 dd
2.40 td1.64 ddd
2.40 td
2.35 dt
6.89 dd
1.27 s
0.83 s
1.82-1.67 m1.15 td
1.54-1.44 m
1.82-1.67 m
1.54-1.44 m2.53 dt
1.42 br d 1.42 br d
6.28 dd
7.35 t
7.22 dd
2.35 dt
δH δH
δC
HMBCNOESY
δH δH
δC
HMBCNOESY
Figure 2.48: NMR assignments and key 2D correlations for 28c (stereoview).
116 CHAPTER 2. ISOLATION.
203040506070 ppm
1.0
1.5
2.0
2.5
3.0
3.5
4.0
Figure 2.49: HMBC spectrum of 28c.
0.80.91.01.11.21.31.41.51.61.71.8
3.8
3.9
4.0
4.1
4.2
4.3
1.31.41.51.61.71.81.92.02.12.22.32.42.5 ppm
0.9
1.0
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
nOe effectdirect coupling
Figure 2.50: NOESY spectrum of 28c.
pared to 28b, the C-1 oxymethine was absent, while the oxymethylene signals
were shifted downfield to δ 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, C22H30O5 (HRESIMS). As with 28b,
NOESY data (Figure 2.50) confirmed the configuration of all stereogenic cen-
tres and permitted full assignment of the H-11, -12 and -17 methylene protons.
The neoclerodane absolute stereochemistry shown for the salvinorins and div-
inatorins is common to all clerodanes isolated from the Lamiaceae,26 including
2.2. STRUCTURE ELUCIDATION. 117
1a24, 25 and 29a.226
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 identification:
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 divinato-
rum. 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 fluent in Latin, and
was intended to mean “of the seers.”4 According to classicist Dr Parshia Lee-
Stecum, 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 divino-
rum: “of the prophets/soothsayers/seers”.
Hofmann’s alternate rendering, derived from the medieval Latin noun “div-
inator”, would have had the same meaning in the medieval period. However,
this word was not used in antiquity, and “divinatorum” would have been in-
terpreted as a form of the verb divino (I foresee):
To someone of ... Virgil’s, Ovid’s or even Tacitus’ time, Salvia div-
inatorum would thus mean “salvia of those who have been foretold”
...
Nonetheless, in honour of Hofmann’s key role in the scientific identification of
the plant, compounds 28a-28c were named divinatorins.
118 CHAPTER 2. ISOLATION.
2.2.7 Subsequent isolations.
1g
O
O
O
HO
O
O O
O
O
O
O
H HOO
O
O O
HHO
R
RHO
OMe
OMe
34a
34b
R
28d
28e
R
CH2OAc
CHO
R
O
H HOH
O O
4
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 definitively confirmed by X-ray crystallography.
Chapter 3
Synthesis.
In order to study the structure-activity relationships of salvinorin A (1a) at
opioid receptors, various synthetic modifications were made to the compound.
3.1 Known derivatives.
1a
O
O
O
H HOO
O
O O
35
O
OH
O
H HOO
O
O O
O
O
O
H HOR1
R2O
O O
R2 R1
36h H H
36e Ac H
36d H Ac
a bc
d
36c Ac Acc
Conditions: a) iBu2AlH, THF, –78◦C, 25 min, 65% (81% borsm); b) NaBH4, EtOH, 40◦C, 56%; c) Ac2O, pyridine, DMAP, rt; d) i) trimethyl orthoacetate, AcOH, 105 ◦C; ii)AcOH, cat. HCl, THF/H2O, rt, 74% over 2 steps.
Scheme 3.1: Preparation of known derivatives.
Several known derivatives of 1a were prepared by published procedures with
minor modifications. Lactol 35 was prepared using iBu2AlH following Brown’s
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 Valdés from 1a using 0.6 equivalents of
NaBH4 in iPrOH;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 purification of the starting diol 36e itself.
Using Ac2O/pyridine/DMAP, direct formation of dihydrosalvinorin C (36c)
from 36h occurs in negligible yield, even after prolonged reflux.262 Valdés et
al’s indirect route via the orthoacetate and 36d27 was therefore used. Satis-
factory purification 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 Condi-
tions.
3.2.1 Previous Reports.
In the NaBH4 reduction of 1a described above, Valdés 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 undeter-
mined stereoisomer” of 1a. Subsequently, Brown identified 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 Valdés et al characterised diol 38h.27
No basis was given for this proposed structure, however.
1a
O
O
O
H HOO
O
O O
88
O
O
O
H HOH
HO
O O
88
O
O
O
H HOHO
O O
37a
88
O
O
O
H HOO
O
O O
37b
88
O
O
O
H HOH
O O
O
O
38h 38e
NaHCO3, DMF
Ac2O / pyr
NaBH4
Ac2O / pyr
PCC
B-
ROH
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 Ac2O/pyridine gave 8-epi-salvinorin A (37a).
The structures of 37a and 37b were elucidated using 2D NMR (HMQC,
122 CHAPTER 3. SYNTHESIS.
1a 11.6, 3.1 37a 5.0, 2.21b 11.7, 3.0 37b 4.8, 2.2
Table 3.1: Coupling constants (Hz) at H-8 for 1a, 1b and 8-epimers.
HMBC, COSY and NOESY). The configuration of H-8 was apparent from the
loss of the trans-diaxial coupling constant, establishing an equatorial configu-
ration (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-11β 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 car-
bonyl absorption reported for pseudoboat δ–lactones (≈ 1760 cm−1).268 How-
ever, those values were obtained in solution, and are subject to solvent effects;
recording the spectrum as a mull gives typical carbonyl absorptions ( ≤ 1730
cm−1).269 The spectra of 37a and 37b were recorded as neat films.
Our initial publication270 contained characterisation data for 37a but not for
37b. Harding et al82 and Lee et al271 subsequently published 1H 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 refluxing MeOH/THF gave 1b
in quantitative yield,272 but in my hands 37b was the major product, as had
been the case with NaHCO3. Valdés noted the same result.273 Additionally,
the epimers were not resolved on silica using Brown’s solvent system (3%
3.2. EPIMERISATION AT C-8 UNDER BASIC CONDITIONS. 123
5.25 dd (12.0, 2.2 Hz)
5.25 dd (12.0, 2.2 Hz)
1.50 dd (15.0, 12.0 Hz)
2.24 br s2.24 br s
2.36 dd (15.0, 2.2 Hz)1.50 dd (15.0, 12.0 Hz)
2.36 dd (15.0, 2.2 Hz)
2.45 dd(5.0, 2.2 Hz)
2.45 dd(5.0, 2.2 Hz)
δH δH
NOESY
δH δH
NOESY
Figure 3.1: Key NOESY correlations for 37a (stereoview).
MeOH/CH2Cl2). Petrol-based systems resolved the epimers, but gave poor
recoveries: some precipitation occurred when loading in these systems. Thus,
effective separation required elution with petrol-based systems for separation,
followed by stripping with MeOH/CH2Cl2 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 Na2CO3 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 Na2CO3 appears to cause less epimerisation than the much
weaker base NaHCO3 (the conjugate acid of Na2CO3). When Tidgewell et
al’s procedure was repeated in oxygen-free MeOH, the supernatant contained
almost pure 37b, which was obtained as an amber resin in 23% yield by evap-
oration. 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 confirm that as expected, epimerisation does occur
124 CHAPTER 3. SYNTHESIS.
under Tidgewell et al’s conditions, but is usually not apparent due to decom-
position in oxygenated solution. Nonetheless, the yield of 1b is not affected, 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 CH2Cl2, 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 differences in MeOH solubility among starting material and
products.
Substituting NaHCO3 for Na2CO3 under Tidgewell et al’s conditions gave no
reaction after one day. Conversely, Na2CO3 in MeOH/CH2Cl2 gave a complex
mixture; the distinctive H-12 signals of 8-epimers were apparent by 1H NMR.
The above deacetylations proceed via base-catalysed solvolysis; the use of po-
lar “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 ba-
sic 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 al’s deacety-
lation conditions to 1c (Na2CO3 in minimal MeOH)169 gave predominantly
3.2. EPIMERISATION AT C-8 UNDER BASIC CONDITIONS. 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 Me2SO, but poorly soluble
in acetone and chloroform, with a tendency to precipitate during preparation
of NMR samples.
O
O
O
H HOR1
R2O
O Oc
b
R2 R1
37c Ac Ac
37d H Ac
37e Ac H
37h H H
b
1c
O
O
O
H HOO
O
O
O O
a
Conditions: a) Na2CO3, MeOH, rt × 4 h. b) Ac2O, pyridine, (± DMAP) c) CDCl3, rtovernight
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 1-
acetate 37d, presumably due to traces of DCl. Filtration of either compound
through a plug of basic Al2O3 also caused this unexpected acetate migration.
Alumina-catalyzed acyl migrations have been reported previously274 in 1,2-
dihydroxy 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 Identification of Epimers.
The epimers gave different colours when visualised with vanillin/H2SO4 in
EtOH. After brief heating, the natural H-8β compounds slowly developed a
pink/purple colour, while the 8α compounds turned blue. These colours were
observed for salvinorins A–E (1a, 1b and 1c–1e) and the diol 1h. Addition-
ally, in each case, the 8-epimer gave a higher Rf in Et2O or EtOAc/petrol.
126 CHAPTER 3. SYNTHESIS.
eluent:1a 37a 1b 1h37b 37h
Et2O Et2O 50% EtOAc/petrol
Figure 3.2: TLC comparison of epimers using vanillin/H2SO4.
These relationships also held for many derivatives to be discussed below; thus,
configuration at C-8 can be confidently inferred from TLC alone. Given the
tendency for salvinorins to epimerise at C-8 under basic conditions, this infor-
mation 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 α-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 configuration of
3.2. EPIMERISATION AT C-8 UNDER BASIC CONDITIONS. 127
H-8 indirectly. This mechanism has been tentatively endorsed by Lee et al.271
O
O
O
H HORO
O O
1010 8899
O
O-
O
ORO
O O
O
O
O
H HORO
O O
B-
H+
Scheme 3.4: Koreeda et al’s 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 definitively established by X-ray
crystallography: columbin (39),265 isocolumbin (40),267 palmarin (42)266 and
jateorin (43).264
O
O
OO
OH
RO
O
OO
OH
R'
88
R
R
H
H
39
40
R R'
H H
H H
H H
41
42
43
44H H
a
aa
O O
O
Conditions: a) ROH, OH−, ∆.
Scheme 3.5: Epimerisation of related natural products with base.
Conclusive evidence against the mechanism can be found in Brown’s own re-
sults, performed under Koreeda’s guidance. Refluxing 1a with KCN in CD3OD
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 CHAPTER 3. SYNTHESIS.
(a weaker base than CO2−3 ).275 Thus their proposed mechanism, in which this
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.
1a
O
O
O
H HOO
O
O O
22 1010 88
O
O
O
D DOHO
O O
D
KCN, CD3OD
45
Scheme 3.6: Brown’s 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 CH2Cl2/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; deacetyla-
tion appeared to have been accompanied by lactone methanolysis (additional
methoxy peaks and an upfield shift of H-12). Since this route is clearly inferior,
the products were not characterised further.
3.3 Simple Derivatives.
3.3.1 Esters (46 and 47).
Treatment of 1b with HCO2H/Ac2O279 in pyridine280 gave formate 46. Treat-
ment in neat HCO2H281 proved too vigorous, giving inseparable byproducts.
NMR assignments of 46 were inferred from the near-identical spectra of 1a.
3.3. SIMPLE DERIVATIVES. 129
46
O
O
O
H HOO
O
O O 47
2211
O
O
O
H HOH
O
O O
O
Br
48
O
O
O
H HOO
O O
Figure 3.3: Ester and ether derivatives.
In the hope of obtaining crystallographic confirmation of the structure and ab-
solute 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 CH2Cl2/Et2O 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 µ opioid agonist,82 for-
mation of the benzyl ether 48 was attempted. Benzyl trichloroacetimidate
and catalytic Me3SiOTf in CH2Cl2283, 284 gave a complex, inseparable mixture
which did not appear to contain 48 by 1H NMR. An alternative route, benzyl
bromide with freshly prepared285 Ag2O in CH2Cl2,286 gave negligible reaction
after 15 hours, despite both reagents being in excess. Addition of Bu4NI287
gave completion in 4 hours. Again a complex mixture of benzylated compounds
resulted, which appeared to contain traces of 48.
Subsequently, Béguin et al reported successful preparation of 48 using BnBr
and Ag2O 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 CHAPTER 3. SYNTHESIS.
3.3.3 17-Deoxy Compounds (49 and 50).
17-Deoxysalvinorin A (49) was synthesised by deoxygenation of lactol 35 using
Et3SiH 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 Et3SiH 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 de-
hydration 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-8β configuration. The furan
substituent may also stabilise the oxonium intermediate through electron do-
nation.
49
6677
881717
O
O
H HOO
O
O O
35
O
OH
O
H HOO
O
O O
O
O
HO
O
O
O O
50
+Et3SiH, BF3•OEt2,
CH2Cl2, 0 °C48% 23%
13C1HHMBC
Et3SiH, Amberlyst 1576%
CH2Cl2
Scheme 3.7: Deoxygenation of lactol 35.
By 1H NMR, the H-17 oxymethylene of 49 appeared as a doublet, δ 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
confirming the configuration 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).
52
OH
O
O
H HOO
O
O O
H
51
O
O
O
H HOO
O
O O
HH13
1aH2, Pd/C H2 (4 atm), Rh/C
59%94%
MeOH MeOH/CH2Cl2
O
O
O
R'
OHR
O
O
R'
OR
O
O
R'
OHR
O
O
+ +
A B C D
H2
cat.
O
O
O
Scheme 3.8: Hydrogenation of 1a and other furanolactones.
O
O
O
OO
OO
O
HH
H
Limonin (53)
OO
O
H
O OAcOAc
H
Montanin C (54)
OO
H
O
H
Teucrin A (55)
O
OH
O
Figure 3.4: Furanolactones 53, 54 and 55.
To explore the role of the furan ring in the effects of 1a, tetrahydrosalvinorin
132 CHAPTER 3. SYNTHESIS.
A (51) was prepared. Valdés had reported15, 23 that catalytic hydrogenation
of 1a over palladium on carbon gave hexahydrosalvinorin A (52) in near-
quantitative yield after 24 hours (Scheme 3.8). Saturation of the furan ring
was accompanied by hydrogenolysis of the lactone. This is typical of the pseu-
dobenzylic bond of furanolactones (A): although high yields of tetrahydro
compounds C have been reported with Pd/C, hexahydro compounds D gen-
erally 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: different groups report very different results for the same sub-
strate under similar conditions (compare the divergent results for 53 and 54).
This may be due to differences 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% H2SO4/MeOH gave complete
conversion to 52 in 10 minutes. The substrate also affects the outcome: differ-
ent compounds sometimes give dramatically different results under identical
conditions (compare 55 and 54 in Ref. 298).
Rhodium on carbon is reputed172, 319, 297 to cause less hydrogenolysis than pal-
ladium catalysts. There have been reports of selective reduction of furanolac-
tones 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 1H NMR, H-12 showed a new cou-
pling to H-13, but its coupling constants to H-11 were scarcely affected, sug-
gesting little change of conformation in the lactone.
Determination of the configuration 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. 133
Catalyst Solvent Time B C D Ref.% % %
30% Pd/C EtOAc 22 0 0 29910% Pd/C MeOH 24 h* 65 298(55)
24 h* 53 20 298(54)14 78 300(54)
8 h 10 76 30135 min 6 89 2681 h 10 84 2681 h 7 85 268
EtOH 30 min 17 56 30220 80 303
45 min 29 67 304AcOH 5 h 50 305
24 h 10 59-69 306(53)12 41 2768 44 276
CH2Cl2 24 h* 48 307(53)dioxane 4 59 308EtOAc 36 h 74 309H2O 30 310
5% Pd/C AcOH 17 h 20 77 311(53)EtOH 6 h 2 51 312(53)
10% Pd/BaSO4 MeOH 22 24 39 31318 20 35 313
EtOH n.s. n.s. 314(55)AcOH 55 314
PtO2 AcOH 5 h 34 10 31520 h 42 <55 3167 h 53 30530 h 30 299
AcOH/dioxane 48 h 88 311(53)96 h 86 311(53)
5% Rh/C AcOH 10 h 42 317(55)EtOAc 48 h 36 318
* = 2 Atm. n.s. = not stated
Table 3.2: Some previously reported furanolactone hydrogenations.
obtained pure, this was not attempted in this case. Since this work, Béguin et
al have synthesised 51 using Pt/C catalyst, albeit in lower yield (36%).125
134 CHAPTER 3. SYNTHESIS.
O
H H
O OR
R
H
Me
ent-29a
ent-29b
Figure 3.5: (+)-Hardwickiic acid (ent-29a).
3.3.5 (+)-Hardwickiic Acid (ent-29a).
(+)-Hardwickiic acid (ent-29a) occurs in commercially-available copaiba bal-
sam. The compound is difficult 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 reflux gave only slow decomposition, but
microwave irradiation on KF/Al2O3320 provided ent-29a in low yield. The
reaction did not proceed in the absence of KF.
3.4 Modification of the Methyl Ester.
3.4.1 Relevant Results from Previous Work.
To explore the pharmacophore of 1a, a selective modification of the C-18
methyl ester was desirable. It was apparent from previous work that hydride
reduction was unsuitable. As discussed above, Valdés and Brown had achieved
selective reductions of the ketone and lactone respectively (Scheme 3.1 on
page 119). Varying the solvent in the iBu2AlH reduction was of no benefit.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. Béguin 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. 135
56
O
OH
O
H HOH
HO
O O 57
OH
OH
O
H HOH
HO
OH 58
O
O
H HOH
OH
OH
1aLiAlH4 LiAlH4
Li/NH3, -100 °C
-78 °Crt
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 conflicting structures for
58; the hybrid structure shown here incorporates those features consistent with
the 1H NMR data. The furan peaks downfield of 6 ppm rule out saturation of
the furan ring. The C-18 oxymethylene δ 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/CD3OD gave deuteration at C-2, -8 and -10, but not
C-4, α to the methyl ester (see Scheme 3.6 on page 128). Clearly none of
these approaches offered 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 first route attempted for methyl ester cleavage was basic hydrolysis. Since
the methyl ester was unaffected by Na2CO3 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 CHAPTER 3. SYNTHESIS.
was enedione 59 (Scheme 3.10). The base-soluble fraction was difficult to
analyze, smearing on TLC and giving a poorly-resolved 1H NMR spectrum.
Surprisingly, at least eight peaks were apparent in the methoxy region. After
methylation with Me3SiCHN2,322 TLC showed only a single spot. 1H NMR
analysis, however, revealed three major compounds (60a, 60b and 60c), which
were separated with difficulty by HPLC. Baseline resolution was not achieved,
necessitating repeated repurification and poor recoveries.
3.4.2.1 Structure Elucidation of the Products.
+1. KOH / MeOH
2. Me3SiCHN2
O
O
O
R'RO
O
O
O
R R'
H H
H H
H H
60a
60b
60c
10 8
59
53%47%
1a
O
O
O
H HOO
O
O O
O
O
O
HOH
O
O O O O
Scheme 3.10: Autoxidation of 1a.
The 1H spectrum of 59 showed two new singlets at δ 6.91 and 6.99 ppm. The
peak at 6.91 showed no COSY or HMQC crosspeak, and exchanged with D2O.
Such strongly deshielded exchangeable peaks are typical of cyclic α-diones,
whose enol tautomers are stabilised by internal H-bonds.323 Consistent with
this, the compound exhibited strong IR absorptions at 3373 and 1651 cm−1
(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 confirmed an extended π 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 vis-
ible wavelengths. The absorptions of α-diones exhibit a bathochromic shift
3.4. MODIFICATION OF THE METHYL ESTER. 137
towards visible wavelengths in basic solution:325, 326, 327 the orange colour ob-
served during formation of 59 in basic MeOH is consistent with this.
22
3344
55
101011
O
O
O
HO
O
1919
2020
H
H
O
O
O
HHO
O
O
O
60aH
59
13C1H
18
O O O O
Figure 3.6: Key HMBC correlations of 59 and 60a.
HRESIMS confirmed 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 β configuration 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 identified 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 confirmed the location of the new methyl
esters at C-1 and 2, although the three esters were not sufficiently 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, confirming the configuration at C-8. HRESIMS
confirmed the molecular formula. The second major product was identified
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 CHAPTER 3. SYNTHESIS.
1c
59
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
20000
400380360340320300280260240220200
1a
O
O
O
H HO
O
O
O O
O
O
O
H HO
O
O
O
O O
O
O
O
HOH
O
O O
1a
1c59
61
ε
λ (nM)
OO
O
Figure 3.7: UV/Visible spectra of 1a, 1c and 59 in MeCN.
again confirmed the molecular formula. Interestingly, although 60a and 60b
cospotted by TLC, they gave the expected colours with vanillin/H2SO4 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 C-
ring configurations 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 configuration. The most plausible
candidate structure was therefore the 10-epimer 60c, since H-10 is not coupled.
Placing a large substituent in an axial configuration would be expected to
affect the conformation of both remaining rings, and hence the chemical shifts
around those rings. By contrast, inversion at C-4 would not be expected
3.4. MODIFICATION OF THE METHYL ESTER. 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 downfield 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 refluxed with KCN in CD3OD, 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 conflict with those published previously. Tidgewell et al169 re-
ported 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 1H NMR spectrum. As shown above, however, the acidic fraction and its
methoxy peaks result from cleavage of the α-hydroxy ketone. When the 1,2-diol
36h23 (Figure 3.8) was refluxed in KOH/MeOH for 30 min, an epimeric mix-
ture at C-8 was recovered in near-quantitative yield, confirming that neither
methyl ester hydrolysis nor lactone methanolysis occur under these conditions.
62
O
O
O
HO
O O
O
O
O
H HOH
HO
O O 36h
Figure 3.8: Diol 36h and proposed autoxidation product 62.
140 CHAPTER 3. SYNTHESIS.
More recently, Lee et al treated 1a with Ba(OH)2 in MeOH, reportedly obtain-
ing 62 in 75% yield.271 The 1H and 13C NMR data quoted for 62 are identical
to those of 59, apart from the omission of the H-6 multiplet at δ 1.77-1.67
ppm; they are evidently the same compound. Their proposed structure 62 is
not consistent with the additional data presented here. Specifically, HRMS es-
tablished the molecular formula as C21H22O7 (59) rather than C21H22O6 (62).
Further, the singlet at 6.91 cannot be attached to C-1, since it exchanges with
D2O, 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 distinc-
tive red shift of α-diones. The presence of the C-1 ketone was confirmed 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 α-hydroxy ketones (“acyloins” or “α-ketols”) to α-diones (“diosphe-
nols”) under basic conditions is well-established.325, 328, 329 The reaction con-
sumes one equivalent of O2, generating H2O2.329 While the autoxidation of
unsubstituted ketones requires stronger bases such as tBuOK, α-hydroxy ke-
tones 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 hy-
droperoxide 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 1H NMR spectrum of the crude product.
Dehydrogenation of 63 to form 59 is unusual. While there have been sev-
eral reports of dehydrogenation of 1,4-diones in alcoholic KOH,331 literature
searches48 revealed no such reaction involving a 4-ketoester. However, α-diones
3.4. MODIFICATION OF THE METHYL ESTER. 141
O
O
O
H HOHO
O
O
O
H HOHO
O2-
O
O
O
H HO
HO
HOOO
O
O
H HOO
O
O
O
H HO
HO
OO
O
O
O
HHO
O
O
HO
O2-OH
1) -OH
O
O
O
H HOO
HOO
H
MeOH/MeO-
HO
59
63
2) O2
O
O
O
H HOO
O2
MeOH
-OH
-OH
MeOH
1a
O
O
O
H HOO
O
O O O O O O
O OO OO O
O O OO
O O 64
Scheme 3.11: Proposed mechanism of the autoxidation.
65
O
O
O
O
HO
O O
H
66
O
O
O
O
HO
O O
H
HO
1b
O
O
O
H HOHO
O O
CrO3
pyridine
-H2O
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 plausi-
ble that 63 should be extremely reactive, and unsurprising that this compound
142 CHAPTER 3. SYNTHESIS.
was not isolated. Consistent with this, Brown’s 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 ob-
taining 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 α-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 effects.328 The reaction was therefore repeated in EtOH, iPrOH andtBuOH. 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 per-
formed 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 refinement 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 com-
pound might provide a route to diol 1h. However, reduction with NaBH4
in EtOH/CH2Cl2 was accompanied by conjugate addition and epimerisation,
giving 8-epi-diol 38h27 in low yield (Scheme 3.13). Attempted Luche reduc-
tion with NaBH4–CeCl3 in MeOH (with335 or without336 sonication) was also
unsuccessful, giving a complex mixture whose unstable major components re-
3.4. MODIFICATION OF THE METHYL ESTER. 143
O
O
O
H HOH
HO
1h59
O
O
O
H HOH
HO
NaBH4 X38h
O O O ONaBH4
CeCl3
Scheme 3.13: Attempted reductions of 59.
tained the characteristic enedione peaks at δ 6.8 and 7.0 ppm. These products
cospotted with the starting material in petrol-based systems, making the reac-
tion difficult to follow, but were resolved by acetone/CH2Cl2 and gave a darker
purple with vanillin. There are precedents for ketones which are smoothly re-
duced by NaBH4 alone, but not in the presence of CeCl3.335 Note that use
of “forcing conditions”337 (reflux for 12 hours) was futile. NaBH4 decomposes
to unreactive tetramethoxyborate within 5 minutes under the standard condi-
tions (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 decomposi-
tion.338 The quality of the CeCl3 used was verified by reduction of 2-cyclohexen-
one,336 which gave 2-cyclohexenol rapidly and quantitatively.
The enols of α-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 ketone–solvent–Ce3+ 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
Et3Si protection was unsuccessful.
144 CHAPTER 3. SYNTHESIS.
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 reac-
tions 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 effective procedures involve
nucleophilic substitution at the alkoxy (or carbinol) carbon, via an SN2 mech-
anism, with the carboxylate as leaving group. This mechanism (BAl2 ester
cleavage) contrasts with the BAc2 mechanism typical of basic hydrolysis, in-
volving attack at the carboxyl carbon (Scheme 3.14).343 The tendencies of
different 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 BAl2 mechanism, while hard nucleophiles such
as F− favour attack at the hard electrophilic carboxyl carbon, via the BAc2
mechanism. Thus soft nucleophiles are less affected by hindrance around the
carboxyl carbon, but more affected by hindered alkyl chains; hence the se-
lectivity for methyl esters. Given the severe hindrance exhibited by the C-18
carboxyl carbon of 1a, the desirability of methyl ester selectivity, and the dis-
astrous effects of hydroxide, the use of soft nucleophiles was clearly preferable.
CH3Nu+
BAl2
OR
O H
HH
Nu-
O-R
O
BAc2
OR
O
OHR
-OO
-OH
+O-R
O
CH3OH
Scheme 3.14: BAl2 and BAc2 ester cleavage mechanisms.
3.4. MODIFICATION OF THE METHYL ESTER. 145
3.4.3.1 Cleavage with Iodide and Cyanide Reagents.
Perhaps the most widely used source for a soft nucleophile is LiI. Early re-
ports 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 SN2 reac-
tion.347 Anhydrous LiI has been reported to give greater selectivity for methyl
esters, but lower yield, than the hydrate.348
1a
O
O
O
H HOO
O
O O 67a
O
O
O
H HOO
O
O OH 67b
O
O
O
H HOHO
O OH
Figure 3.9: O-Demethyl salvinorins A and B.
An initial trial of anhydrous LiI in dry DMF, after refluxing for 24 h, gave a
29% yield of the epimerised acid 67a. The dark brown colour and fishy 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 refluxing
overnight was not improved.
N
2,6-lutidine
O
N
HMPA
N N
O
DMPUDMF
N N
O
NP
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 CHAPTER 3. SYNTHESIS.
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 fishy
smell suggestive of decomposition.
A run at a higher temperature (190 ◦C) gave a different base-soluble product
which was not identified. The NMR spectrum showed an epimeric mixture
lacking the acetyl peaks of 67a. It did not match 67b, however, and was not
acetylated by Ac2O/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 K2CO3 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
3.4. MODIFICATION OF THE METHYL ESTER. 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 nBuLi using a simplified
version of published procedures for LiSMe354 and LiSPr.357 Again, DMPU was
substituted for the carcinogen HMPA.
68
O
O
O
H HOO
O
O O
O
67b
O
O
O
H HOHO
O OH
Ac2O, DMAP
DMPU
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 deacetyla-
tion, 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 Ac2O/DMAP gave 67a in four hours, unfortunately accompanied by the
mixed anhydride 68 (crude 1H NMR of the neutral fraction showed additional
characteristic peaks at δ 2.24 and 2.25 ppm; Scheme 3.15). The anhydrides
proved remarkably stable. Some starting material remained after reflux in
THF/H2O for 1 hour; addition of NaHCO3 was necessary for complete hydrol-
148 CHAPTER 3. SYNTHESIS.
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.
Yield of 67aNu Base Solvent T time Isolated borsm*
◦C h % %LiI - DMF 150 24 29LiI NaOAc 130-150 21 26LiI - DMPU 150 25 27 79LiI - 190 24 0NaI - 150 14 23
NaCN - 90 70 0KCN - 90 27 0
p-MePhSH cat. K2CO3 200 16 0LiSEt - 55 23 73 (after acetylation)
* = based on recovered starting material
Table 3.3: Summary of results - nucleophilic cleavage of 1a methyl ester.
3.4.3.3 Confirmation of Structure of 67a.
The mixed acids 67a streaked on TLC, and reduction by BH3 was also ex-
pected to cause epimerisation, so chromatographic separation was not at-
tempted. The structure of 67a was definitively confirmed by methylation with
CH2N2, giving a mixture of 1a and 37a by 1H NMR. The products cospotted
with authentic material in both petrol– and CH2Cl2–based TLC systems and
gave identical colours with vanillin dip.
The 1H NMR spectrum of 67a lacked the methoxy peak, but was otherwise
near-identical to a mixture of 1a and 37a. One interesting difference was the
first-order H-4 multiplet at 2.80 (dd, J = 5.2, 3.5). The H-4 multiplet of 1a is
slightly non-first 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. 149
3.4.3.4 Subsequent Developments and Discussion.
Since this work, Me3SnOH has been reported to cleave methyl esters selec-
tively 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, Me3SnOH 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 refluxing in
pyridine for 36 hours. The LiI was presumably a hydrate since no drying is
mentioned. If the higher yield proves reproducible, this method offers 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 effects discussed in Section 3.4.3.1 on page 145.
By the same token, the alkylthiolate route has not been optimised. This
route offers the inherent advantage of proceeding at or near room temperature.
Given the accumulation of the basic carboxylate during BAl2 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 purified before use.
On several occasions after the cleavage and acetylation, 2-hydroxyethyl ac-
etate362 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 nBuLi used
(from exposure to moisture) would also have been present in the thiolate. ThenBuLi 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 CHAPTER 3. SYNTHESIS.
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 O-Demethyl-18-deoxysalvinorin A (77).
3.4.4.1 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 reduc-
tions of polyfunctional acids such as 70366 and 71.367 However, some acids
are much less reactive. For example, many aromatic acids such as 72368 re-
quire excess BH3, longer reaction times and higher temperatures.369 In some
extreme cases, refluxing 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 different protocols have been used, but generally borane is added drop-
wise 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 selectiv-
ity 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.
3.4. MODIFICATION OF THE METHYL ESTER. 151
X
OH
X
HO
4 eq, reflux, 4h, 94%
X
HO
OR
OR
OR
OR X
OH
4 eq, reflux, 3h, 77%
O
O
O
O
OHX
XHO
3.5 eq, rt, 48h, 71%
O O
O O
OHX
1.3 eq, 0°C, 30 min, 66%
OHX
1.3 eq, 0°C to rt, 1h, 95%
OXH
H
H
O
H
H
XH
HOH
3 eq, rt, 3 days, 55%
1 eq, rt, 30 min, 80%
XO
H2BH3•THF
Br
OO
OH
X
1 eq, 0 °C, 3h, 86%
XO
HO
70 7169
73 72
74 (R= Me/Et)
75 76
H2
Scheme 3.16: Some previously reported BH3·THF reductions.
Gradual warming to 0 ◦C and addition of excess borane had no effect. 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 first visible signs of H2 evolution. The actual
152 CHAPTER 3. SYNTHESIS.
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.
Refluxing with excess borane gave 78 accompanied by unidentified 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.
67a
O
O
O
H HOO
O
O OH
O
O
O
H ROO
O
OH
BH3•THF (1.3 eq)
dropwise, rt x 5 min,then 55 °C x 90 min
R
H
H
77
78
23%
25%
1H1HCOSY
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 mate-
rial in the TLC system used previously (1% AcOH in 20% acetone/CH2Cl2).
Thus, in earlier experiments, clean formation of the desired product may have
gone undetected, while decomposition of that product may have been mis-
taken 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
reflux 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. 153
3.4.4.2 Structure Elucidation of 77.
Although the H-4 multiplet of 77 (δ 1.89 ppm) was too complex for determi-
nation of coupling constants, the COSY spectrum showed cross peaks with the
diastereotopic H-18 oxymethylene (δ 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), confirming the configuration at C-8. A
strong infrared absorption at 3468 cm−1 confirmed the presence of a hydroxyl.
HRMS confirmed the molecular formula.
3.5 Modification of the Ketone.
Selective reduction of the C-1 ketone in 1a had already been reported (36e).15, 23
Since the 1α-hydroxy of this compound was axial, however, any change in bi-
ological activity would be ambiguous; attributable either to the loss of the
ketone, or the presence of a new protruding H-bond donor. Other modifica-
tions 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 effect on the conformation of
ring A, albeit with a much larger substituent.
3.5.1 Attempted Methylenation.
Initially, Wittig olefination was attempted. Treatment of 1a with the ylide
formed from Ph3PMeBr375 and nBuLi (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 CHAPTER 3. SYNTHESIS.
O
O
O
H HO
O
O O 79
Figure 3.11: Methylenated target compound 79.
for which Wittig conditions fail. Treatment of 1a with Zn-CH2Br2-TiCl4reagent378 (room temperature, CH2Cl2, 1 hour) gave an extremely viscous,
black reaction mixture. After standard workup, however, only starting mate-
rial 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 effervesced vigorously.
3.5.2 Attempted Direct Deoxygenation.
R R
O
R R
NNH
SO
O
R R
TsNHNH2 NaBH4
Scheme 3.18: Ketone deoxygenation via a tosylhydrazone.
The next modification of the ketone to be attempted was deoxygenation. The
Clemmensen reduction (Zn/HCl) was not suitable, since α-acetoxy groups are
eliminated,380 even under the mildest conditions.381 An alternative mild ap-
proach 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
effective for ketones.385 However, some ketones react readily when the flask is
supported in an alumina bath, generating much higher temperatures.386 Since
these conditions proved ineffective, 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 Conditions Time Ref.THF - reflux 6 h 388
AcOH - rt 18 h 389AcOH - ultrasound, rt 2 h 390EtOH basic Al2O3 reflux 3 h 391
- - microwave, rt to 115 ◦C 5 min 386- silica gel microwave, rt to 115 ◦C 5 min 387- basic Al2O3 microwave, rt to 115 ◦C 5 min 387- acidic Al2O3 microwave, rt to 115 ◦C 5 min 387
Ultrasound: 40 kHz, 50 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
confirmed by TLC and 1H NMR.393
No other promising methods of tosylhydrazone formation were located. El-
Sayed’s recent review of sulfonohydrazides,394 although published in 2004, ap-
pears to have been written decades earlier. The only post-1971 references cited
are the author’s own, and these are cited only in the final sentence.
3.5.3 Indirect Deoxygenation.
3.5.3.1 Formation of Cyclic Thionocarbonate (80).
Given the failure of the direct reductions, deoxygenation of 36h was at-
tempted. The usual approaches to hydroxyl deoxygenation,395 involving either
156 CHAPTER 3. SYNTHESIS.
36h
O
O
O
H HOH
HO
O O
N
S
NN N
80
O
O
O
H HO
O
O O
S
67% from 1a
DMF, 90 °C
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 un-
reactive 1α-hydroxyl group. While direct acetylation has been achieved using
NEt3,82 this is ineffective 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 condi-
tions;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 dif-
ficult. 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 thiono-
carbonate 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 interme-
diate 36h and 38h. By 1H NMR, the H-1 and -2 signals of 80 were shifted
downfield, and the characteristic 13C peak (191 ppm) of cyclic thionocarbon-
ates was present.
3.5. MODIFICATION OF THE KETONE. 157
3.5.3.2 Unsuccessful Radical Deoxygenation Attempts.
Although radical reductions have traditionally been performed with organos-
tannanes, these compounds are highly toxic, and purification of the products
tends to be difficult.397 This has led to extensive and fruitful research into al-
ternatives, such as silanes and phosphites.396 Perhaps the most promising396 of
these alternatives is hypophosphorous acid, H3PO2. Typically, this is buffered
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 H3PO2 and NEt3 in refluxing 1,4-dioxane were used; how-
ever, benzoyl peroxide was substituted for AIBN because the latter was not
at hand. (BzO)2 gives superior results to AIBN with alkyl phosphites.396 De-
spite adding seven portions of (BzO)2 over five hours at reflux, 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 reflux for 40 minutes
gave no reaction. Addition of activated401 Mg turnings had no effect after a
further two hours’ reflux. Starting material was recovered (94%).
3.5.3.3 Radical Deoxygenation using nBu3SnH.
The traditional nBu3SnH/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 nBu3SnH used was prepared by a published procedure408 with reduced
reaction time (10 minutes),409 distilled and stored in darkness under argon at
158 CHAPTER 3. SYNTHESIS.
O
O
O
H RHO
O O
36f
O
O
O
H H
O O
83
O
O
O
H HO
O
O O
OOH
80 + +
detected by LCMS
4%
1) nBu3SnH
AIBNtoluene
80 °C, 6 h
2) silica gel
H
H
81b
82b
22%
25%
R
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 first contained a complex
mixture of products, heavily contaminated with organotin compounds. Analy-
sis of this mixture by reverse-phase liquid chromatography/mass spectrometry
(RP-LC/MS) was performed by Claudio Medana at Turin University. Com-
parison of the early fractions with 81b and 82b confirmed 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 in-
creased. One simple and effective way of removing organostannane reagents
3.5. MODIFICATION OF THE KETONE. 159
0
20
40
60
80
100
0
20
40
60
80
100total ion count
50 ≤ m/z ≤ 700 (early fractions) re
lativ
e ab
und
ance
18.60
m/z = 377 (early fractions)
m/z = 377 (pure 81b/82b)
82b
81b
19.03
15.95
19.3915.21
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44
Time (min)
15.95
15.24
0
20
40
60
80
100
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/H2O confirmed the effectiveness 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, smear-
ing badly. It would not be advisable to apply either of these methods to the
crude reaction mixture: the initial reaction products are themselves alkyltin-
substituted thiocarbonates, which are cleaved on silica to give the desired
alcohols.405 Two more complex methods of organostannane removal, involv-
ing treatment with NaBH3CN413 or I2 and KF,397 are harsher and have the
potential for side reactions.
160 CHAPTER 3. SYNTHESIS.
Thus, the organostannane route used above has serious disadvantages: the
requirement for freshly prepared reagents, the difficulty of purification, the
low recovery of products and most importantly toxicity. While nBu3SnH
is less toxic than other organostannanes,414 it is prepared from the highly
toxic (nBu3Sn)2O,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).
81a
22 101011
O
O
O
H HO
O
O O
Figure 3.13: 1-Deoxysalvinorin A (81a).
Acetylation of 81b (Ac2O/pyridine/DMAP, room temperature) gave 81a.
Both H-2 (δ 4.74, tt) and H-10 (δ 1.10, dd) showed new trans-diaxial cou-
plings to H-1α (δ 1.50, td), and smaller couplings to H-1β (δ 1.95-1.89, m).
The HMQC spectrum correlated these peaks with a new 13C peak at 26.7 ppm.
The DEPT spectrum confirmed C-1 as a methylene, and HRMS confirmed 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 salvi-
norins, display a broad range of activities against insects.2 Some act as in-
sect 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) fit an empirical pharmacophore for antifeedant
activity against Tenebrio molitor:421 an α,β-unsaturated carbonyl approxi-
mately 10 Å 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 ac-
tivity. 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 fibreglass
discs422, 423 was used, with one modification. 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 su-
crose, 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 cal-
culation of the antifeedant index, AI:
AI =100 × (∆C − ∆T )
∆C + ∆T
Antifeedant indices fall between +100 (maximal antifeedant effect) and -100
(maximal phagostimulant effect); a value of zero indicates no effect.
The results are shown in Table 4.1.
Compound AI SEMsalvinorin A 1a 0 30salvinorin C 1c 10 26
divinatorin A 28a 6 11divinatorin B 28b 2 23divinatorin C 28c 16 33
Table 4.1: Antifeedant test results.
None of these compounds had a statistically significant effect. While the fact
that four of the five compounds had a slightly positive AI suggests a weak
(nonsignificant) effect, 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 variabil-
ity 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 significant differences 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 (1a-
1d) and divinatorins A-C (28a-28c) for inhibition of eukaryotic protein syn-
thesis. The assay used has been described in detail elsewhere.424 Briefly, lu-
ciferase, the protein responsible for chemiluminescence in fireflies, 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 (firefly and Renilla) was
used. None of the compounds tested caused significant inhibition at 50 µM
(significant inhibition is defined as a relative light output of 30% or less). The
positive control, anisomycin, caused complete inhibition at 10 µM.
Relative light output (%)
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
anisomycin (+ve control)
1a
1b
1c
1d
28a
28b
28c
RenillaFirefly
Figure 4.1: Luciferase assay results for salvinorins and divinatorins (50 µM).
164 CHAPTER 4. BIOASSAYS.
4.3 Antimicrobial Activity.
4.3.1 Bacteria and Fungi.
28a28b28b29a
R1
OH OH H H
R2
O
H HR1
18
R2
HOHOAcH
R3
HMeHHO 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 there-
fore screened against standard antibiotic susceptible strains of Escherichia coli,
Staphylococcus aureus, Bacillus subtilis, and Candida albicans, using broth mi-
crodilution425, 426 and disk diffusion427 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 Mel-
bourne.
No activity was apparent against any of the test organisms at 100 µg/mL or 100
µg/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 sufficient, 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 concen-
tration (MIC) 25 µg/mL) and B. subtilis (MIC 12.5 µg/mL, 10 mm zone
of inhibition), but much less potent than its enantiomer (MIC 0.78 µg/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 diffusion assay.
An undergraduate investigation published without peer review429 reported that
4.3. ANTIMICROBIAL ACTIVITY. 165
E. coli(+)-hardwickiic acid (ent-29a)
(+)-hardwickiic acid (ent-29a)
MICMIC
1005025
12.56.253.12
1.56
0.78
0.390.19
µg/mL
B. subtilis S. aureus
E. coli
C. albicans
Staph. aureus
B. subtilis
Crude extract
streptomycin sulphate(positive control)
acetone(negative control)
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 confirm these results. The acetone extract of the commercial
material, as well as 1a, showed no activity at 100 µg/mL or 100 µg/disk.
Apparently insufficient 29a and oleanolic acid (31) were present to elicit an
effect (31, like 29a, is active against B. subtilis and Staph. aureus).245
4.3.2 HIV-1.
Several κ 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 offer still greater syn-
ergy. KOR antagonists, which do not themselves inhibit viral replication,
have been shown to enhance the effects of the standard therapy azidovudine
(AZT).433
Beside their effect on viral replication, κ opioids also appear to counteract the
166 CHAPTER 4. BIOASSAYS.
neurotoxic effects of HIV in infected cells.434, 435 Finally, since nausea is a com-
mon side effect of standard therapies,436 the possible antiemetic activity437, 438
of κ opioids may be advantageous.
4.3.2.1 NL4.3 and AD8 Strains.
NL4.3
HIV strain:
AD8
1a (10-6 M)
Me2SO only
control
101
102
103
104
105
106
107
0 2 4 6 8
Days post infection
HIV
co
pie
s /
106 c
ells
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 University’s 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 pe-
ripheral blood mononuclear cells (PBMCs) with 1 µM salvinorin A had no
effect 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 µM) had no effect 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.
20
40
60
80
100
120
140
160
0.00 0.02 0.07 0.23 0.73 2.30 7.27 23.0 72.7 230
0.00 0.10 0.32 1.00 3.16 10.0 31.6 100 316.2 1000
CONCENTRATION (nM)
CONCENTRATION (µM)
20
40
60
80
100
120
140
160
% VIRUS CONTROL
% CELL CONTROL
% VIRUS CONTROL
% CELL CONTROL
0
20
40
60
80
100
120
140
0
20
40
60
80
100
120
140 PE
RC
EN
T O
F C
EL
L C
ON
TR
OL
PE
RC
EN
T O
F V
IRU
S C
ON
TR
OL
PE
RC
EN
T O
F V
IRU
S C
ON
TR
OL P
ER
CE
NT
OF
CE
LL
CO
NT
RO
L
AZT
salvinorin A (1a)
Figure 4.5: HIV-1 replication assays (ROJO isolate).
4.3.2.3 Discussion.
These results were surprising given the established activity of the κ opioids
U50,488 and U69,593, which are of comparable potency to 1a. However, the
extremely potent endogenous κ 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 A1−13, 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 A1−17, which has a half life in plasma of three hours,444 also caused
168 CHAPTER 4. BIOASSAYS.
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 κ opioid receptor
internalisation than U50,488, despite their similar potencies.81 Such differences
may involve the apparent subtypes445 of κ 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 sufficient
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 10−5 M in all cases (Figure
4.6). In the one apparent exception, 28b exhibited a GI50 of just under 10−6
M against the SF-539 brain tumour cell line. However, a significant inhibition
of growth was already apparent at 10−8 M (59% of control), which scarcely
increased up to 10−5 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. GI50values 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. 169
All Cell Lines
% G
row
th
-8 -7 -6 -5 -4
Log10
Concentration (M)Log10
Concentration (M)
-8 -7 -6 -5 -4
-8 -7 -6 -5 -4
All Cell Lines
-8 -7 -6 -5 -4
100
50
-50
-100
0
100
50
-50
-100
0
100
50
-50
-100
0
100
50
-50
-100
0
100
50
-50
-100
0
% G
row
th
-8 -7 -6 -5 -4
salvinorin B (1b)
(NSC D737715)
divinatorin A (28a)
(NSC D737768)
divinatorin B (28b)
(NSC D737769)
divinatorin C (28c)
(NSC D737770)
salvinorin C (1c)
(NSC D737716)
Figure 4.6: NCI 60 cell line results for salvinorins and divinatorins
in 8 cell lines. The “Total Growth Inhibition” concentration (TGI) was above
10−5 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 cyto-
toxicity: the SNB-75 CNS tumour line, with an LC50 slightly below 10−4 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 CHAPTER 4. BIOASSAYS.
Log10
Concentration (M)
100
50
-50
-100
0
-8 -7 -6 -5 -4
Log10
Concentration (M)
100
50
-50
-100
0
% G
row
th
-8 -7
SF-268SNB-75 SNB-19
SF-295 SF-539U251
-6 -5 -4
salvinorin B (1b)
(NSC D737715)
divinatorin B (28b)
(NSC D737769)
CNS tumour cell line:
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 tu-
mour cells (lymphocytic leukaemia); test animals were injected with 1a daily
for 5 days. Doses of 1a up to 76 mg/kg had no effect, 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 κ Opioid Receptor.
Previous work found that 1b was inactive at the κ opioid receptor, while re-
placement of the acetoxy group in 1a with more hindered esters dramatically
reduced binding affinity.135 No information was available about other func-
tional groups.
In order to further explore salvinorin A’s structure-activity relationships, iso-
lated compounds and derivatives were tested in vitro for binding affinity at
cloned opioid receptors using radioligand assays. Those compounds with sub-
micromolar affinity were also screened for agonist potency and efficacy using
4.5. ACTIVITY AT THE κ 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 Health’s Psychoactive Drug Screening Program.
No compound showed affinity for µ or δ subtypes (K i > 1 µM), and the
following discussion will therefore discuss the κ 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 affinities and potencies relative to
1a:
Krel(x) =Ki(x)
Ki(1a)
and
ECrel(x) =EC50(x)
EC50(1a)
A value of 10 thus signifies a potency or affinity 10 × lower than 1a.
O
O
O
H HOO
O
O O
1a
4 ± 1 nM
46 ± 8 nM
Ki
EC50
Figure 4.8: KOR binding affinity and potency of salvinorin A.
4.5.1 Other Salvinorins and Divinatorins.
4.5.1.1 Radioligand Binding Results.
There have been conflicting results on salvinorin B (1b). Initial work by the
Roth group found it inactive (K i > 10 µM),135 but subsequent work by Béguin
et al reported high affinities (K i = 66,137 111271 or 155125 nM). Retesting by the
Roth group indicated very weak affinity (3.1 µM versus 1a control 44 nM).452
172 CHAPTER 4. BIOASSAYS.
K i s.e.m. K rel EC50 s.e.m. ECrel Emax s.e.m.nM nM %
1a 4 1 46 8 100 191b 3,153* 71*1c 1,022 262 2551d >10,000 >2,5001e >10,000 >2,500
28a >10,000 >2,50028b >10,000 >2,50028c >10,000 >2,50046 18 2 4.5 315 35 6.8 108 11
37a 163 50 41 244 102 5.3 78 1736e 1,125 28136c >10,000 >2,50051 156 18 39 126 36 2.7 108 1435 59 11 15 78 21 1.7 107 549 6 1 1.5 223 60 4.8 103 1350 6 2 1.5 624 200 14 116 1077 347 53 87 >10,000 >217
81a 18 2 4.5 141 43 3 122 2759 >10,000 >2,500
60a 2,900 400 725* = direct comparison to 1a (44 nM) against [3H]diprenorphine
Table 4.2: KOR radioligand and functional assay results.
O
O
O
H HOHO
O O
O
O
O
H HOR1
R2O
O O
28a28b28c
R1
OH OH H
R2
O
H HR1
R2
HOHOAc
R3
HMeH
O OR3
Krel
>2,500>2,500>2,500
1c1d1e
R2
Ac H Ac
R1
AcAcH
Krel
255>2,500>2,500
1b
Krel
71
Figure 4.9: KOR binding affinities of salvinorins and divinatorins.
The discrepancy is not as great as it appears, since the two groups obtained
different absolute affinities for 1a. Béguin et al reported values of K rel = 66,137
85271 or 120.125 The Roth group’s latest result (K rel = 71)452 is concordant.
4.5. ACTIVITY AT THE κ OPIOID RECEPTOR. 173
The nonconcordant earlier result (K rel > 500) may have resulted from the use
of a different radioligand – [3H]bremazocine rather than [3H]diprenorphine.
Moreover, in all cases the binding affinity of 1b was negligible relative to 1a.
Salvinorin C (1c) also showed negligible binding affinity compared to 1a (K rel
= 255, Figure 4.9). Salvinorins D (1d) and E (1e) and divinatorins A-C (28a-
28c) showed no affinity (K rel > 2,500). Based on these results, we tentatively
concluded270 that 1a is the sole κ 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 affinities
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
effectively the sole active principle of S. divinorum.
The weak activity of 28d shows that the C ring can be cleaved without total
loss of affinity. Apart from this greater conformational freedom, 28d not only
lacks the 2-acetoxy function, but also possesses a 1α-hydroxy group, both of
which drastically reduce activity in analogues of 1a, as will be discussed be-
low. The structure-activity relationships of this compound are thus markedly
different 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 Modification of the Ketone.
Reduction of the ketone to an α-hydroxy group (giving 36e) dramatically
reduced binding affinity (Figure 4.10). Acetylation (giving 36c) abolished
binding entirely. These results were surprising, since molecular modelling had
indicated that these modifications would not affect binding.67 Although the
very low affinities of 1e, 1c, 36e, and 36c might appear to suggest that the
174 CHAPTER 4. BIOASSAYS.
O
O
O
H HOO
O
O
O O
O
O
O
H HOH
O
O
O O 36c
>2,500Krel
36e
281Krel
O
O
O
H HO
O
O O 81a
4.5
3
Krel
ECrel
Figure 4.10: KOR activity after ketone modifications.
ketone is part of the pharmacophore, the relatively high affinity and potency
of 1-deoxy compound 81a show that it is not. This suggests that 1α-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) affinity. The effect of
this double bond itself therefore remains unclear.
4.5.3 Modification of the Acetoxy Group.
O
O
O
H HOO
O
O O 46
4.5
6.8
Krel
ECrel
O
O
O
HOH
O
O O
O
O
O
O
HO
O O
O
O
H
60a
725Krel
59
>2,500Krel
Figure 4.11: KOR activity after acetoxy group modifications.
It had previously been shown that substituting more hindered esters for the
2-acetoxy group reduced binding affinity.135 This suggested that the less-
4.5. ACTIVITY AT THE κ OPIOID RECEPTOR. 175
hindered formate 46 (Figure 4.11 on the preceding page) might prove more
potent, but in fact both affinity 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 affinity 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 affinity dramatically.
4.5.4 Modification of the Methyl Ester.
O
O
O
H HOO
O
OH 77
87
>217
Krel
ECrel
O
O
O
H HOO
O
O OH 67a
-
Figure 4.12: KOR activity after methyl ester modifications.
The role of the methyl ester in binding was strongly confirmed 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-8β epimer was inactive at the KOR, while the H-8α epimer had high affinity
(K i = 48 nM).361
176 CHAPTER 4. BIOASSAYS.
4.5.5 Modification of the Lactone.
O
OH
O
H HOO
O
O O
O
O
H HOO
O
O O
O
O
HO
O
O
O O50
1.5
14
49
1.5
4.8
35
15
1.7
Krel
ECrel
Krel
ECrel
Krel
ECrel
Figure 4.13: KOR activity after lactone modifications.
The high affinities of lactol 35, ether 49, and enol ether 50 (all full agonists)
show that the lactone carbonyl is not essential for binding or activity. How-
ever, the latter two compounds, especially 50, showed reduced potency in the
functional assay.
4.5.6 Modification of the Furan Ring.
O
O
O
H HOO
O
O O
HH13
51
39
2.7
Krel
ECrel
O
O
O
H HOO
O
O O
8
37a
41
5.3
Krel
ECrel
Figure 4.14: KOR activity after furan modifications.
The substantially reduced affinity of tetrahydrofuran 51 suggests that the fu-
ran ring is involved in binding. As described earlier, one 13-epimer of 51
was isolated for characterisation. The binding affinity did not differ signif-
icantly from the 1:1 mixture (123 nM vs 156 nM). The reduced affinity of
4.5. ACTIVITY AT THE κ OPIOID RECEPTOR. 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 difference between this compound and 1a. The reduction in potency
of these compounds was far smaller than the reduction in affinity, particularly
in 51.
4.5.7 Incorporation into a Revised Binding Model.
O
O
O
H HOO
O
O O
OTyr 313
OTyr 312
Gln 115N
OH
H
H
H
O
Tyr 139
H
Figure 4.15: Westkaemper’s 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 affinities 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 finished 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 CHAPTER 4. BIOASSAYS.
modification of the receptor itself, identifying the residues involved in binding
to 1a.453 The revised model is shown in Figure 4.16, visualised from coordi-
nates kindly provided by Dr Westkaemper. For clarity, hydrogen atoms are
not shown.
Tyr 119 Tyr 119
Tyr 320
- - - H-bonds*
*not present simultaneously *not present simultaneously
Tyr 320
Glu 297 Glu 297
Tyr 313 Tyr 313
Ile 294- - - hydrophobic interactions
- - - hydrophobic interactions
- - - H-bonds*
Ile 294
Figure 4.16: Westkaemper’s 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 affect binding affinity.453 Three
of the five 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 five residues involved
(isoleucine 294, glutamic acid 297 and tyrosine 313) are unique to the κ sub-
type. This provides a plausible explanation for the striking κ selectivity of
4.5. ACTIVITY AT THE κ OPIOID RECEPTOR. 179
1a and analogues; very few derivatives show submicromolar affinity δ or µ
subtypes, and none of these are selective.25, 82, 136
4.5.8 Subsequent Results.
O
O
O
H HORO
O O
O OO
O
O
O
O
O
H HOO
O
O OH67a (H-8b)
>770Krel
O
O
O
H HOO
O
O O
HHO
34a
102ECrel
O
OHO
Krel
ECrel
84
6
4
85
0.31
0.13
86
12
0.13
Ki (m)
Krel
(m/k)
Figure 4.17: KOR binding affinities and potencies of recent derivatives.
Since publication of these results, numerous 1a analogues have been tested
in vitro by other groups. These results again confirm that replacement of the
2-acetoxy group (with other esters, ethers, carbonates, amides etc) almost in-
variably reduces binding affinity.95, 271, 82 There are exceptions, however. The
high affinity and potency of ethyl ether 8495 indicate that the acetyl carbonyl
is not essential for activity. This result provides strong support for the hy-
drophobic interaction at C-2 in Westkaemper’s revised model (Figure 4.17).
Another notable derivative is methoxymethyl ether 85,271 to date the only
derivative significantly more potent than 1a. Surprisingly, benzoate 8682 and
some related derivatives activate the µ 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 modifications at C-18,361, 95 but all esters and amides
180 CHAPTER 4. BIOASSAYS.
tested showed large reductions in affinity and potency. The loss of affinity on
O-demethylation (giving 67a)361 strongly supports the carbinol hydrophobic
interaction in Westkaemper’s model over the more intuitive carbonyl H bond.
Even less data has been reported on furan modifications, but the greatly re-
duced potency of salvinicin A (34a)25 confirms that hindrance in this region
interferes with binding. Some synthetic modifications of the furan ring have
been prepared,136 but biological data have not been reported.
Chapter 5
Experimental.
5.1 General Conditions
5.1.1 Instruments and Procedures.
Flash Chromatography: Flash column chromatography was performed ac-
cording to Leonard’s194 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 difficult separations (∆hRf < 5). The ac-
tivated carbon used was Merck aktivkohle 2183 powder (20:1 mass ratio
versus crude product).
HPLC: Spherex 5 µm silica column (250 × 10 mm), flow rate 2 mL min−1
with refractive index detection. Solvent front reached detector at 6.8
min.
HRESIMS: Bruker 4.7T BiOAPEX FTMS.
InChIs: IUPAC International Chemical Identifiers454 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 films on NaCl discs.
181
182 CHAPTER 5. EXPERIMENTAL.
LCMS: Phenomenex Luna 3 µm C-18 column (150 × 2 mm). Gradient elu-
tion: 0.05 % HCO2H in 20-100% MeCN/H2O over 40 minutes. ESI in-
terface 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 (run-
ning TopSpin 1.3). 1H NMR are 400 MHz, and 13C NMR 100 MHz, un-
less otherwise stated. 13C multiplicities are based on DEPT experiments.
Assignments are given according to the standard32 clerodane numbering
scheme below. Where 13C assignments are given, all assignments (1H and13C) are based on 2D NMR data (COSY, HMQC, HMBC). In other cases,1H 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 first-order multiplets were analysed using Hoye’s al-
gorithms,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 δ
values of solvent residual peaks (eg. CHCl3 = 7.26 ppm) and impurities
were taken from Gottlieb et al.458
22
3344
55
101011
66
77
8899 1717
12121111
1313
1616
1515
1414
18181919
2020
Polarimetry: JASCO DIP-1000. Concentration c is in g/l00 mL; the units
5.1. GENERAL CONDITIONS 183
of the specific rotation are (◦·mL·g−1·dm−1).
SEM: Samples were sputter-coated with gold. Fresh leaves were fixed in 2.5%
glutaraldehyde in 0.1 M phosphate buffer at rt for 24 h. The samples were
rinsed in the buffer 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 4Å 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 Mex-
ico, were purchased from Salvia Space Ethnobotanicals (Berkeley, Cal-
ifornia). Voucher specimens were deposited at the National Herbarium
of Victoria (accession number MEL 2101361) and the University of Mel-
bourne Herbarium (MELU s.n.)
Australian material Additional S. divinorum plants were cultivated in Mel-
bourne. A voucher specimen of the dried leaves was deposited at the
National Herbarium of Victoria (accession number MEL 2145478).
184 CHAPTER 5. EXPERIMENTAL.
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.
κ: rat KORs with [3H]diprenorphine (50 Ci/mmol, PerkinElmer Inc) or [3H]U69,593
(41.4 Ci/mmol, PerkinElmer Inc) as radioligand.
δ: human DORs with [3H]DADLE (51.5 Ci/mmol, PerkinElmer Inc) as radi-
oligand.
µ: human MORs with [3H]diprenorphine as radioligand. K i values were calcu-
lated using Prism 4.01 (GraphPad Software, Inc) as the mean ±SEM of
quadruplicate (n ≥ 4) determinations. Nonspecific binding was defined
using 10 µM 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 Gα16. Ca2+ mobilisation was quantified
using a 96-well FlexStationII with the calcium flux 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 qua-
druplicate (n ≥ 4) determinations.
Antimicrobial Tests. The extract and compounds were tested against Es-
cherichia coli (ATCC 25922), Staphylococcus aureus (ATCC 29213),
Bacillus subtilis (ATCC 6633), and Candida albicans (ATCC 90028) us-
ing standard broth microdilution426, 425 (100 - 0.19 µg/mL using two-fold
serial dilutions) and disc-diffusion427 assays (100 µg/disk). All measure-
ments were performed in duplicate. Streptomycin sulfate and ampho-
tericin B were used as positive controls.
5.2. ISOLATION 185
Insect Antifeedant Tests. A standard choice assay employing sweetened fi-
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 filter (3.6 cm2) were treated with 100 µL of 0.05 M
sucrose solution. Test discs were then treated with 10 µg of test com-
pound in acetone, usually423 expressed as 100 µL of a 100 ppm (100 mg
L−1) solution. Control discs were treated with the same volume of ace-
tone. 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 Me2SO before addition. Me2SO at the same concen-
tration 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 Isolation
5.2.1 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 purified by flash column chromatography on
an equal mixture of activated carbon and diatomite filter 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 purified by flash column chromatography on silica gel
186 CHAPTER 5. EXPERIMENTAL.
(5-50% acetone/CH2Cl2 gradient). This was divided, based on TLC (10%
acetone/CH2Cl2), into four series: A (656 mg), B (150 mg), C (359 mg) and
D (77 mg).
TLC: hRf of series A B C D
10% acetone/CH2Cl2 62 36-53 18 12
Series A: flash column chromatography, eluting with a gradient from 50-80%
Et2O/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: flash column chromatography on silica gel (35 g) in 70-90% Et2O/petrol,
and recrystallisation from MeOH, gave 1b (13 mg, 0.015 g/kg).
Series C: Trituration in hot Et2O gave 1d (75 mg). Flash column chromatog-
raphy of the mother liquor (60-100% Et2O/petrol) gave four fractions based
on TLC (70% Et2O/petrol): C1 (55 mg), C2 (119 mg), C3 (57 mg) and C4
(39 mg).
TLC: hRf of fraction C1 C2 C3 C4
70% Et2O/petrol 40-43 28-21 17 13
Fraction C1: Repeated flash column chromatography (20% acetone/petrol
and 40-60% Et2O/petrol) gave 28c (23 mg) and 31 (3 mg).
Fraction C2: Repeated flash column chromatography (25% acetone/petrol
and 60-100% Et2O/petrol) gave 28b (41 mg).
Fraction C3: Extensive flash column chromatography (Et2O/petrol, ace-
tone/petrol and EtOAc/petrol) gave additional 28b (total yield 41 mg) and a
mixture of 1e and 1f. Final purification 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 flash column chromatography (60% Et2O/petrol and 4%
MeOH/CH2Cl2) 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 purified by vacuum filtration through a 1:1
mixture of activated carbon (75 g) and diatomite filter aid, eluting with a
gradient from 50-20% EtOAc/petrol, to give series E (97 mg) and F (279 mg)
based on TLC (70% Et2O/petrol).
TLC: hRf of series E F
70% Et2O/petrol 55-68 16
Series E: Repeated flash column chromatography (1% acetone/CH2Cl2 and
20% Et2O/petrol) gave 33 (1 mg). Further flash column chromatography
(0.75% MeOH/CH2Cl2 and 1% EtOH/CHCl3) gave 32 (23 mg) and 30 (12
mg).
Series F: Two recrystallisations from MeOH gave 1a (126 mg).
yield 70% Et2O/ 10% acetone/(mg/kg) petrol CH2Cl2
Salvinorin A 1a 3,400 24 57B 1b 15 14 37C 1c 254 31 60D 1d 132 18 25E 1e 3 23 47F 1f 1 24 40
Divinatorin A 28a 42 37 15B 28b 48 31 31C 28c 27 50 39
(–)-Hardwickiic acid 29a 8 64 45Oleanolic acid 31 3 66 65
Presqualene alcohol 32 3 44 34Peplusol 33 4 75 73
(E)-Phytol 30 3 59 58
Table 5.1: Yields and TLC data (hRf ) of isolated compounds.
188 CHAPTER 5. EXPERIMENTAL.
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,5,7,9-10H2,1-
4H3/t14-,15-,16-,17-,19-,22-,23-/m0/s1
1a
O
O
O
H HOO
O
O O
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): λmax (log ε) 208 (3.76) nm;
lit.23 UV (MeOH): λmax (log ε) 211 (3.72) nm;
1H NMR (800 MHz, CDCl3): δ 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, CO2CH 3), 2.74 (1H, ∼dd, J ≈ 11.3, 5.6 Hz, H-4), 2.50 (1H, dd, J =
13.5, 5.2 Hz, H-11α), 2.31-2.28 (2H, m, H-3), 2.18 (1H, br s, H-10), 2.17-2.14
(1H, m, H-7β), 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-6α), 1.64 (1H, tdd, J = 13.5, 12.1, 3.4 Hz,
H-7α), 1.58 (1H, td, J = 13.5, 0.9 Hz, H-6β), 1.57 (1H, ddd, J = 13.5, 11.7,
0.8 Hz, H-11β), 1.45 (3H, s, H-20), 1.11 (3H, s, H-19);
13C 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
1b
O
O
O
H HOHO
O O
colourless crystals, mp (from hot MeOH) 239-240 ◦C; (from cold Et2O/petrol)
244-245 ◦C;
lit. (from hot MeOH) 213-216 ◦C;23 251-254 ◦C;15, 260 (from cold MeOH) 211-
214◦C;169
1H and 13C NMR (CDCl3) data matched previously reported values.23
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
1c
O
O
O
H HOO
O
O
O O
190 CHAPTER 5. EXPERIMENTAL.
clear resin;
TLC: See Table 5.1 on page 187.
[α]16D +70 (c 0.6, CHCl3);
[α]22D +49 (c 0.6, CHCl3) lit;27
UV (MeCN): λmax (log ε) 208 (4.10) nm;
FTIR (film): ν̃max 3145, 2952, 1740, 1643, 1506, 1435, 1372, 1315, 1226, 1174,
1142, 1075, 1041, 961, 950, 910, 875, 789, 754, 695, 667 cm−1;
1H and 13C NMR (CDCl3) data matched previously reported values.27
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
1d
O
O
O
H HOHO
O
O O
fine colourless crystals, mp (from cold Et2O) 185-187 ◦C;
TLC: See Table 5.1 on page 187.
[α]17D +67 (c 1.0, CH2Cl2);
FTIR (film): ν̃max 3475, 3146, 2952, 2861, 1723, 1505, 1435, 1371, 1315, 1228,
1195, 1174, 1142, 1071, 1056, 1027, 949, 875, 788, 744, 699 cm−1;
1H NMR (CDCl3): δ 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, CO2CH 3), 2.56 (1H, dt, J
= 13.3, 3.4 Hz, H-6α), 2.54 (1H, dd, J = 13.1, 5.9 Hz, H-11α), 2.17-2.09 (1H,
m, H-7β), 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-7α), 1.69
(3H, s, H-19), 1.64 (1H, ddd J = 13.1, 11.2, 1.0 Hz, H-11β), 1.42 (1H, br s,
H-10), 1.22 (3H, s, H-20), 1.20 (1H, td, J = 13.5, 3.5 Hz, H-6β);
13C NMR (CDCl3): δ 171.57 (C, C-17/OCOCH3), 171.55 (C, C-17/OCOCH3),
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, CO2CH3),
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, OCOCH3), 18.3 (CH2, C-7), 15.6 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 455.1672 (calcd for C23H28O8Na+, 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
1e
O
O
O
H HOH
O
O
O O
clear resin;
TLC: See Table 5.1 on page 187.
HPLC: tR (min) 1e 1f
60% EtOAc/petrol 9.8 10.7
192 CHAPTER 5. EXPERIMENTAL.
[α]17D +46 (c 0.14, CHCl3);
FTIR (film): ν̃max 3510, 3144, 2952, 2858, 1722, 1641, 1505, 1436, 1374, 1316,
1228, 1142, 1070, 1029, 949, 935, 897, 875, 805, 788, 763, 708, 690, 679 cm−1;
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH 3), 2.52 (1H,
ddd, J = 12.5, 3.0, 2.6 Hz, H-6α), 2.46 (1H, dd, J = 13.1, 6.0 Hz, H-11α),
2.17 (3H, s, OCOCH 3), 2.18-2.07 (2H, m, H-7β & 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-7α), 1.72 (3H, s, H-19),
1.62 (1H, dd, J = 13.1, 11.2 Hz, H-11β), 1.47 (3H, s, H-20), 1.30 (1H, br s,
H-10), 1.19 (1H, td, J = 13.3, 3.7 Hz, H-6β);
13C NMR (CDCl3): δ 171.8 (C, OCOCH3), 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, C-
1), 54.0 (CH, C-10), 51.8 (CH3, CO2CH3), 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,
OCOCH3), 18.4 (CH2, C-7), 16.2 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 455.1687 (calcd for C23H28O8Na+, 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 193
1f
O
O
O
H HOH
O O
clear resin;
TLC: See Table 5.1 on page 187.
HPLC: see table on page 191.
[α]16D −20 (c 0.05, CHCl3);
FTIR (film): ν̃max 3514, 3147, 2951, 2857, 1712, 1637, 1505, 1436, 1372, 1318,
1232, 1194, 1144, 1070, 1028, 978, 947, 896, 875, 797, 756, 686 cm−1;
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH 3), 2.60 (1H, ddd, J = 20.1, 5.5, 3.0 Hz, H-2β), 2.53
(1H, dt, J = 13.2, 3.2 Hz, H-6α), 2.46 (1H, dd, J = 13.2, 5.8 Hz, H-11α), 2.35
(1H, ddt, J = 20.1, 4.6, 1.1 Hz, H-2α), 2.17-2.08 (2H, m, H-7β, 8), 1.82 (1H,
dtd, J = 15.0, 13.2, 3.5 Hz, H-7α), 1.71 (3H, s, H-19), 1.62 (1H, ddd, J =
13.2, 11.4, 0.9 Hz, H-11β), 1.48 (3H, s, H-20), 1.29 (1H, dd, J = 4.7, 0.7 Hz,
OH ), 1.25 (1H, br s, w 12
= 3.8 Hz, H-10), 1.18 (1H, tdd, J = 13.3, 3.6, 0.9 Hz,
H-6β);
13C NMR (CDCl3): δ 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, C-
14), 71.7 (CH, C-12), 63.9 (CH, C-1), 54.8 (CH, C-10), 52.2 (CH, C-8), 51.5
(CH3, CO2CH3), 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 C21H26O6Na+, 397.1622).
194 CHAPTER 5. EXPERIMENTAL.
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)10-
7-14-8-11-24-12-14/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
28a
O
H H
O OH
OH
amber resin;
TLC: See Table 5.1 on page 187.
[α]19D −53 (c 1.8, CH2Cl2);
FTIR (film): ν̃max 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 cm−1;
1H NMR (CDCl3): δ 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-2β), 2.43-2.33 (2H, m, H-2α, 6α), 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-7α, 8), 1.47-1.42
(1H, m, H-7β), 1.45 (1H, br s, H-10), 1.23-1.17 (1H, m, H-6β), 1.15 (3H, s,
H-20), 0.84 (3H, d, J = 6.2 Hz, H-17);
13C NMR (CDCl3): δ 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 C20H28O4Na+, 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
28b
OH
O
H HOH
O O
amber resin;
TLC: See Table 5.1 on page 187.
[α]20D −54 (c 2.1, CHCl3);
FTIR (film): ν̃max 3434, 2930, 2881, 1714, 1503, 1461, 1437, 1385, 1356, 1236,
1196, 1162, 1097, 1067, 1025, 979, 943, 921, 874, 781, 759, 734 cm−1;
1H NMR (500 MHz, CDCl3): δ 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, CO2CH 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-2β), 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-6α),
2.31 (1H, ddt, J = 19.9, 4.9, 1.4 Hz, H-2α), 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-7β), 1.77 (1H, ddd, J = 15.0, 12.6, 4.6 Hz,
196 CHAPTER 5. EXPERIMENTAL.
H-11-pro-R), 1.66 (3H, s, H-19), 1.64-1.49 (2H, m, H-8 & 7α), 1.49 (2H, br s,
OH ), 1.45 (1H, br s, H-10), 1.19-1.13 (1H, m, H-6β), 1.18 (3H, s, H-20);
13C NMR (CDCl3): δ 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, C-
1), 63.9 (CH2, C-17), 51.3 (CH3, CO2CH3), 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 C21H30O5Na+, 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
28c
O
O
H H
O OH
O
amber resin;
TLC: See Table 5.1 on page 187.
[α]25D −110 (c 1.1, CHCl3);
FTIR (film): ν̃max 2960, 2938, 2873, 2650, 1738, 1681, 1629, 1502, 1459, 1420,
1385, 1367, 1236, 1161, 1064, 1025, 1000, 975, 873, 785, 759, 730 cm−1;
1H NMR (500 MHz, CDCl3): δ 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-6α), 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-2α),
2.24-2.16 (2H, m, H-2β & 12-pro-S), 2.03 (3H, s, OCOCH 3), 1.82-1.67 (4H,
m, H-1β, 7β, 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-1α & 7α), 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-6β), 0.83 (3H, s, H-20);
13C NMR (CDCl3): δ 171.9 (C, C-18), 171.2 (C, OCOCH3), 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, OCOCH3), 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 C22H30O5Na+, 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)11-8-
15-9-12-23-13-15/h5,9,12-14,17H,4,6-8,10-11H2,1-3H3,(H,21,22)/t14-,17-,19+,
20+/m1/s1
O
H H
O OR
R
HMe
29a29b
1H 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 CH2N2 in
Et2O gave the methyl ester 29b;
198 CHAPTER 5. EXPERIMENTAL.
InChI=1/C21H30O3/c1-15-8-11-21(3)17(19(22)23-4)6-5-7-18(21)20(15,2)12-
9-16-10-13-24-14-16/h6,10,13-15,18H,5,7-9,11-12H2,1-4H3/t15-,18-,20+,21+
/m1/s1
[α]18D −86 (c 0.04, CHCl3);
[α]23D −104 (c 1.1, CHCl3) lit;217
1H and 13C NMR (CDCl3), FTIR (film) and EIMS (70 eV) matched previ-
ously reported values.218, 217 The 1H 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
O
OH
HOH
H 31
HRESIMS: [M + Na]+ m/z 479.3516 (calcd for C30H48O3Na+, 479.3496);
1H and 13C 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 199
OH
H
32
[α]21D +45 (c 1.2, CHCl3);
[α]20D +49 (c 4.8, CHCl3) lit;223
1H NMR (CDCl3),222 13C NMR (C6D6)223 and FTIR (film)222 matched previ-
ously reported values. The 1H NMR spectrum in C6D6 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
[α]20D −6 (c 0.07, CHCl3);
[α]25D −18 (c 0.74, iPrOH) lit;224
1H and 13C NMR (CDCl3) and FTIR (film) matched previously reported val-
ues.224
5.3 Synthesis
5.3.1 Salvinorin C (1c) via acetylation of salvinorin D
(1d).
Ac2O (250 µL, 2.6 mmol) was added to a solution of 1d (11.0 mg, 25.4
µmol) in dry pyridine (2.5 mL) under Ar. After stirring for 3.5 h, TLC
200 CHAPTER 5. EXPERIMENTAL.
(10% acetone/CH2Cl2) indicated completion. The reaction mixture was di-
luted with ice water and extracted with Et2O (× 3). The organic phase was
washed with saturated NaHCO3, saturated CuSO4, water and brine. Drying
(MgSO4), evaporation in vacuo and flash column chromatography on silica gel
(55% Et2O/petrol) gave 1c as a clear resin (8.4 mg, 70%);
TLC: See Table 5.1 on page 187. Cospotted with isolated material.
[α]16D +69 (c 0.4, CHCl3).
Other spectra (FTIR, 1H and 13C NMR) superimposable with those of the
isolated material.
5.3.2 Salvinorins D (1d) and E (1e) via acetylation of
1h.
1h (2.8 mg, 7.2 µmol), Ac2O (7 µL, 74 µmol) and catalytic DMAP (�1 mg)
were stirred in pyridine (3 mL) for 3 h, when TLC (10% acetone/CH2Cl2,
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.
1H 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 µmol), Ac2O (8 µL, 85 µmol) and catalytic DMAP (�1
mg) were stirred in pyridine (500 µL) at 45 ◦C for 75 min, when TLC (10%
acetone/CH2Cl2, visualised with KMnO4/H2SO4 dip) showed no starting ma-
terial. Evaporation in vacuo and flash column chromatography (2.5 – 10%
acetone/CH2Cl2) gave 1c and 1e; the latter was contaminated by an insepa-
rable byproduct.
5.3. SYNTHESIS 201
TLC: See Table 5.1 on page 187 and table on this page. Cospotted with
isolated materials.
1H 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 µmol) and Na2CO3 (5.1 mg, 41.1 µmol) in
CH2Cl2 (1 mL) was added MeOH (1 mL), and the solution stirred at rt for 2
h, when TLC (10% acetone/ CH2Cl2) showed considerable starting material.
After heating at 45 ◦C for a further 90 min, TLC indicated completion. The
solution was partitioned between brine (acidified with 10% HCl) and CH2Cl2(× 3). Drying (MgSO4), evaporation in vacuo, and flash column chromatogra-
phy (loaded in CH2Cl2, eluted with 33 - 50% EtOAc/petrol, then 25% MeOH/
CH2Cl2) 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
1h
O
O
O
H HOH
HO
O O
TLC: hRf 1c 1h
10% acetone/CH2Cl2 60 18
[α]18D +27 (c 0.2, CH2Cl2);
FTIR (film): ν̃max 3456, 2951, 1714, 1504, 1435, 1379, 1314, 1229, 1177, 1144,
1075, 1049, 1027, 949, 875, 788, 736, 685 cm−1;
202 CHAPTER 5. EXPERIMENTAL.
1H NMR (CDCl3): δ 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, CO2CH 3), 2.49 (1H, dd, J = 13.2, 6.0 Hz,
H-11α), 2.50-2.45 (1H, m, H-6α), 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-7β), 1.82 (1H, dtd, J = 15.0, 13.2, 3.3
Hz, H-7α), 1.70 (3H, s, H-19), 1.60 (1H, ddd, J = 13.0, 11.1, 0.8 Hz, H-11β),
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-6β);
13C NMR (CDCl3): δ 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, C-
14), 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, CO2CH3), 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 C21H26O7Na+, 413.1571).
5.3.5 (+)-Hardwickiic acid (ent-29a).
Following a modified version of Costa et al’s procedure,218 the acid fraction
of copaiba balsam was methylated with CH2N2 in Et2O, and the methyl ester
(ent-29b) was isolated. This ester (32 mg, 97 µmol) was dissolved in acetone.
KF/Al2O3 (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 filtered, and the filter cake was rinsed with water
(× 2). Rinsing of the filter cake with CHCl3 (× 3) gave, after drying (MgSO4)
and evaporation, starting material ent-29b (11 mg). The aqueous filtrate was
acidified with 10% HCl and extracted with CHCl3 (× 4). The pooled organic
extracts were dried (MgSO4) and evaporated to give 14 mg crude product.∗After D2O exchange.
5.3. SYNTHESIS 203
Flash column chromatography in 80% Et2O/petrol gave ent-29a (5 mg, 16
µmol) as a semicrystalline film;
InChI=1/C20H28O3/c1-14-7-10-20(3)16(18(21)22)5-4-6-17(20)19(14,2)11-8-
15-9-12-23-13-15/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 H
O OR
R
H
Me
ent-29a
ent-29b
TLC: hRf ent-29a ent-29b
5% acetone/CH2Cl2 31 76
[α]19D +81 (c 0.2, CHCl3);
[α]23D −85 (CHCl3) lit. value for 29a;217
FTIR, 1H and 13C NMR data217 matched reported values.
5.3.6 Salvinorin A lactol (35).
Following the published procedure,260 1a (15.8 mg, 36.5 µmol) was warmed in
THF (1 mL) under Ar until fully dissolved, then cooled to -78 ◦C. iBu2AlH
(1M in THF, 0.5 mL, 500 µmol) was added dropwise. The solution was stirred
for 25 minutes, then quenched (sat. aq. NH4Cl dropwise), evaporated in vacuo
until thick, diluted in water and extracted into Et2O (× 3). Washing (brine),
drying (MgSO4), evaporation in vacuo, and flash column chromatography (6-
10% acetone/CH2Cl2 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 CHAPTER 5. EXPERIMENTAL.
35
O
OH
O
H HOO
O
O O
TLC: hRf 1a 35
10% acetone/CH2Cl2 57 25
FTIR (film): ν̃max 3446, 2953, 2256, 1730, 1503, 1438, 1379, 1272, 1237, 1203,
1161, 1123, 1053, 1010, 978, 914, 875, 785, 732, 647 cm−1;
1H NMR, major (17β) anomer (CDCl3): δ 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, CO2CH 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-11α), 2.07 (1H, d, J =
0.9 Hz, H-10), 1.80 (1H, dq, J = 13.9, 3.3 Hz, H-7β), 1.70 (1H, dt, J = 13.5,
3.2 Hz, H-6α), 1.64-1.57 (1H, m, H-6β), 1.38 (3H, s, H-20), 1.43-1.32 (1H, m,
H-7α), 1.21 (1H, ddd, J = 13.2, 11.6, 0.9 Hz, H-11β), 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
13C NMR, major (17β) anomer (CDCl3): δ 202.5 (C, C-1), 171.9 (C, C-18),
169.9 (C, OCOCH3), 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, CO2CH3), 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, OCOCH3), 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 purification;
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
36d
O
O
O
H HOHO
O
O O
TLC: hRf 36h Orth.Ac. 36d
Et2O 33 60 29
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.41 (1H, t, J = 1.7 Hz, H-15),
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, CO2CH 3), 2.49 (1H, dd, J
= 13.1, 5.6 Hz, H-11α), 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);
1H NMR [(CD3)2CO] matched values previously reported for “CDCl3”.27
Acetylation of 36d with Ac2O/pyridine27 and purification 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)34-
19(15-7-9-31-12-15)11-25(16,4)21(24)20(18)33-14(2)27/h7,9,12,16-21H,6,8,10-
11H2,1-5H3/t16-,17-,18-,19-,20-,21-,24-,25-/m0/s1
206 CHAPTER 5. EXPERIMENTAL.
36c
O
O
O
H HOO
O
O
O O
TLC: hRf 36d 36c
10% acetone/CH2Cl2 33 77 (developed in KMnO4).
HPLC: tR = 18.5 min (40% EtOAc/petrol).
1H NMR (CDCl3): δ 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, CO2CH 3),
2.43 (1H, dd, J = 13.3, 5.4 Hz, H-11α), 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 H2O), 1.69 (1H, ddd, J = 13.4, 11.6, 0.9 Hz, H-11β), 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 (assign-
ments differ). 1H NMR [(CD3)2CO]23 and 13C NMR (CDCl3)27 matched pre-
viously reported values.
5.3.8 (4R)-3,4-Dihydrosalvinorin E (36e).
Following the published procedure,23 acetylation in Ac2O/pyridine27 of 36h
(which had been purified 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 207
36e
O
O
O
H HOH
O
O
O O
TLC: hRf 36h 36e
50% EtOAc/petrol 20 48 (developed in KMnO4).
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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,
CO2CH 3), 2.41 (1H, dd, J = 13.2, 5.4 Hz, H-11α), 2.31 (1H, q, J = 12.6 Hz,
H-3α), 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 12
= 11 Hz, OH ), 1.82 (1H, dddd, J = 12.5,
4.9, 2.5, 1.1 Hz, H-3β), 1.76 (1H, dq, J = 13.3, 3.2 Hz, H-7β), 1.74-1.69 (1H,
m, H-6a), 1.64 (1H, ddd, J = 13.0, 11.6, 0.9 Hz, H-11β), 1.45 (3H, s, H-20),
1.38 (3H, s, H-19), 1.38-1.26 (2H, m, H-6b,7α), 1.00 (1H, br s, H-10);
these data are consistent with incomplete data published previously.23 13C
NMR (CDCl3) matched previously reported values.27
5.3.9 (4R)-Dideacetyl-3,4-dihydrosalvinorin C (36h).
A slight modification of a published procedure was used:23, 260 to 1a (37.6 mg,
86.9 µmol) and NaBH4 (4.4 mg, 116 µmol) was added CH2Cl2 (200 µL), fol-
lowed by EtOH (3 mL), and the cloudy solution stirred under Ar at 40 ◦C.
After 4 h, TLC (Et2O) indicated completion. The solution was cooled to 0◦C, and 0.5% H2SO4/MeOH added dropwise until effervescence ceased. The
solution was concentrated to ≈ 500 µL in vacuo, then partitioned between
brine (acidified with 10% HCl) and CH2Cl2 (× 3). Drying (MgSO4), evapora-
tion in vacuo and flash column chromatography (40-66% EtOAc/petrol) gave
36h15, 23 (19.1 mg, 56%);
208 CHAPTER 5. EXPERIMENTAL.
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
36h
O
O
O
H HOH
HO
O O
TLC: hRf 1a 36h 38h
Et2O 45 14 25
HPLC: tR = 8.7 min (EtOAc).
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.42 (1H, t, J = 1.7 Hz, H-15),
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, CO2CH 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-11α), 2.25-2.07 (4H, m),
1.91 (2H, br s, w 12
= 70 Hz, OH ), 1.77-1.69 (3H, m), 1.62 (1H, ddd, J = 13.3,
11.6, 1.0 Hz, H-11β), 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-6α), 0.92 (1H, d, J = 1.9 Hz, H-10);
1H NMR ([CD3]2CO) 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 µmol)
and NaHCO3 (30.1 mg, 358 µmol), 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 flash column chromatography (30% - 50% EtOAc/petrol
gradient) monitored by TLC (Et2O) 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
37a
O
O
O
H HOO
O
O O
8
TLC: hRf 1a 37a
Et2O 39 50
vanillin/H2SO4 purple blue
[α]13D −53 (c 0.6, CHCl3);
FTIR (film): ν̃max 3018, 2951, 2883, 1732, 1504, 1450, 1437, 1375, 1323, 1238,
1202, 1161, 1124, 1084, 1047, 1024, 997, 970, 937, 876, 783, 756, 667, 601 cm−1;
1H NMR (CDCl3): δ 7.43 (1H, br s, H-16), 7.38 (1H, t, J = 1.7 Hz, H-15),
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, CO2CH 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-11α), 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-7β), 2.15 (3H, s, OCOCH 3), 2.00 (1H, td, J = 13.7, 3.9 Hz, H-6β),
1.83 (1H, tdd, J = 14.2, 5.0, 3.9 Hz, H-7α), 1.62 (3H, s, H-20), 1.54 (1H, dt,
J = 13.7, 3.4 Hz, H-6α), 1.50 (1H, dd, J = 15.0, 12.0 Hz, H-11β), 1.07 (3H,
s, H-19);
13C NMR (CDCl3): δ 202.3 (C, C-1), 173.4 (C, C-17), 171.8 (C, C-18), 169.8
(C, OCOCH3), 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, C-
4), 51.8 (CH3, CO2CH3), 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,
OCOCH3), 17.6 (CH2, C-7), 15.2 (CH3, C-19);
210 CHAPTER 5. EXPERIMENTAL.
HRESIMS: [M + Na]+ m/z 455.1683 (calcd for C23H28O8Na+, 455.1676).
5.3.11 8-epi-Salvinorin B (37b).
MeOH (2 mL) was refluxed under Ar for 15 minutes, and cooled to rt. 1a (32.3
mg, 74.7 µmol) and Na2CO3 (30.4 mg, 245 µmol) 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 CH2Cl2 (× 4). Drying (MgSO4) and evaporation in vacuo gave an off-
white 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
37b
O
O
O
H HOHO
O O
8
TLC: hRf 1a 1b 37b
Et2O 40 24 40
50% EtOAc/petrol 37 23 37
10% acetone/CH2Cl2 69 52 54
vanillin/H2SO4 purple purple blue
[α]13D −27 (c 0.3, CHCl3);
FTIR (film): ν̃max 3475, 2952, 1728, 1504, 1438, 1386, 1274, 1201, 1156, 1120,
1051, 1026, 997, 970, 876, 786, 755 cm−1;
5.3. SYNTHESIS 211
1H NMR (CDCl3): δ 7.44 (1H, m, H-16), 7.40 (1H, t, J = 1.8 Hz, H-15), 6.37
(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, CO2CH 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-3β), 2.42 (1H, dd, J = 14.8, 2.2 Hz, H-11α), 2.23 (1H, d,
J = 1.3 Hz, H-10), 2.21 (1H, dq, J = 14.4, 3.1 Hz, H-7β), 2.01 (1H, td, J =
13.5, 11.9 Hz, H-3α), 1.99 (1H, td, J = 13.8, 0.8 Hz, H-6β), 1.85 (1H, tdd, J
= 14.0, 4.8, 3.7 Hz, H-7α), 1.65 (3H, s, H-20), 1.54 (1H, dt, J = 13.5, 3.5 Hz,
H-6α), 1.46 (1H, ddd, J = 14.8, 11.7, 0.7 Hz, H-11β), 1.06 (3H, s, H-19);
these data are inconsistent with those of Harding et al82 and Lee et al.271
13C NMR (CDCl3): δ 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, C-
2), 70.0 (CH, C-12), 63.7 (CH, C-10), 52.4 (CH, C-4), 51.7 (CH3, CO2CH3),
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 C21H26O7Na+, 413.1571).
5.3.12 8-epi-Salvinorin C (37c).
37d (1.7 mg, 3.9 µmol), Ac2O (0.25 mL) and catalytic DMAP (�1 mg)
were stirred in pyridine (0.5 mL) at 50 ◦C for 90 min, when TLC (10%
acetone/CH2Cl2, visualised with KMnO4/H2SO4 dip) showed no starting ma-
terial. In a separate flask, 37e ( 2.8 mg, 6.5 µmol) 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/CH2Cl2) 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)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+/m1/s1
212 CHAPTER 5. EXPERIMENTAL.
37c
O
O
O
H HOO
O
O
O O
8
TLC: hRf 37c 37d 37e
3% acetone/CH2Cl2 35 8 26
vanillin/H2SO4 blue blue blue
[α]23D +9 (c 0.2, CH2Cl2);
FTIR (film): ν̃max 2926, 2854, 1741, 1716, 1645, 1556, 1504, 1435, 1371, 1332,
1229, 1198, 1172, 1159, 1093, 1032, 962, 875, 802 cm−1;
1H NMR (CDCl3): δ 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, CO2CH 3), 2.49 (1H, dd, J = 5.3, 2.3 Hz, H-8), 2.33 (1H, dt, J =
13.4, 3.4 Hz, H-6α), 2.26-2.16 (1H, m, H-7β), 2.14 (1H, dd, J = 14.3, 1.6 Hz,
H-11α), 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-7α), 1.70 (3H, s, H-19), 1.61-1.50 (m, H-11β, 6β, H2O),
1.35 (3H, s, H-20), 1.25 (1H, br s, H-10);
13C NMR (CDCl3): δ 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 C25H30O9Na+, 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 flash column chromatography (5-10% acetone/CH2Cl2) 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
37d
O
O
O
H HOHO
O
O O
8
TLC: See table on the preceding page.
[α]23D +15 (c 0.1, CH2Cl2);
FTIR (film): ν̃max 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 cm−1;
1H NMR (500 MHz, CDCl3): δ 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, CO2CH 3), 2.48 (1H, dd, J
= 5.3, 2.3 Hz, H-8), 2.30 (1H, dt, J = 13.5, 3.8 Hz, H-6α), 2.22 (1H, dtd, J
= 14.7, 3.5, 2.1 Hz, H-7β), 2.17 (1H, dd, J = 14.1, 1.5 Hz, H-11α), 2.13 (3H,
s, OCOCH 3), 1.96 (1H, tdd, J = 14.2, 5.3, 3.9 Hz, H-7α), 1.70-1.64 (1H, m,
H-11β), 1.67 (3H, s, H-19), 1.62-1.52 (m, H-6β, H2O), 1.37 (3H, s, H-20), 1.25
(1H, br s, H-10);
13C NMR (CDCl3): δ 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 C23H28O8Na+, 455.1676).
214 CHAPTER 5. EXPERIMENTAL.
5.3.14 8-epi-Salvinorin E (37e).
37h (6.6 mg, 16.9 µmol) was dissolved in pyridine (2 mL) and Ac2O (250
µL), and stirred at rt for 3 h, when TLC (10% acetone/CH2Cl2) showed no
starting material. H2O (10 mL) was added, and the solution extracted with
Et2O (× 2). The pooled organic layers were washed (sat. NaHCO3, sat.
CuSO4, and brine). Drying (MgSO4), evaporation in vacuo and flash column
chromatography (80% Et2O/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
37e
O
O
O
H HOH
O
O
O O
8
TLC: See tables on page 212 and on the next page.
[α]22D +14 (c 0.1, CH2Cl2);
FTIR (film): ν̃max 3514, 2951, 1737, 1722, 1503, 1461, 1435, 1373, 1322, 1257,
1227, 1199, 1155, 1126, 1092, 1047, 1031, 931, 900, 837, 843, 806, 780, 735
cm−1;
1H NMR (500 MHz, CDCl3): δ 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, CO2CH 3), 2.49 (1H, dd, J =
5.2, 2.1 Hz, H-8), 2.36 (1H, dt, J = 13.3, 3.6 Hz, H-6α), 2.21 (1H, dtd, J =
14.5, 3.7, 2.1 Hz, H-7β), 2.14 (3H, s, OCOCH 3), 2.12 (1H, dd, J = 13.7, 1.5
Hz, H-11α), 1.99 (1H, tt, J = 14.3, 4.4 Hz, H-7α), 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-11β), 1.62
(3H, s, H-20), 1.54 (1H, td, J = 13.1, 3.4 Hz, H-6β), 1.30 (1H, t, J = 0.9 Hz,
H-10);
13C NMR (CDCl3): δ 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 C23H28O8Na+, 455.1676).
5.3.15 8-epi-Dideacetylsalvinorin C (37h).
1c (11.0 mg, 23.2 µmol) and KCN∗ (7.5 mg, 115 µmol) were dissolved in MeOH
(1.5 mL) and stirred at rt for 30 min, when TLC (10% acetone/CH2Cl2) showed
no starting material. The solution was evaporated in vacuo, and the residue
partitioned between H2O and CH2Cl2 (× 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
37h
O
O
O
H HOH
HO
O O
8
TLC: hRf 1c 37h 1h
10% acetone/CH2Cl2 63 23 18
50% EtOAc/petrol - 28 19
vanillin/H2SO4 purple blue purple
∗Use of this poisonous reagent is not recommended. K2CO3 and other bases are equallyeffective.
216 CHAPTER 5. EXPERIMENTAL.
mp (from acetone) 243-245 ◦C;
[α]22D +15 (c 0.04, CH2Cl2);
FTIR (film): ν̃max 3461, 2952, 2926, 1716, 1507, 1459, 1437, 1378, 1259, 1228,
1200, 1180, 1156, 1095, 1052, 875, 803 cm−1;
1H NMR (CDCl3): δ 7.48 (1H, m, H-16), 7.42 (1H, t, J = 1.6 Hz, H-15), 6.44
(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, CO2CH 3), 2.48 (1H, dd, J = 5.5, 2.2
Hz, H-8), 2.23 (1H, dt, J = 13.5, 2.2 Hz, H-6α), 2.23-2.17 (1H, m, H-7β), 2.15
(1H, dd, J = 13.9, 1.5 Hz, H-11α), 1.98 (1H, tdd, J = 14.5, 5.3, 3.8 Hz, H-7α),
1.81 (2H, br s, w 12≈ 50 Hz, OH ), 1.68 (3H, s, H-19), 1.64 (3H, s, H-20), 1.62
(1H, dd, J = 13.8, 12.1 Hz, H-11β), 1.51 (1H, td, J = 12.9, 3.8 Hz, H-6β),
1.21 (1H, br s, H-10);
13C NMR (CDCl3): δ 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 C21H26O7Na+, 413.1571).
5.3.16 Salvinorin B formate (46).
A mixture of Ac2O (0.25 mL) and HCO2H (0.7 mL) was stirred at 45 ◦C
for 40 minutes. Meanwhile 1b (18.0 mg, 46.1 µmol) was added to dry pyri-
dine (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/CH2Cl2, 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
flash column chromatography (2-4% MeOH/CH2Cl2 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
46
O
O
O
H HOO
O
O O
TLC: hRf 1b 46
20% acetone/CH2Cl2 65 85
[α]26D −54 (c 0.6, CH2Cl2);
FTIR (film): ν̃max 3147, 2952, 2854, 1726, 1504, 1451, 1439, 1396, 1372, 1278,
1224, 1163, 1108, 1072, 1024, 1008, 941, 899, 875, 786, 735, 701 cm−1;
1H NMR (CDCl3): δ 8.14 (1H, s, CHO), 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, CO2CH 3),
2.79-2.74 (1H, m, H-4), 2.51 (1H, dd, J = 13.5, 5.2 Hz, H-11α), 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-7α), 2.08 (1H,
dd, J = 11.5, 3.0 Hz, H-8), 1.80 (1H, ∼dt, J ≈ 12.9, 2.9 Hz, H-6α), 1.65 (1H,
tdd, J = 13.3, 11.8, 3.2 Hz, H-7β), 1.63-1.58 (1H, m, H-6β), 1.60 (1H, ddd, J
= 13.5, 11.7, 0.8 Hz, H-11β), 1.46 (3H, s, H-20), 1.13 (3H, s, H-19);
13C NMR (CDCl3): δ 200.8 (C, C-1), 171.4 (C, C-18), 171.0 (C, C-17), 159.4
(CH, CHO), 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, CO2CH3), 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 C22H26O8Na+, 441.1520).
218 CHAPTER 5. EXPERIMENTAL.
5.3.17 Dideacetylsalvinorin C 2-O-(4-bromobenzoate) (47).
To a solution of 1h (3.8 mg, 9.7 µmol), 4-bromobenzoyl chloride (8.7 mg, 39.6
µmol) and DMAP (3.1 mg, 25.4 µmol) in dry CH2Cl2 (1.5 mL) was added
NEt3 (100 µL). The bright yellow solution was stirred under Ar for 90 min,
when TLC (Et2O, visualised in KMnO4) indicated completion. The solution
was diluted in Et2O, 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
47
O
O
O
H HOH
O
O O
O
Br
TLC: hRf 1h 47
10% acetone/CH2Cl2 40 78
mp (from cold Et2O) 161-166 ◦C;
[α]18D +86 (c 0.4, CH2Cl2);
FTIR (film): ν̃max 3509, 2950, 1717, 1590, 1504, 1484, 1434, 1398, 1269, 1233,
1175, 1143, 1104, 1071, 1012, 946, 875, 849, 786, 758, 679 cm−1;
1H NMR (CDCl3): δ 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, CO2CH 3), 2.54 (1H, dt, J = 13.4, 3.6 Hz, H-6α), 2.48 (1H,
5.3. SYNTHESIS 219
dd, J = 13.3, 6.1 Hz, H-11α), 2.17 (1H, dd, J = 12.0, 3.5 Hz, H-8), 2.12 (1H,
dq, J = 12.0, 3.5 Hz, H-7β), 1.96 (1H, br s, w 12
= 14.6 Hz, OH ), 1.92-1.80
(1H, m, H-7α), 1.76 (3H, s, H-19), 1.67 (1H, dd, J = 12.9, 11.2 Hz, H-11β),
1.47 (3H, s, H-20), 1.38 (1H, s, H-10), 1.22 (1H, td, J = 13.1, 3.7 Hz, H-6β);
13C NMR (CDCl3): δ 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 C28H29O8BrNa+,
595.0938, 597.0917).
5.3.18 17-Deoxysalvinorin A (49).
Et3SiH (20 µL, 125 µmol) was added to a stirred solution of lactol 35 (18.3
mg, 42.1 µmol) in dry CH2Cl2 (1 mL) under Ar. The solution was cooled to
0 ◦C, and BF3·Et2O (10 µL, 79 µmol) was added. The light brown solution
was stirred at 0 ◦C for 2 h, when TLC (10% acetone/CH2Cl2) showed no 35.
The reaction was quenched (0.5 mL sat. NaHCO3), and partitioned between
Et2O and brine. Drying (MgSO4), evaporation in vacuo and flash column
chromatography (0-4% acetone/CH2Cl2 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,5,7,9-10,12H2,1-
4H3/t15-,16-,17-,18-,20-,22-,23-/m0/s1
49
O
O
H HOO
O
O O
220 CHAPTER 5. EXPERIMENTAL.
TLC: hRf 35 49 50
10% acetone/CH2Cl2 38 72 72
2% acetone/CH2Cl2 - 24 42
[α]20D −81 (c 0.4, CH2Cl2);
FTIR (film): ν̃max 2950, 2927, 2857, 1730, 1506, 1236, 1201, 1161, 1107, 1042,
1018, 932, 889, 875, 784, 736 cm−1;
1H NMR (CDCl3): δ 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, CO2CH 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-11α), 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-11β), 1.08 (3H, s, H-19);
13C NMR (CDCl3): δ 202.5 (C, C-1), 171.9 (C, C-18), 169.9 (C, OCOCH3),
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, CO2CH3), 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, OCOCH3), 19.6
(CH2, C-7), 16.8 (CH3, C-19), 13.7 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 441.1881 (calcd for C23H30O7Na+, 441.1884).
5.3.19 8,17-Didehydro-17-deoxysalvinorin A (50).
Et3SiH (25 µL, 156 µmol) and Amberlyst 15 resin (22 mg) were added to
a solution of lactol 35 (15.1 mg, 34.7µmol) in CH2Cl2 (1 mL). The sealed
flask was stirred at rt for 24 h. Filtration, evaporation and flash column
chromatography (1% acetone/CH2Cl2) 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
50
O
O
HO
O
O
O O
TLC: See table on the facing page.
[α]23D −60 (c 0.5, CH2Cl2);
FTIR (film): ν̃max 2927, 2855, 1731, 1664, 1504, 1437, 1380, 1275, 1237, 1203,
1164, 1113, 1098, 1044, 1024, 1000, 962, 944, 875, 786, 737 cm−1;
1H NMR (CDCl3): δ 7.38 (1H, br s, H-16), 7.36 (1H, t, J = 1.9 Hz, H-15),
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, CO2CH 3), 2.74-2.69 (1H, m, H-4), 2.33 (1H, dd, J = 13.6, 2.1 Hz, H-11α),
2.32-2.24 (1H, m, H-7α), 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-7β), 1.69
(1H, ddd, J = 13.2, 4.4, 2.6 Hz, H-6α), 1.52 (3H, s, H-20), 1.50 (1H, td, J =
13.2, 4.6 Hz, H-6β), 1.40 (1H, ddd, J = 13.6, 11.6, 0.8 Hz, H-11β), 1.15 (3H,
s, H-19);
13C NMR (CDCl3): δ 202.9 (C, C-1), 172.0 (C, C-18), 169.9 (C, OCOCH3),
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, CO2CH3), 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, OCOCH3), 15.4 (CH3, C-19);
HRESIMS: [M + Na]+ m/z 439.1728 (calcd for C23H28O7Na+, 439.1727).
222 CHAPTER 5. EXPERIMENTAL.
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 µmol) in
50% CH2Cl2/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/CH2Cl2). The solution was filtered through
diatomite filter 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
51
O
O
O
H HOO
O
O O
HH13
TLC: hRf 51 1a
20% acetone/CH2Cl2 46 74
HPLC: tR (min) 51
EtOAc 11.4, 11.75
[α]19D −39 (c 0.4, CHCl3);
FTIR (film): ν̃max 2953, 2879, 1730, 1452, 1438, 1379, 1277, 1235, 1203, 1165,
1110, 1078, 1050, 1005, 951, 912, 895, 755 cm−1;
1H NMR (CDCl3): δ 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, CO2CH 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-11α), 2.17 (3H, s, OCOCH 3), 2.17-2.12 (1H,
m, H-7α), 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-6α), 1.65-1.56 (1H, m, H-7β), 1.60-1.50 (1H,
m, H-6β), 1.35 (3H, s, H-20), 1.26 (1H, ddd, J = 13.3, 11.7, 0.8 Hz, H-11β),
1.09 (3H, s, H-19);
13C NMR (CDCl3): δ 202.0 (C, C-1), 171.5 (C, C-18), 171.3 (C, C-17), 169.9
(C, OCOCH3), 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, CO2CH3), 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, OCOCH3), 18.1 (CH2,
C-7), 16.3 (CH3, C-19), 15.1 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 459.1984 (calcd for C23H32O8Na+, 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 µmol) in minimal
CH2Cl2 (≈ 250 µL), and oxygen bubbled through the resulting orange solution
for 20 minutes, when TLC (Et2O) 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 CH2Cl2 (× 3). Drying (MgSO4), evapora-
tion in vacuo and flash column chromatography (20 – 50% Et2O/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 CHAPTER 5. EXPERIMENTAL.
59
O
O
O
HOH
O
O O
TLC: hRf 1a 1b 59
Et2O 54 42 69
[α]14D +58 (c 0.7, CH2Cl2);
UV (CH3CN): λmax (log ε) 215 (4.30), 249 (3.77), 324 (3.57) nm;
FTIR (film): ν̃max 3373, 3149, 2955, 1726, 1651, 1601, 1504, 1456, 1435, 1408,
1380, 1331, 1244, 1203, 1162, 1068, 1022, 911, 875, 793, 736, 703 cm−1;
1H NMR (CDCl3): δ 7.48 (1H, br s, H-16), 7.41 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH 3), 3.11 (1H, ddd, J =
14.8, 2.9, 1.2 Hz, H-11α), 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-11β), 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);
13C NMR (CDCl3): δ 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, CO2CH3), 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 C21H22O7Na+, 409.1258).
The aqueous layer was acidified with 10% HCl until opaque white, then ex-
tracted into CH2Cl2 (× 3), which was dried (MgSO4) and filtered. MeOH (10
mL) and Me3SiCHN2 in Et2O (2.0 M, 200 µL) 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
identified, based on HRESIMS and NMR (1H 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
60c(tentative)
O
O
O
O
HO
O O
O
O
10
H
TLC: hRf 59 60a 60b 60c
80% Et2O/petrol 50 41 41 41
HPLC: tR (min) 60c 60a 60b
36% EtOAc/petrol 17.0 17.4 18.3
1H NMR (CDCl3): δ 7.46 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH3), 3.67 (3H, s, CO2CH3), 3.65 (3H, s, CO2CH3), 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-11α), 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-7α), 1.86 (1H, dd, J = 14.9, 11.8 Hz, H-11β), 1.78 (1H, dq, J = 14.4, 3.8
Hz, H-7β), 1.60-1.41 (2H, m, H-6 [obscured by H2O]), 1.41 (3H, s, H-19), 1.20
(3H, s, H-20);
HRESIMS: [M + Na]+ m/z 473.1781 (calcd for C23H30O9Na+, 473.1782).
Further elution gave 60a;
226 CHAPTER 5. EXPERIMENTAL.
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
60a
O
O
O
O
HO
O O
O
O
H
TLC, HPLC: see table on the previous page.
[α]24D +6 (c 0.1, CH2Cl2);
FTIR (film): ν̃max 2953, 1732, 1506, 1436, 1393, 1373, 1261, 1226, 1202, 1163,
1137, 1079, 1025, 875, 792 cm−1;
1H NMR (500 MHz, CDCl3): δ 7.44 (1H, br s, H-16), 7.43 (1H, t, J = 1.9 Hz,
H-15), 6.41 (1H, m, H-14), 5.46 (1H, dd, J = 11.6, 5.4 Hz, H-12), 3.69 (3H, s,
CO2CH3), 3.68 (3H, s, CO2CH3), 3.65 (3H, s, CO2CH3), 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-7β, 8), 2.01 (1H,
dd, J = 13.7, 5.5 Hz, H-11α), 1.84 (1H, ddd, J = 13.7, 12.0, 0.8 Hz, H-11β),
1.74-1.63 (2H, m, H-6a, 7α), 1.52-1.48 (1H, m, H-6b), 1.38 (3H, s, H-19), 1.30
(3H, s, H-20);
13C NMR (CDCl3): δ 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, CO2CH3), 51.9
(CH, C-4), 51.7 (CH3, CO2CH3), 51.4 (CH3, CO2CH3), 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 C23H30O9Na+, 473.1782).
5.3. SYNTHESIS 227
Further elution gave 60b;
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
60b
O
O
O
O
HO
O O
O
O
8
H
TLC, HPLC: see table on page 225.
[α]23D +4 (c 0.4, CH2Cl2);
FTIR (film): ν̃max 2953, 1733, 1504, 1436, 1392, 1361, 1247, 1200, 1166, 1091,
1062, 1025, 998, 968, 909, 875, 849, 794, 735 cm−1;
1H NMR (500 MHz, CDCl3): δ 7.47 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz,
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, CO2CH3), 3.66 (3H, s, CO2CH3), 3.64 (3H, s, CO2CH3),
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-7β), 2.02 (1H, dd, J = 14.5, 11.9 Hz,
H-11β), 1.94-1.86 (2H, m, H-6β, 7α), 1.80 (1H, dd, J = 14.4, 2.1 Hz, H-11α),
1.48 (3H, s, H-20), 1.32-1.27 (1H, m, H-6α), 1.29 (3H, s, H-19);
13C NMR (CDCl3): δ 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, CO2CH3), 51.7 (CH3,
CO2CH3), 51.7 (CH, C-4), 51.5 (CH3, CO2CH3), 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 C23H30O9Na+, 473.1782).
228 CHAPTER 5. EXPERIMENTAL.
5.3.22 NaBH4 reduction of 59.
Enedione 59 (41.3 mg, 107 µmol) and NaBH4 (10 mg, 264 µmol) were dissolved
in CH2Cl2 (500 µL), 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 (Et2O) indicated completion. The solution was cooled to 0 ◦C, and 0.5%
H2SO4/MeOH added dropwise until effervescence ceased. The solution was
concentrated to ≈ 500 µL in vacuo, then partitioned between brine (acidified
with 10% HCl) and CH2Cl2 (× 3). Drying (MgSO4), evaporation in vacuo
and flash column chromatography (70-100% Et2O/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
38h
O
O
O
H HOH
HO
O O
8
TLC: hRf 38h 59
Et2O 53 69
1H NMR (CDCl3): δ 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, CO2CH 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-7α), 1.73-1.53 (m, obscured by H2O &
OH ), 1.66 (3H, s, H-20), 1.32 (3H, s, H-19), 0.90 (1H, d, J = 1.6 Hz, H-10);
1H NMR ([CD3]2CO) and 13C 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 nBuLi 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 nBuLi was added, swirling the flask
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 flask under Ar for up to a year without losing activity.
1a (42.2 mg, 97.6 µmol) and LiSEt (13.7 mg, 201 µmol) under Ar were dis-
solved in DMPU (1 mL). The yellow solution was stirred at 55 ◦C for 23
h, when TLC (1% H2SO4/10% MeOH/CH2Cl2) 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 acidified at 0 ◦C with 10% HCl, then
extracted into CH2Cl2 (× 3). Drying (MgSO4) and evaporation in vacuo gave
an amber resin, which was treated with Ac2O (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. NH4Cl). Drying (MgSO4) and evaporation in vacuo gave the mixed
acids 67a (H-8α : β, ∼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,4,6,8-9H2,1-
3H3,(H,25,26)/t13u,14-,15-,16-,18-,21-,22-/m0/s1/f/h25H
67a
O
O
O
H HOO
O
O OH
230 CHAPTER 5. EXPERIMENTAL.
TLC: hRf 1a 1b 67a 77 78
1% H2SO4/10% MeOH/CH2Cl2 71 63 0-20
1% NEt3/EtOAc 0-30 56 56
Et2O 20 30
vanillin/H2SO4 purple purple purple blue
1H NMR (CDCl3): δ 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-11α), 2.45 (1H, dd, J = 4.9, 2.0 Hz, H-8∗), 2.37 (1H, dd, J = 14.8, 2.0
Hz, H-11α∗), 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-7α∗), 1.73 (1H, dtd, J = 13.7, 3.5, 0.8 Hz, H-6α), 1.67-1.52 (m), 1.63 (3H,
s, H-20∗), 1.50 (1H, dd, J = 15.1, 12.1 Hz, H-11β∗), 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 µmol) were stirred in dry THF (1 mL) at
rt, under Ar, in a flask fitted with a reflux condenser. BH3·THF (1.0 M, 110 µL,
110 µmol) 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 CH2Cl2 (×3). Drying
(MgSO4), evaporation in vacuo and flash column chromatography (10-25%
acetone/Et2O gradient) monitored by TLC (Et2O) 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(13-
5-7-27-11-13)9-22(15,3)19(21)18(16)25/h5,7,11,14-17,19,23H,4,6,8-10H2,1-
3H3/t14-,15+,16-,17-,19-,21-,22-/m0/s1
∗Unnatural (H-8α) epimer.
5.3. SYNTHESIS 231
78
O
O
O
H HOO
O
OH
8
TLC: see table on the preceding page.
[α]24D −31 (c 0.3, CH2Cl2);
FTIR (film): ν̃max 3547, 3148, 2942, 2880, 1725, 1504, 1468, 1450, 1440, 1376,
1238, 1202, 1159, 1085, 1050, 1024, 950, 934, 875, 802, 784, 734, 703 cm−1;
1H NMR (CDCl3): δ 7.44 (1H, br s, H-16), 7.38 (1H, t, J = 1.8 Hz, H-15),
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-11α), 2.24 (1H, d, J = 0.9 Hz, H-10), 2.22-2.09 (2H, m, H-6β,7β), 2.15 (3H,
s, OCOCH 3), 1.91-1.72 (4H, m, H-3b, 4, 6α, 7α), 1.64 (3H, s, H-20), 1.52 (1H,
dd, J = 14.8, 12.0 Hz, H-11β), 0.92 (3H, s, H-19);
13C NMR (CDCl3): δ 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 C22H28O7Na+, 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(13-
5-7-27-11-13)9-22(15,3)19(21)18(16)25/h5,7,11,14-17,19,23H,4,6,8-10H2,1-
3H3/t14-,15-,16-,17-,19-,21-,22-/m0/s1
232 CHAPTER 5. EXPERIMENTAL.
77
O
O
O
H HOO
O
OH
TLC: see table on page 230.
[α]25D −19 (c 0.3, CH2Cl2);
FTIR (film): ν̃max 3468, 2944, 1727, 1504, 1452, 1376, 1237, 1162, 1095, 1053,
1023, 952, 875, 794, 735, 703 cm−1;
1H NMR (CDCl3): δ 7.41 (1H, br s, H-16), 7.39 (1H, t, J = 1.9 Hz, H-15), 6.38
(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-11α), 2.17 (1H, d, J = 0.9 Hz, H-10), 2.16
(3H, s, OCOCH 3), 2.20-2.13 (1H, m, H-7β), 2.06 (1H, dd, J = 12.1, 3.0 Hz,
H-8), 1.99 (1H, dt, J = 13.3, 3.3 Hz, H-6α), 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-7α),
1.58 (1H, ddd, J = 13.3, 11.7, 0.9 Hz, H-11β), 1.45 (3H, s, H-20), 1.43 (1H,
td, J = 13.2, 3.7 Hz, H-6β), 0.96 (3H, s, H-19);
13C NMR (CDCl3): δ 203.6 (C, C-1), 171.3 (C, C-17), 170.0 (C, OCOCH3),
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, OCOCH3), 18.1 (CH2, C-7), 16.7 (CH3,
C-19), 15.3 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 427.1729 (calcd for C22H28O7Na+, 427.1727).
5.3. SYNTHESIS 233
5.3.24 1-Deoxysalvinorin A (81a).
A solution of NaBH4 (11.6 mg, 307 µmol) and 1a (108 mg, 249 µmol) in dry
EtOH (9 mL)/CH2Cl2 (2 mL) was stirred under Ar at 40 ◦C. The cloudy
solution gradually cleared. At 4 h, TLC (Et2O) indicated completion (for hRf
values, see table on page 208). The solution was evaporated in vacuo, and
the residue partitioned between brine and CH2Cl2 (× 3). Drying (MgSO4)
and evaporation in vacuo gave an approximately equal mixture of 36h23 and
38h27 (total 103 mg), which was used without purification.
This crude mixture and 1,1’-thiocarbonyldiimidazole (118 mg, 662 µmol) were
dissolved in dry DMF and stirred under Ar at 90 ◦C for 6 h, when TLC
(40% acetone/ CH2Cl2) indicated completion. The cooled solution was diluted
in EtOAc/Et2O and washed (10% HCl × 3, then brine). Drying (MgSO4)
and evaporation in vacuo followed by flash column chromatography (1-5%
acetone/CH2Cl2 gradient) gave the cyclic thionocarbonates 80 (H-8α : β, ≈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
80
O
O
O
H HO
O
O O
S
TLC: hRf 36h/38h 80 83 81b 82b 81a
Et2O 47 40 47 74
10% acetone/CH2Cl2 55 70 20 20
40% acetone/CH2Cl2 62 86
vanillin/H2SO4 purple blue purple
234 CHAPTER 5. EXPERIMENTAL.
1H NMR (CDCl3): δ 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, CO2CH 3), 3.67 (3H,
s, CO2CH 3∗), 2.56 (1H, dd, J = 5.3, 1.9 Hz, H-8∗), 2.45 (1H, dd, J = 12.8,
5.2 Hz, H-11α), 2.27 - 2.10 (m), 1.94 (1H, tdd, J = 14.2, 4.8, 3.9 Hz, H-7α∗),
1.82 - 1.69 (m), 1.71 (1H, dd, J = 13.5, 11.0 Hz, H-11β), 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∗);
13C NMR (CDCl3): δ 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.
nBu3SnH 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 nBu3SnH408 (150 µL, 550 µmol) and AIBN (5.4 mg, 33 µmol) 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/ CH2Cl2)
indicated completion. The solution was cooled and loaded directly onto silica
gel, rinsing the flask with CH2Cl2 (× 2). Repeated flash column chromatog-
raphy (50-70% EtOAc/petrol gradient) monitored by TLC (Et2O) gave two
fractions, A and B. Fraction A was subjected to flash column chromatography
(10% acetone/ CH2Cl2) 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-8α) epimer.
5.3. SYNTHESIS 235
83
O
O
O
H HO
O
O O
O
TLC: see table on page 233.
[α]22D +5 (c 0.1, CH2Cl2) ;
FTIR (film): ν̃max 2952, 1797, 1727, 1502, 1452, 1441, 1370, 1276, 1229, 1200,
1163, 1073, 1043, 1023, 956, 875, 790, 736 cm−1;
1H NMR (CDCl3): δ 7.46 (1H, br s, H-16), 7.43 (1H, t, J = 1.9 Hz, H-15), 6.40
(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, CO2CH 3), 2.46-2.41 (2H, m), 2.27-2.10 (4H, m), 1.87-1.67 (m, overlapping
with H2O), 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);
13C NMR (CDCl3): δ 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 C22H26O8Na+, 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
82b
O
O
O
H HHO
O O
8
236 CHAPTER 5. EXPERIMENTAL.
TLC: see table on page 233.
[α]21D −4 (c 0.1, CH2Cl2);
FTIR (film): ν̃max 3432, 2948, 2871, 1728, 1502, 1450, 1438, 1390, 1363, 1277,
1255, 1199, 1167, 1091, 1054, 1022, 1000, 875, 840, 785, 727 cm−1;
1H NMR (CDCl3): δ 7.47 (1H, br s, H-16), 7.42 (1H, t, J = 1.7 Hz, H-15), 6.41
(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, CO2CH 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-7β), 2.12 (1H, dd, J = 13.2, 3.6 Hz,
H-4), 1.99 (1H, dd, J = 14.0, 1.6 Hz, H-11α), 1.91 (1H, dddd, J = 12.9, 5.0,
3.3, 1.9 Hz, H-1β), 1.87 - 1.58 (m), 1.81 (1H, dd, J = 13.0, 11.0 Hz, H-11β),
1.52 (1H, dt, J = 13.6, 3.2 Hz, H-6α), 1.43 (1H, td, J = 12.6, 10.7 Hz, H-1α),
1.28-1.24 (1H, m, H-6β), 1.24 (3H, s, H-19), 1.08 (3H, s, H-20), 1.00 (1H, dd,
J = 12.7, 2.2 Hz, H-10);
13C NMR (CDCl3): δ 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 C21H28O6Na+, 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
81b
O
O
O
H HHO
O O
TLC: see table on page 233.
5.3. SYNTHESIS 237
1H NMR (CDCl3): δ 7.43 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH 3), 2.32 (1H, dd, J = 13.4,
5.8 Hz, H-11α), 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-7β), 1.96-1.82 (3H, m, H-1β, 3), 1.74 (1H, dt, J = 13.6, 3.2
Hz, H-6α), 1.65-1.58 (m, overlapping with H2O), 1.44 (1H, td, J = 12.6, 10.9
Hz, H-1α), 1.28 (1H, td, J = 13.6, 4.0 Hz, H-6β), 1.10 (3H, s, H-19), 1.07 (3H,
s, H-20), 1.04 (1H, dd, J = 12.8, 2.5 Hz, H-10);
13C NMR (CDCl3): δ 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 Ac2O (0.3 mL) with a crystal of DMAP. The solution was stirred at rt
for 2.5 h, when TLC (Et2O, visualised in KMnO4) indicated completion, then
evaporated in vacuo. Flash column chromatography (60% Et2O/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
81a
22 101011
O
O
O
H HO
O
O O
TLC: see table on page 233.
[α]21D −8 (c 0.1, CH2Cl2);
FTIR (film): ν̃max 2952, 1729, 1503, 1450, 1434, 1365, 1319, 1245, 1202, 1153,
1099, 1024, 954, 901, 875, 783, 734 cm−1;
238 CHAPTER 5. EXPERIMENTAL.
1H NMR (CDCl3): δ 7.43 (1H, br s, H-16), 7.42 (1H, t, J = 1.8 Hz, H-15),
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, CO2CH 3), 2.28 (1H,
dd, J = 13.5, 5.9 Hz, H-11α), 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-7β), 2.05 (3H, s,
OCOCH 3), 1.95-1.89 (1H, m, H-1β), 1.92-1.83 (2H, m, H-3), 1.74 (1H, dt, J
= 13.5, 3.3 Hz, H-6α), 1.68-1.57 (1H, m, H-7α), 1.64 (1H, dd, J = 13.6, 11.5
Hz, H-11β), 1.50 (1H, td, J = 12.8, 11.2 Hz, H-1α), 1.31 (1H, td, J = 13.6,
3.7 Hz, H-6β), 1.11 (3H, s, H-19), 1.10 (1H, dd, J = 12.9, 2.2 Hz, H-10), 1.06
(3H, s, H-20);
13C NMR (CDCl3): δ 172.7 (C, C-18), 171.9 (C, C-17), 170.5 (C, OCOCH3),
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, CO2CH3), 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, OCOCH3),
18.2 (CH2, C-7), 15.0 (CH3, C-19), 14.6 (CH3, C-20);
HRESIMS: [M + Na]+ m/z 441.1893 (calcd for C23H30O7Na+, 441.1884).
Bibliography
[1] Photo by Max Hem, Melbourne.
[2] Esquivel, B.; Sánchez, A. A.; Aranda, E.: Natural products of agrochem-ical interest from Mexican Labiatae. In Phytochemicals and Phytophar-maceuticals, F. Shahidi; C.-T. Ho, eds., AOCS Press, Champaign, IL,pp. 371–385 2000.
[3] Rodriguez-Hahn, L.; Esquivel, B.; Cardenas, J.: Neo-clerodane diter-penoids from American Salvia species. Recent Adv. Phytochem. 1995,29, 311–332.
[4] Epling, C.; Játiva-M., C. D.: A New Species of Salvia from Mexico.Bot. Mus. Leafl. Harvard Univ. 1962, 20, 75–76. Unofficial transcriptionavailable online.URL http://www.sagewisdom.org/epling&jativa.html
[5] Wasson, R. G.: A New Mexican Psychotropic Drug from the Mint Fam-ily. Bot. Mus. Leafl. Harvard Univ. 1962, 20, 77–84. Unofficial transcrip-tion available online.URL http://www.sagewisdom.org/wasson1.html
[6] Reisfield, A. S.: The Botany of Salvia divinorum (Labiatae). Sida 1993,15, 349–366. HTML adaptation available online.URL http://www.sagewisdom.org/reisfield.html
[7] Ott, J.: Ethnopharmacognosy and Human Pharmacology of Salvia di-vinorum and Salvinorin A. Curare 1995, 18, 103–129. See pp. 112 (useby Mexican teenagers), 116 - 118 (pipiltzintzintli), 126 (failure of animaltesting). Unofficial transcription available online.URL http://www.sagewisdom.org/ott2.html
[8] Photo by Slava Olcheski, Virginia.
[9] Valdés, III, L. J. J.: The early history of Salvia divinorum. EntheogenRev. 2001, 10, 73–75,80.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). Unofficial transcription available online.URL http://www.druglibrary.org/schaffer/LSD/child.htm
[11] Valdés, III, L. J. J.; Díaz, J. L.; Paul, A. G.: Ethnopharma-cology of ska Maria Pastora (Salvia divinorum, Epling and Játiva-M.). J. Ethnopharmacol. 1983, 7, 287–312. http://dx.doi.org/10.1016/0378-8741(83)90004-1. Unofficial transcription available online.URL http://www.sagewisdom.org/valdes83.html
[12] Siebert, D. J.: The History of the First Salvia divinorum Plants Culti-vated Outside of Mexico. Entheogen Rev. 2003, 12, 117–118.URL http://www.sagewisdom.org/salviahistory.html
[13] Valdés, III, L. J. J.; Hatfield, G. M.; Koreeda, M.; Paul, A. G.: Studies ofSalvia divinorum (Lamiaceae), an Hallucinogenic Mint from the SierraMazateca in Oaxaca, Central Mexico. Econ. Bot. 1987, 41, 283–291.Unofficial transcription available online.URL http://www.sagewisdom.org/valdes87.html
[14] Valdés, III, L. J. J.: Salvia divinorum and the Unique Diterpene Hal-lucinogen, Salvinorin (Divinorin) A. J. Psychoactive Drugs 1994, 26,277–283. Unofficial transcription available online.URL http://www.sagewisdom.org/valdes94.html
[15] Valdés, III, L. J. J.: The Pharmacognosy of Salvia divinorum (EplingAnd Játiva-M): an Investigation of Ska María Pastora. Ph.D. thesis,University Of Michigan, Ann Arbor, MI 1983. ProQuest PublicationNumber: AAT 8402393 (document ID: 750217621). See p. 49 - 50 (“re-medial use”).URL http://wwwlib.umi.com/dissertations/fullcit/8402393
[16] Ref 15, pp. 80, 129.
[17] Valdés, III, L. J. J.: La Divina Pastora (The Divine Shepherdess), aBackground to Studies of Salvia divinorum. Salvia Divinorum Magazine2003, 1(2), 6–9.URL http://salvia.us/magazine.html
[18] Siebert, D. J.: Salvia divinorum and salvinorin A: New pharmaco-logic findings. J. Ethnopharmacol. 1994, 43, 53–56. 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 OtherHallucinogens of Mexico. Bot. Mus. Leafl. Harvard Univ. 1963, 20, 161–212. Unofficial transcription available online.URL http://pages.prodigy.com/GBonline/liquix.htm
BIBLIOGRAPHY 241
[20] Díaz, J. L.: Ethnopharmacology of Sacred Psychoactive Plants Used bythe 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 trans-Neoclerodane Diterpene from Salvia divinorum (Labiatae). J. Chem.Soc., Perkin Trans. 1 1982, pp. 2505–2508.URL http://dx.doi.org/10.1039/P19820002505
[23] Valdés, III, L. J. J.; Butler, W. M.; Hatfield, G. M.; Paul, A. G.; Koreeda,M.: Divinorin A, a Psychotropic Terpenoid, and Divinorin B from theHallucinogenic Mexican Mint Salvia divinorum. J. Org. Chem. 1984, 49,4716–4720.URL http://dx.doi.org/10.1021/jo00198a026
[24] Koreeda, M.; Brown, L.; Valdés, III, L. J. J.: The absolute stereochem-istry of salvinorins. Chem. Lett. 1990, pp. 2015–2018.
[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 Salviadivinorum. Org. Lett. 2005, 7, 3017–3020.URL http://dx.doi.org/10.1021/ol0510522
[26] Rodriguez-Hahn, L.; Esquivel, B.; Cardenas, J.: Clerodane diterpenesin Labiatae. Prog. Chem. Org. Nat. Prod. 1994, 63, 107–196. See p. 110(absolute stereochemistry).
[27] Valdés, III, L. J. J.; Chang, H. M.; Visger, D. C.; Koreeda, M.: Salvi-norin C, a New Neoclerodane Diterpene from a Bioactive Fraction ofthe Hallucinogenic Mexican Mint Salvia divinorum. Org. Lett. 2001, 3,3935–3937.URL http://dx.doi.org/10.1021/ol016820d
[28] Valdés, 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.; Man-gin, P.: Salvia divinorum: an hallucinogenic mint which might becomea new recreational drug in Switzerland. Forensic Sci. Int. 2000, 112,143–150.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, 243–287.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 andthe reversal of the currently accepted absolute configuration of clerodin.J. Chem. Soc., Chem. Commun. 1979, pp. 97–99.URL http://dx.doi.org/10.1039/C39790000097
[32] McCrindle, R.; Overton, K. H.: The chemistry of cyclic diterpenoids. InAdvances in Organic Chemistry: Methods and Results., R. A. Raphael;E. C. Taylor; H. Wynberg, eds., Wiley Interscience, New York, vol. 5,pp. 47–105 1965.
[33] Cane, D. E.: Isoprenoid Biosynthesis: Overview. In Comprehensive Nat-ural Products Chemistry, D. E. Cane; D. H. R. Barton; K. Nakanishi;O. Meth-Cohn, eds., Elsevier, New York, vol. 2, pp. 1–14 1999.URL http://www.amazon.com/gp/reader/0080431542
[34] MacMillan, J.; Beale, M. H.: Diterpene Biosynthesis. In ComprehensiveNatural Products Chemistry, D. E. Cane; D. H. R. Barton; K. Nakanishi;O. Meth-Cohn, eds., Elsevier, New York, vol. 2, pp. 221–243 1999.URL http://www.amazon.com/gp/reader/0080431542
[35] Nabeta, K.; Ishikawa, T.; Kawae, T.; Okuyama, H.: Biosynthesis of het-eroscyphic acid A in cell cultures of Heteroscyphus planus: nonequivalentlabelling of C-5 units in diterpene biosynthesis. J. Chem. Soc., Chem.Commun. 1995, pp. 681–682.URL http://dx.doi.org/10.1039/C39950000681
[36] Akhila, A.; Rani, K.; Thakur, R. S.: Biosynthesis of the clerodanefurano-diterpene lactone skeleton in Tinospora cordifolia. Phytochem-istry 1991, 30, 2573–2576.URL http://dx.doi.org/10.1016/0031-9422(91)85103-7
[37] Rohmer, M.: A Mevalonate-independent Route to Isopentenyl Diphos-phate. In Comprehensive Natural Products Chemistry, D. E. Cane;D. H. R. Barton; K. Nakanishi; O. Meth-Cohn, eds., Elsevier, New York,vol. 2, pp. 45–68 1999.URL http://www.amazon.com/gp/reader/0080431542
[38] Eisenreich, W.; Bacher, A.; Arigoni, D.; Rohdich, F.: Biosynthesis ofisoprenoids via the non-mevalonate pathway. Cell. Mol. Life Sci. 2004,61, 1401–1426.URL http://dx.doi.org/10.1007/s00018-004-3381-z
[39] SEM image created in collaboration with Dr. Simon Crawford, Universityof Melbourne.
[40] Siebert, D. J.: Localization of Salvinorin A and Related Compounds inGlandular Trichomes of the Psychoactive Sage, Salvia divinorum. Ann.Bot. 2004, 93, 763–771.URL http://dx.doi.org/10.1093/aob/mch089
BIBLIOGRAPHY 243
[41] Díaz, J. L.: Etnofarmacología de algunos psicotrópicos vegetales deMéxico. In Etnofarmacología de Plantas Alucinógenas Latinoamericanas,J. L. Díaz, ed., Centro Mexicano de Estudios en Farmacodependencia,Mexico, vol. 4 of Cuadernos Científicos CEMEF, pp. 135–201 1975. Seepp. 149 - 153 (isolation and testing), pp. 153 - 154 (Leonurus sibiricusas Cannabis substitute), pp. 196 - 197 (English abstract).
[42] Ref 15, pp. 20 - 26. Note: “HAc” is intended to refer to AcOH.
[43] Valdés, III, L. J. J. Email, 18th July 2003.
[44] Habib, A. A.: False-positive alkaloid reactions. J. Pharm. Sci. 1980, 69,37–43. And references therein.
[45] Lüdy-Tenger, F.: Versuche zur mikrochemischen Identifizierung einigerPyridinderivate. Pharm. Acta Helv. 1953, 28, 22–26.
[46] McConnell Davis, T. W.; Farmilo, C. G.; Genest, K.: Analysis of animpure heroin seizure. Bull. Narc. 1962, 14, 47–57.URL http://www.unodc.org/unodc/bulletin/bulletin_1962-01-01_3_page006.html
[47] Díaz, J. L.: Ethnopharmacology and taxonomy of Mexican psy-chodysleptic plants. J. Psychedelic Drugs 1979, 11, 71–101.
[48] Scifinder Scholar (Chemical Abstracts Service). Searched under the exactphrases “salvia” along with “alkaloid” or “nitrogenous” or “nitrogen-containing”. Accessed March 2006.URL http://www.cas.org/SCIFINDER/scicover2.html
[49] Web of Science (Thomson Scientific). Searched under “salvia” along with“alkaloid” or “nitrogenous” or “nitrogen-containing”. Accessed March2006.URL http://scientific.thomson.com/products/wos/
[50] Dictionary of Natural Products. Chapman and Hall/CRC Press, webversion 2004(2) ed. 2004. Searched for “N” in molecular formula field,and “salvia” in “biological source” field. Accessed March 2006.URL http://www.chemnetbase.com/scripts/dnpweb.exe
[51] Raffauf, R. F.: Plant Alkaloids: A Guide to Their Discovery and Distri-bution. Food Products Press, Binghamton, NY 1996. See pp. 111 - 113(S. apiana, azurea, ballotaeflora and leucophylla).
[52] Williams, C. H.; Hines, H. J. G.: The toxic properties of Salvia reflexa.Aust. Vet. J. 1940, 16, 14–20.
[53] Raffauf, R. F.: Handbook of Alkaloids and Alkaloid-Containing Plants.Wiley Interscience, New York 1970.
244 BIBLIOGRAPHY
[54] Werle, E.; Raub, A.: Über Vorkommen, Bildung Und Abbau BiogenerAmine Bei Pflanzen Unter Besonderer Berücksichtigung Des Histamins.Biochem. Z. 1948, 318, 538–553. English abstract available via Scifinder(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, 1066–1070.URL http://dx.doi.org/10.1021/np0500934
[56] Ref 15, p. 131.
[57] Valdés, III, L. J. J.: Comments on the Report of Alkaloids in Salviadivinorum by J. L. Díaz 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 ofsalvinorins and divinatorins in Salvia divinorum leaves by liquid chro-matography/multistage mass spectrometry. Rapid Commun. Mass Spec-trom. 2006, 20, 131–136.URL http://dx.doi.org/10.1002/rcm.2288
[60] Silverstein, R. M.; Webster, F. X.; Kiemle, D.: Spectrometric Identifica-tion 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 his-tory. Natural Products Co., Kennewick, WA 1993. See esp. p. 153 (fail-ure of animal testing).
[63] Ref 15, pp. 103 - 105.
[64] Coughenour, L. L.; McLean, J. R.; Parker, R. B.: A new device forthe rapid measurement of impaired motor function in mice,. Pharmacol.Biochem. Behav. 1977, 6, 351–353.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 plottedare divinorin A, TLC 2 (impure fraction), TLC 3 (much lower in 1a),“CHCl3/MeOH” and 90% MeOH. The 1a contents of the mixed frac-tions were estimated as follows. 1) “CHCl3/MeOH” fraction (actuallyCH2Cl2/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 levelwas “much higher” in TLC 2 than TLC 3 (p. 159). The CH2Cl2/MeOH
BIBLIOGRAPHY 245
fraction (“oily solids”) therefore contained between 60 and 126 mg of1a (6–12%), probably closer to 6%. 2) 90% MeOH Fraction: as notedon p. 165, see the partitioning on p. 137. Chromatography on 9.36 g ofthis fraction gave 1.51 g of yellowish crystals, which yielded 892 mg of1a after recrystallization. The 90% MeOH fraction therefore contained9.5–16% of 1a.
[66] Brimblecombe, R. W.; Green, A. L.: Effect of Monoamine Oxidase In-hibitors on the Behaviour of Rats in Hall’s 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 potentnaturally occurring nonnitrogenous κ-opioid selective agonist. Proc. Natl.Acad. Sci. U. S. A. 2002, 99, 11934–11939.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 deliv-ery systems. Eur. J. Pharm. Sci. 2000, 11, S93–98.URL http://dx.doi.org/10.1016/S0928-0987(00)00167-6
[72] ACD/Labs v8.14 for Solaris. See Scifinder (Ref. 48).URL http://www.acdlabs.com
[73] Ref 15. See the CH2Cl2/MeOH (p. 122) and 90% MeOH fractions (p.135).
[74] Hancock, B. C.; Parks, M.: What is the True Solubility Advantage forAmorphous Pharmaceuticals? Pharm. Res. 2000, 17, 397–404. See alsoreferences therein.URL http://dx.doi.org/10.1023/A:1007516718048
[75] Griesser, U.; Stowell, J. G.: Solid-state analysis and polymorphism. InPharmaceutical Analysis, D. C. Lee; M. Webb, eds., Blackwell, Oxford,pp. 240–294 2003.URL http://www.amazon.com/gp/reader/0849328144
[76] Eder, M.; Mehnert, W.: Bedeutung Pflanzlicher Begleitstoffe in Extrak-ten. Pharmazie 1998, 53, 285–293.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 ef-fect of penetration enhancers on drug delivery through skin: a QSARstudy. J. Controlled Release 2004, 99, 113–125. See also referencestherein.URL http://dx.doi.org/10.1016/j.jconrel.2004.06.010
[78] Zhang, Y.; Butelman, E.; Schlussman, S.; Ho, A.; Kreek, M.: Effects ofthe plant-derived hallucinogen salvinorin A on basal dopamine levels inthe caudate putamen and in a conditioned place aversion assay in mice:agonist actions at kappa opioid receptors. Psychopharmacology 2005,179, 551–558.URL http://dx.doi.org/10.1007/s00213-004-2087-0
[79] Fantegrossi, W. E.; Kugle, K. M.; Valdés, III, L. J. J.; Koreeda, M.;Woods, J. H.: Kappa-opioid receptor-mediated effects of the plant-derived hallucinogen, salvinorin A, on inverted screen performance inthe mouse. Behav. Pharmacol. 2005, 16, 627–633.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 profile of salvinorin A, a structurally unique kappa opi-oid receptor agonist. Pharmacol. Biochem. Behav. 2006, 83, 109–113.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 Phar-macological Activities of Three Distinct κ-ligands (Salvinorin A, TRK-820 and 3FLB) on κ Opioid Receptors in vitro and their Antipruriticand Antinociceptive Activities in vivo. J. Pharmacol. Exp. Ther. 2005,312, 220–230.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.: NeoclerodaneDiterpenes as a Novel Scaffold for µ Opioid Receptor Ligands. J. Med.Chem. 2005, 48, 4765–4771. Mesylate = compound 15.URL http://dx.doi.org/10.1021/jm048963m
[83] Butelman, E. R.; Harris, T. J.; Kreek, M. J.: The plant-derived hallu-cinogen, salvinorin A, produces κ-opioid agonist-like discriminative ef-fects in rhesus monkeys. Psychopharmacology 2004, 172, 220–224.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.: Phar-macokinetics of the plant-derived κ-opioid hallucinogen salvinorin A innonhuman primates. Synapse 2005, 58, 208–210.URL http://dx.doi.org/10.1002/syn.20191
BIBLIOGRAPHY 247
[85] Turner, D. M.: Salvinorin - The Psychedelic Essence of Salvia Divino-rum. 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.; Butel-man, E. R.; Kreek, M. J.; Murry, D. J.: Determination of Salvinorin Ain body fluids by high performance liquid chromatography-atmosphericpressure chemical ionization. J. Chromatogr. B. 2005, 818, 221–225.URL http://dx.doi.org/10.1016/j.jchromb.2004.12.041
[88] Pichini, S.; Abanades, S.; Farré, M.; Pellegrini, M.; Marchei, E.; Pacifici,R.; de la Torre, R.; Zuccaro, P.: Quantification of the plant-derived hal-lucinogen Salvinorin A in conventional and non-conventional biologicalfluids by gas chromatography/mass spectrometry after Salvia divinorumsmoking. Rapid Commun. Mass Spectrom. 2005, 19, 1649–1656.URL http://dx.doi.org/10.1002/rcm.1970
[89] Aardvark, D., ed.: Salvia Divinorum and Salvinorin A: the Best of theEntheogen 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 Wirkstoffe. PlantaMed. 1964, 12, 341–352.URL http://www.erowid.org/references/refs_view.php?A=ShowDoc1&ID=2715
[93] Ref 15, p. 19.
[94] Valdés, III, L. J. J. Email, 19th Aug. 2002.
[95] Béguin, 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 invitro pharmacological evaluation of salvinorin A analogues modified atC(2). Bioorg. Med. Chem. Lett. 2005, 15, 2761–2765.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, 5129–5132.URL http://dx.doi.org/10.1074/jbc.R400030200
[97] Sheffler, D. J.; Roth, B. L.: Salvinorin A: the “magic mint” hallucinogenfinds a molecular target in the kappa opioid receptor. Trends Pharmacol.
248 BIBLIOGRAPHY
Sci. 2003, 24, 107–109.URL http://dx.doi.org/10.1016/S0165-6147(03)00027-0
[98] Sneader, W.: Drug Discovery: A History. Wiley, New York 2005. Seeesp. 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 organicmolecules. Chem. Commun. 2002, pp. 1159–1168.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 medicinalchemistry : the rational design, mechanistic study and therapeutic appli-cation of chemical compounds, J. C. Emmett; C. Hansch; P. G. Sammes;J. B. Taylor, eds., Pergamon Press, New York, vol. 3, pp. 805–846 1990.
[102] Dhawan, B.; Cesselin, F.; Raghubir, R.; Reisine, T.; Bradley, P.; Por-toghese, P.; Hamon, M.: International Union of Pharmacology. XII. Clas-sification of opioid receptors. Pharmacol. Rev. 1996, 48, 567–592.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, 14091–14096.URL http://dx.doi.org/10.1073/pnas.0405430101
[104] Szmuszkovicz, J.: U-50,488 and the κ receptor. A personalized accountcovering the period 1973 to 1990. Prog. Drug Res. 1999, 52, 167–195.URL http://books.google.com/books?ie=UTF-8&vid=ISBN376435979X
[105] Szmuszkovicz, J.: U-50,488 and the κ receptor. Part 2. 1991-1998. Prog.Drug Res. 1999, 53, 1–51.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, 341–396.URL http://pharmrev.aspetjournals.org/cgi/content/abstract/51/2/341
[107] Carlezon Jr., W. A.; Béguin, 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 Effects of the κ-Opioid Recep-tor Agonist Salvinorin A on Behavior and Neurochemistry in Rats. J.
BIBLIOGRAPHY 249
Pharmacol. Exp. Ther. 2005, 316, 440–447. And references therein.URL http://dx.doi.org/10.1124/jpet.105.092304
[108] Pande, A. C.; Pyke, R. E.; Greiner, M.; Wideman, G. L.; Benjamin,R.; Pierce, M. W.: Analgesic efficacy of enadoline versus placeboor morphine in postsurgical pain. Clin. Neuropharmacol. 1996, 19,451–456.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 hy-dromorphone in humans. Psychopharmacology 2001, 157, 151–162.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 effects of the selectivekappa agonist spiradoline in Tourette’s syndrome: a pilot study. Psychi-atry Res. 1993, 47, 267–280.URL http://dx.doi.org/10.1016/0165-1781(93)90084-T
[111] Gadano, A.; Moreau, R.; Pessione, F.; Trombino, C.; Giuily, N.; Sinnas-samy, P.; Valla, D.; Lebrec, D.: Aquaretic effects of niravoline, a κ-opioidagonist, in patients with cirrhosis. J. Hepatol. 2000, 32, 38–42.URL http://dx.doi.org/10.1016/S0168-8278(00)80187-7
[112] Pfeiffer, A.; Brantl, V.; Herz, A.; Emrich, H. M.: Psychotomimesis me-diated by κ opiate receptors. Science 1986, 233, 774–776.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 withcyclazocine, morphine and placebo. J. Pharmacol. Exp. Ther. 1986, 238,960–968.URL http://jpet.aspetjournals.org/cgi/content/abstract/238/3/960
[114] Greenwald, M. K.; Stitzer, M. L.: Butorphanol agonist effects and acutephysical dependence in opioid abusers: comparison with morphine. DrugAlcohol Depend. 1998, 53, 17–30.URL http://dx.doi.org/10.1016/S0376-8716(98)00104-5
[115] Barber, A.; Gottschlich, R.: Novel developments with selective, non-peptidic kappa-opioid receptor agonists. Expert Opin. Invest. Drugs1997, 6, 1351–1368.URL http://www.ingentaconnect.com/content/apl/eid/1997/00000006/00000010/art00004
[116] Bush, K. A.; Kirkham, B. W.; Walker, J. S.: The κ-opioid agonist, asi-madoline, alters cytokine gene expression in adjuvant arthritis. Rheuma-tology (Oxford). 2001, 40, 1013–1021.
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, 1331–1334.URL http://dx.doi.org/10.1038/sj.bjp.0705763
[118] Dortch-Carnes, J.; Potter, D. E.: Bremazocine: a κ-opioid agonist withpotent analgesic and other pharmacologic properties. CNS Drug Rev.2005, 11, 195–212.URL http://www.nevapress.com/cnsdr/contents/1102.html
[119] Sorbera, L. A.; Castaner, J.; Leeson, P. A.: Nalfurafine hydrochloride.Drugs Fut. 2003, 28, 237–242.URL http://dx.doi.org/10.1358/dof.2003.028.03.723584
[120] Wadenberg, M. L.: A review of the properties of spiradoline: a potentand selective κ-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.: Effects of U-50,488H, a κ-opioidreceptor agonist, on the learned helplessness model of depression in mice.J. Neural Transm. 2002, 109, 1221–1225.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 depressive-like states in mice. Brain Res. Protoc. 2005, 16, 70–78.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 δ-opioid receptor agonists reduce immobilityin the forced swim assay in rats. Neuropsychopharmacology 2002, 26,744–755.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 Effects of κ-Opioid Receptor Antagonistsin the Forced Swim Test in Rats. J. Pharmacol. Exp. Ther. 2003, 305,323–330.URL http://dx.doi.org/10.1124/jpet.102.046433
[125] Béguin, C.; Carlezon, W.; Cohen, B. M.; He, M.; Lee, D. Y.-W.;Richards, M. R.: Salvinorin derivatives and uses thereof 2005. Inter-national 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 effects of the herb Salvia divinorum: acase report. J. Clin. Psychopharmacol. 2001, 21, 634–635.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. MAPSBulletin 2003, 13, 18–20.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 se-lective κ-opioid agonist, produces potent antinociception in cynomolgusmonkeys. Jpn. J. Pharmacol. 2001, 85, 282–290.URL http://dx.doi.org/10.1254/jjp.85.282
[129] Mori, T.; Nomura, M.; Nagase, H.; Narita, M.; Suzuki, T.: Effects of anewly synthesized κ-opioid receptor agonist, TRK-820, on the discrimi-native stimulus and rewarding effects of cocaine in rats. Psychopharma-cology 2002, 161, 17–22.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 OpioidsEnadoline 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 κ-opioid agonist with subtype selec-tivity 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 κ3-opioid receptors? Br. J.Pharmacol. 2006, 147, 349–350.URL http://dx.doi.org/10.1038/sj.bjp.0706603
[133] Schultes, R.; Hofmann, A.: The Botany and Chemistry of Hallucinogens.Charles C. Thomas, Springfield, IL, 2nd ed. 1980. See esp. p. 26 restructural types. Note that the phenylalkylamine skeletal types containfar 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 centralnervous system drugs and their receptors. J. Med. Chem. 1986, 29, 453–462. Figures of 82% and 58% taken from the appendix. Other figurescalculated 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 Ac-tive Component of the Hallucinogenic Sage Salvia divinorum Is a HighlyEfficacious κ-Opioid Receptor Agonist: Structural and Functional Con-siderations. J. Pharmacol. Exp. Ther. 2004, 308, 1197–1203.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.: SyntheticStudies of Neoclerodane Diterpenes from Salvia divinorum: Semisyn-thesis of Salvinicins A and B and Other Chemical Transformations ofSalvinorin A. J. Nat. Prod. 2006, 69, 107–112.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 theleaves of Salvia divinorum and their binding affinities for human κ-opioidreceptors. Bioorg. Med. Chem. 2005, 13, 5635–5639.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 novelopioid κ-agonist derived from 4,5-epoxymorphinan. Chem. Pharm. Bull.1998, 46, 366–369.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.; Na-gase, H.; Satoh, M.: Pharmacological properties of TRK-820 on clonedµ-, δ- and κ-opioid receptors and nociceptin receptor. Eur. J. Pharma-col. 1999, 376, 159–167.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.: Valida-tion of World Health Organization Guidelines for cancer pain relief: a10-year prospective study. Pain 1995, 63, 65–76.URL http://dx.doi.org/10.1016/0304-3959(95)00017-M
[141] Australian and New Zealand College of Anaesthetists: Acute Pain Man-agement: Scientific Evidence. National Health and Medical ResearchCouncil, Canberra, 2nd ed. 2005.URL http://www.nhmrc.gov.au/publications/_files/cp104.pdf
[142] Ref 15, pp. 103, 112.
[143] Ref 15, p. 165.
[144] Lewis, Sr, R. J.: Niacin (NCQ900): record number 16623. In Sax’s Dan-gerous Properties of Industrial Materials, Wiley, New York, 10th ed.
BIBLIOGRAPHY 253
2000.URL http://www.knovel.com/knovel2/Toc.jsp?BookID=707
[145] ten Tije, A. J.; Verweij, J.; Loos, W. J.; Sparreboom, A.: Phar-macological effects of formulation vehicles : implications for cancerchemotherapy. Clin. Pharmacokinet. 2003, 42, 665–685.URL http://www.ingentaconnect.com/content/adis/cpk/2003/00000042/00000007/art00005
[146] Mowry, M.; Mosher, M.; Briner, W.: Acute physiologic and chronichistologic changes in rats and mice exposed to the unique hallucinogensalvinorin A. J. Psychoactive Drugs 2003, 35, 379 – 382. Manuscriptavailable online. Accessed March 2006.URL http://www.sagewisdom.org/mowryetal.pdf
[147] Boire, R. G.; Russo, E.; Fish, A. R.; Bowman, J.: Salvia divinorum: in-formation concerning the plant and its active principle. Center for Cog-nitive 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, SanFrancisco 1973.
[150] Record of the Reasons, 33rd Meeting, 20-22 November 2001. NationalDrugs and Poisons Schedule Committee, Therapeutic Goods Adminis-tration, Commonwealth Department of Health and Ageing, Canberra2001. 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 toalt.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. PureAppl. Chem. 1998, 70, 143–216. Accessed March 2006.URL http://www.chem.qmul.ac.uk/iupac/fusedring/
[153] Munro, T.: The Prohibition of Diviner’s Sage 2002. Accessed March2006.URL http://www.thomasmunro.com/salvia.htm
[154] Leask, E.: Should Diviner’s Sage be banned? Australian Vital 2002,1(3), 149–151.
[155] Record of the Reasons, 36th Meeting, 15-17 October 2002. NationalDrugs and Poisons Schedule Committee, Therapeutic Goods Adminis-tration, 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 Classification of Publications. Office of Film and Lit-erature Classification, Canberra 2005. See p. 16. Accessed March 2006.URL http://www.oflc.gov.au/resource.html?resource=63&filename=63.pdf
[158] Andrews, K.: Salvinorin A (Question No. 1963), 11th August 2003.In Parliamentary Debates, House of Representatives, Commonwealth ofAustralia, Canberra, pp. 18142–18143. 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 KarlHanes, 7th January 2005.
[161] BEK nr 714 af 14/08/2003 (Gældende). Retsinformation, Legal Infor-mation Division, Ministry of Justice, Denmark.URL http://www.retsinfo.dk/delfin/html/b2003/0071405.htm
[162] Siebert, D. J.: The legal status of Salvia divinorum 2006. AccessedMarch 2006.URL http://www.sagewisdom.org/legalstatus.html
[163] United States. Hallucinogen Control Act of 2002, HR 5607, 107thCongress, 2nd session, 10th October 2002.URL http://thomas.loc.gov/cgi-bin/query/z?c107:H.R.5607:
[164] Gruber, J. W.: Quantification of Salvinorin A from Tissues of Salviadivinorum (Epling & Játiva-M.). Master’s thesis, Philadelphia College ofPharmacy and Science, Philadelphia, PA 1997. ProQuest document ID:740502451. Publication Number: AAT 1385726. See pp. 46 - 50 (effectof temperature on recovery).URL http://wwwlib.umi.com/dissertations/fullcit/1385726
[165] Valdés, III, L. J. J. Email, 22nd July 2005.
[166] ‘Sphere’: Salvia Divinorum Extractions using Chilled Acetone. AccessedMarch 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 deuteriumlabeled salvinorin A: synthesis of [2,2,2-2H3]-salvinorin A. Bioorg. Med.Chem. Lett. 2004, 14, 5099–5102.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. 2p.19 (partitioning chlorophylls).
[172] Vogel, A. I.; Furniss, B. S.; Hannaford, A. J.; Rogers, V.; Smith, P.W. G.; Tatchell, A. R.: Vogel’s 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.: Purification with Activated Carbon: industrial, commer-cial, environmental. Chemical Pub. Co., New York, 3rd ed. 1974. Seeesp. pp. 15 ,16 (general factors governing adsorption).
[174] Sanghi, R.; Bhattacharya, B.: Review on decolorisation of aqueous dyesolutions by low cost adsorbents. Color. Technol. 2002, 118, 256–269.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.; Woodfield,K. L.: Selective adsorption of organic homologues onto activated carbonfrom dilute aqueous solutions: Solvophobic interaction approach: PartIV. Effect of simple structural modifications with aliphatics. AIChE J.1984, 30, 197–207.URL http://dx.doi.org/10.1002/aic.690300205
[176] Vailaya, A.; Horváth, C.: Retention in reversed-phase chromatography:partition or adsorption? J. Chromatogr., A 1998, 829, 1–27. 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 Coeffi-cient of Organic Solutes between Water and Activated Carbon. Environ.Sci. Technol. 1996, 30, 143–152. See esp. Table 1, entries 129, 160, 165and 174; 277 and 289; 348, 358, 361 and 366 (adsorption constants ofbenzene, toluene, p-xylene and o-dichlorobenzene from 3 studies - theirrefs. 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 AdsorptionPotential Theory to Adsorption from Solution on Activated Carbon. IV.Steric Factors, as Illustrated by the Adsorption of Planar and OctahedralMetal Acetylacetonates. J. Phys. Chem. 1973, 77, 809–813.URL http://dx.doi.org/10.1021/j100625a015
[179] Dąbrowski, A.; Podkościelny, P.; Hubicki, Z.; Barczak, M.: Adsorption ofphenolic compounds by activated carbon - a critical review. Chemosphere2005, 58, 1049–1070.URL http://dx.doi.org/10.1016/j.chemosphere.2004.09.067
[180] Gordon, A. J.; Ford, R. A.: The chemist’s companion: a handbook ofpractical data, techniques, and references. Wiley, New York 1972. Seepp. 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 puffer fish Fugu rubripes. Toxicon 2005, 45,851–859.URL http://dx.doi.org/10.1016/j.toxicon.2005.02.005
[182] Naundorf, A.; Ajisaka, K.: Purification of α-N -acetyl-galactosaminidasefrom Aspergillus niger and its use in the synthesis of GalNAc-α-(1→O)-serine. Enzyme Microb. Technol. 1999, 25, 483–488.URL http://dx.doi.org/10.1016/S0141-0229(99)00073-3
[183] Aksu, Z.; Tunç, O.: Application of biosorption for penicillin G removal:comparison with activated carbon. Process Biochem. 2005, 40, 831–847.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 ultrafiltrationto improve the extraction of antibiotics. Sep. Purif. Technol. 2004, 34,115–123.URL http://dx.doi.org/10.1016/S1383-5866(03)00185-0
[185] Scrivens, W. A.; Bedworth, P. V.; Tour, J. M.: Purification of GramQuantities 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 ColumnPurification of Gram Quantities of C70. J. Am. Chem. Soc. 1994, 116,6939–6940.URL http://dx.doi.org/10.1021/ja00094a060
[187] Athanasiadou, M.; Jensen, S.; Klasson Wehler, E.: Preparative Fraction-ation of a Commercial PCB Product. Chemosphere 1991, 23, 957–970.URL http://dx.doi.org/10.1016/0045-6535(91)90123-U
BIBLIOGRAPHY 257
[188] Kočan, A.; Petrík, J.; Chovancová, J.; Drobná, B.: Method forthe group separation of non-ortho-, mono-ortho- and multi-ortho-substituted polychlorinated biphenyls and polychlorinated dibenzo-p-dioxins/polychlorinated dibenzofurans using activated carbon chro-matography. J. Chromatogr., A 1994, 665, 139–153.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 poly-chlorinated biphenyls by carbon column chromatography. Chemosphere1991, 23, 915–924.URL http://dx.doi.org/10.1016/0045-6535(91)90096-V
[190] Zupančič-Kralj, L.; Jan, J.; Marsel, J.: Fractionation of chloroorganiccompounds on a carbon cartridge. Chemosphere 1991, 23, 841–843.URL http://dx.doi.org/10.1016/0045-6535(91)90089-V
[191] Lombardo, M. E.; Viscelli, T. A.; Mittelman, A.; Hudson, P. B.: TheRemoval of Non-steroidal Pigments of Urinary Extracts by Adsorptionon Charcoal. J. Biol. Chem. 1955, 212, 353–360.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 Identification ofSome Organophosphate Insecticides in the Presence of Plant Extracts.J. Agric. Food Chem. 1963, 11, 174–178.URL http://dx.doi.org/10.1021/jf60126a023
[194] Leonard, J.; Lygo, B.; Procter, G.: Advanced Practical Organic Chem-istry. 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 vegetableoil in hexane to activated earths and carbons. Eur. Food Res. Technol.2001, 213, 132–138.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 ControlledSage Salvia divinorum. J. Nat. Prod. 2003, 66, 1242–1244.URL http://dx.doi.org/10.1021/np030313i
[197] Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.: Experi-mental and computational approaches to estimate solubility and perme-ability in drug discovery and development settings. Adv. Drug DeliveryRev. 1997, 23, 3–25.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.: Con-stituents of sage (Salvia officinalis) with in vitro affinity to human brainbenzodiazepine receptor. Planta Med. 2003, 69, 113–117.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 flavonoidsas high-affinity benzodiazepine site ligands. Biochem. Pharmacol. 2003,66, 2397–2407.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 flavonoid ligands ofthe central benzodiazepine receptor have no effect on memory tasks inrats. Pharmacol. Biochem. Behav. 1997, 58, 887–891.URL http://dx.doi.org/10.1016/S0091-3057(97)00054-3
[201] Zanoli, P.; Avallone, R.; Baraldi, M.: Behavioral characterization of theflavonoids apigenin and chrysin. Fitoterapia 2000, 71, S117–S123.URL http://dx.doi.org/10.1016/S0367-326X(00)00186-6
[202] González, A. G.; Aguiar, Z. E.; Luis, J. G.; Ravelo, A. G.; Vázquez,J. T.; Domínguez, X. A.: Flavonoids from Salvia texana. Phytochemistry1989, 28, 2871–2872.URL http://dx.doi.org/10.1016/S0031-9422(00)98114-7
[203] Valdés, 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 byGABAA and GABAC receptors. Eur. J. Pharmacol. 2003, 461, 79–87.URL http://dx.doi.org/10.1016/S0014-2999(03)01309-8
[206] Lutterodt, G. D.: Inhibition of gastrointestinal release of acetylcholineby quercetin as a possible mode of action of Psidium guajava leaf extractsin the treatment of acute diarrhoeal disease. J. Ethnopharmacol. 1989,25, 235–247.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 divinorumand its active ingredient salvinorin A inhibit enteric cholinergic transmis-sion in the guinea-pig ileum. Neurogastroenterol. Motil. 2006, 18, 69–75.URL http://dx.doi.org/10.1111/j.1365-2982.2005.00725.x
[208] Singh, A.; Naidu, P. S.; Kulkarni, S. K.: Quercetin, a bioflavonoid,reverses development of tolerance and dependence to morphine. Drug
BIBLIOGRAPHY 259
Dev. Res. 2002, 57, 167–172.URL http://dx.doi.org/10.1002/ddr.10119
[209] Ma, M.-H.; Lin, C.-I.: Adsorption kinetics of β-carotene from soy oilusing regenerated clay. Sep. Purif. Technol. 2004, 39, 201–209.URL http://dx.doi.org/10.1016/j.seppur.2003.12.007
[210] Tanin, S.; Gurgey, I.: Bleaching of cottonseed and sunflower oils by activecarbons obtained from rice hulls. Chim. Acta Turc. 1988, 16, 209–219.
[211] Ortega, A. Email, 15th Oct. 2005.
[212] Delgado, G.; Romo de Vivar, A.; Ortega, A.; Cárdenas, J.; Schlemper,E. O.: Diterpenoids from Viguiera insignis. Phytochemistry 1983, 22,1227–1230.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 withBentonite. J. Nat. Prod. 1995, 58, 424–427.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 heterocyclicchemistry using tonsil, a bentonitic clay: 1981 to 2003. Trends Hetero-cycl. Chem. 2003, 9, 195–235.
[215] Munro, T. A.; Rizzacasa, M. A.: Salvinorins D-F, New NeoclerodaneDiterpenoids from Salvia divinorum, and an Improved Method for theIsolation of Salvinorin A. J. Nat. Prod. 2003, 66, 703–705.URL http://dx.doi.org/10.1021/np0205699
[216] Ohsaki, A.; Yan, L. T.; Ito, S.; Edatsugi, H.; Iwata, D.; Komoda, Y.: Theisolation and in vivo potent antitumor activity of clerodane diterpenoidfrom the oleoresin of the Brazilian medicinal plant, Copaifera langsdorfiiDesfon. Bioorg. Med. Chem. Lett. 1994, 4, 2889–2892.URL http://dx.doi.org/10.1016/S0960-894X(01)80834-9
[217] McChesney, J. D.; Clark, A. M.; Silveira, E. R.: Antimicrobial diterpenesof Croton sonderianus, 1. Hardwickic and 3,4-secotrachylobanoic acids.J. Nat. Prod. 1991, 54, 1625–1633.URL http://dx.doi.org/10.1021/np50078a021
[218] Costa, M.; Tanaka, C. M. A.; Imamura, P. M.; Marsaioli, A. J.: Iso-lation and synthesis of a new clerodane from Echinodorus grandiflorus.Phytochemistry 1999, 50, 117–122.URL http://dx.doi.org/10.1016/S0031-9422(98)00464-6
[219] Spectral Database for Organic Compounds (SDBS), National Institute ofAdvanced 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.: Effect of magnesium coordinationon the 13C and 15N magnetic resonance spectra of chlorophyll a. TheRelative energies of nitrogen nπ* states as deduced from a completeassignment of chemical shifts. J. Am. Chem. Soc. 1974, 96, 7058–7066.URL http://dx.doi.org/10.1021/ja00829a038
[221] Huang, K.-F.; Hsu, C.-J.: Constituents of stem bark of Erythrina ar-borescens. 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 terpenebiosynthesis. Synthesis and resolution of presqualene and prephytoenealcohols. J. Am. Chem. Soc. 1978, 100, 6174–6182.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, 941–945.URL http://dx.doi.org/10.1021/jo00109a026
[224] Giner, J.-L.; Berkowitz, J. D.; Andersson, T.: Nonpolar Components ofthe Latex of Euphorbia peplus. J. Nat. Prod. 2000, 63, 267–269.URL http://dx.doi.org/10.1021/np990081g
[225] Misra, R.; Pandey, R. C.; Dev, S.: The chemistry of the oleo resin fromhardwickia pinnata: a series of new diterpenoids. Tetrahedron Lett. 1964,5, 3751–3759.URL http://dx.doi.org/10.1016/S0040-4039(01)89373-4
[226] Misra, R.; Pandey, R. C.; Dev, S.: Absolute stereochemistry of hard-wickiic acid and its congeners. Tetrahedron Lett. 1968, 9, 2681–2684.URL http://dx.doi.org/10.1016/S0040-4039(00)89672-0
[227] Misra, R.; Pandey, R. C.; Dev, S.: Higher isoprenoids–VIII: Diterpenoidsfrom the oleoresin of hardwickia pinnata part 1: hardwickiic acid. Tetra-hedron 1979, 35, 2301–2310.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, C19H27OCOOH. Z. Kristallogr. - New Cryst. Struct.2004, 219, 111–113. (accessed Oct 2005).URL http://www.oldenbourg.de/verlag/zkristallogr/mn-ncsc0402.htm#p111
BIBLIOGRAPHY 261
[229] Ortega, A.; Cárdenas, J.; Gage, D. A.; Maldonado, E.: Abietane andclerodane 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.: Isolationand insecticidal activity of (-)-hardwickiic acid from Croton aromaticus.Planta Med. 1987, 53, 575.
[231] Cocker, W.; Moore, A. L.; Pratt, A. C.: Dextrorotatory hardwickiicacid, an extractive of Copaifera officinalis. Tetrahedron Lett. 1965, 6,1983–1985.URL http://dx.doi.org/10.1016/S0040-4039(01)83897-1
[232] Cascon, V.; Gilbert, B.: Characterization of the chemical compositionof oleoresins of Copaifera guianensis Desf., Copaifera duckei Dwyer andCopaifera multijuga Hayne. Phytochemistry 2000, 55, 773–778.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 inducesprogrammed cell death in human lymphoid leukemia Molt 4B cells. Int.J. Mol. Med. 1999, 4, 377–380.
[234] Tomita, Y.: Immunological role of vitamin A and its related substancesin prevention of cancer. Nutr. Cancer 1983, 5, 187–194.
[235] Köhler, I.; Jenett-Siems, K.; Kraft, C.; Siems, K.; Abbiw, D.; Bien-zle, 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, 1022–1027.URL http://www.znaturforsch.com/sc/57c/s57c1022.pdf
[236] Rajab, M. S.; Cantrell, C. L.; Franzblau, S. G.; Fisher, N. H.: An-timycobacterial activity of (E)-phytol and derivatives. A preliminarystructure-activity study. Planta Med. 1998, 64, 2–4.
[237] Saludes, J. P.; Garson, M. J.; Franzblau, S. G.; Aguinaldo, A. M.: An-titubercular constituents from the hexane fraction of Morinda citrifoliaLinn. (Rubiaceae). Phytother. Res. 2002, 16, 683–685.URL http://dx.doi.org/10.1002/ptr.1003
[238] Arnhold, T.; Elmazar, M. M. A.; Nau, H.: Prevention of Vitamin A Ter-atogenesis by Phytol or Phytanic Acid Results from Reduced Metabolismof Retinol to the Teratogenic Metabolite, All-trans-retinoic Acid. Toxi-col. Sci. 2002, 66, 274–282.URL http://dx.doi.org/10.1093/toxsci/66.2.274
262 BIBLIOGRAPHY
[239] Topçu, G.: Bioactive Triterpenoids from Salvia Species. J. Nat. Prod.2006, 69, 482–487.URL http://dx.doi.org/10.1021/np0600402
[240] Ovesná, Z.; Vachálková, A.; Horváthová, K.; Táthová, D.: Pentacyclictriterpenoic acids: new chemoprotective compounds. Neoplasma 2004,51, 327–333.URL http://www.neoplasma.sk/abstrakt.asp?idn=645
[241] Liu, J.: Pharmacology of oleanolic acid and ursolic acid. J. Ethnophar-macol. 1995, 49, 57–68.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 effects on HIV-1 protease of constituents from the wood ofXanthoceras sorbifolia. J. Nat. Prod. 2000, 63, 238–242.URL http://dx.doi.org/10.1021/np9902441
[243] Watanabe, M.; Kobayashi, Y.; Ogihara, J.; Kato, J.; Oishi, K.: HIV-1reverse transcriptase-inhibitory compound in Salvia officinalis. Food Sci.Technol. Res. 2000, 6, 216–220.
[244] Ikeda, T.; Yokomizo, K.; Okawa, M.; Tsuchihashi, R.; Kinjo, J.; Nohara,T.; Uyeda, M.: Anti-herpes virus type 1 activity of oleanane-type triter-penoids. Biol. Pharm. Bull. 2005, 28, 1779–1781.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, 70–75.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 effects of Baccharisconferta and some of its constituents. J. Pharm. Pharmacol. 2002, 54,99–104.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 penta-cyclic triterpenoids from plants of Argentina and Chile. Phytomedicine1999, 6, 341–345.
[248] Tan, N.; Kaloga, M.; Radtke, O. A.; Kiderlen, A. F.; Öksüz, S.; Ulubelen,A.; Kolodziej, H.: Abietane diterpenoids and triterpenoic acids fromSalvia cilicica and their antileishmanial activities. Phytochemistry 2002,61, 881–884.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.: An-tifeedant activity of some pentacyclic triterpene acids and their fattyacid ester analogs. J. Agric. Food Chem. 2003, 51, 1952–1955.URL http://dx.doi.org/10.1021/jf020691d
[250] Blagg, B. S. J.; Jarstfer, M. B.; Rogers, D. H.; Poulter, C. D.: Re-combinant Squalene Synthase. A Mechanism for the Rearrangement ofPresqualene Diphosphate to Squalene. J. Am. Chem. Soc. 2002, 124,8846–8853.URL http://dx.doi.org/10.1021/ja020411a
[251] Faure, S.; Connolly, J. D.; Fakunle, C. O.; Piva, O.: Structure and Syn-thesis of Anhydrobisfarnesol from Euphorbia lateriflora and AsymmetricSynthesis of (R)-Sesquilavandulol. Tetrahedron 2000, 56, 9647–9653.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, 567–571.URL http://www.iupac.org/publications/pac/2001/pdf/7303x0567.pdf
[253] Connolly, J. D.: Natural products from around the world. Rev. Lati-noamer. Quím. 1997, 25, 77–85.
[254] Van Tamelen, E. E.; Leopold, E. J.: Mechanism of presqualenepyrophosphate-squalene biosynthesis. II. Synthesis of bifarnesol. Tetra-hedron Lett. 1985, 26, 3303–3306.URL http://dx.doi.org/10.1016/S0040-4039(00)98283-2
[255] Valdés, III, L. J. J.: Divinorin C, a new neoclerodane diterpene froma 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 Bulletin2001, 11, 22–35. 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 ESide Chains. II. Benzoyl Triflate and its Application in the Determina-tion of the Absolute Configuration of Divinorin A and B, and Terrecyclicacid. Ph.D. thesis, University of Michigan, Ann Arbor, MI 1984. Pro-Quest Publication Number: AAT 8422201 (Document ID: 748992941).See pp. 78 – 80, 196 – 200 (hydride reductions).URL http://wwwlib.umi.com/dissertations/fullcit/8422201
264 BIBLIOGRAPHY
[261] Ref 260, pp. 84, 85 (compound 123).
[262] Ref 260, p. 76 (direct diacetylation).
[263] Ref 260, pp. 72 – 75 (BH−4 reductions, epimerizations and proposed mech-
anism).
[264] Goddard, R.; Akhtar, F.: The structure of jateorin. Acta Crystallogr.,Sect. C: Cryst. Struct. Commun. 1986, C42, 1217–1220.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 Crys-tallogr., Sect. C: Cryst. Struct. Commun. 1989, C45, 300–303.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 ofpalmarin. Liebigs Ann. Chem. 1989, pp. 485–487.
[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-p-iodophenyl-3-phenylpyrazoline adduct of isocolumbin. J. Chem. Soc., B1966, pp. 853–861.URL http://dx.doi.org/10.1039/J29660000853
[268] Overton, K. H.; Weir, N. G.; Wylie, A.: The stereochemistry of thecolombo root bitter principles. J. Chem. Soc., C 1966, pp. 1482–1490.See isocolumbin, dihydrocolumbin and dihydroisocolumbin (hydrogena-tions); table 2 (furanolactone infrared absorptions).URL http://dx.doi.org/10.1039/J39660001482
[269] Yonemitsu, M.; Fukuda, N.; Kimura, T.; Komori, T.: Studies on theconstituents of Jateorhiza palmata Miers (colombo root). I. Separationand structure of a new furanoid diterpene glucoside (palmatoside A).Liebigs Ann. Chem. 1986, pp. 1327–1333. See compounds 3, 4 (nujolmulls) 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 κ-OpioidReceptor Agonist. J. Med. Chem. 2005, 48, 345–348.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.; Béguin, C.; Carlezon, W. A.; Cohen,B.: Synthesis and in vitro pharmacological studies of new C(2) modifiedsalvinorin A analogues. Bioorg. Med. Chem. Lett. 2005, 15, 3744–3747.Epimerization is discussed on p. 3746 c.1.URL http://dx.doi.org/10.1016/j.bmcl.2005.05.048
BIBLIOGRAPHY 265
[272] Ref 260, p. 83 (deacetylations with KCN); cf. p. 202 (yield).
[273] Valdés, III, L. J. J. Email, 28th Apr. 2003.
[274] Khong, P. W.; Lewis, K. G.: Acetyl migration in urs-12-ene-3α,24-diol.Aust. J. Chem. 1975, 28, 201–206. See also references 1 - 6 therein.URL http://dx.doi.org/10.1071/CH9750201
[275] Smith, M. B.; March, J.: March’s advanced organic chemistry: reactions,mechanisms, and structure. Wiley, New York, 5th ed. 2001. See Table8.1, 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 princi-ples. IV. Investigations on the constitution of palmarin. J. Chem. Soc.1962, pp. 4809–4815. 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.: Diterpenoidbitter principles. V. Constitution of palmarin and its congeners. J. Chem.Soc. 1962, pp. 4816–4820.URL http://dx.doi.org/10.1039/JR9620004816
[278] Ref 260, pp. 90 – 92 (deuteration). The NMR data given are consistentwith this structure. The 13C peak at δ 171.6 is incorrectly assigned to thelactone, however, on the basis of an apparent reduction in intensity rel-ative to 1a. The apparent reduction was probably due to an inadequatesignal-to-noise ratio; five peaks were not observed. The claimed yield of89% is implausible given the epimerization we and Valdés observed withKCN.
[279] Strazzolini, P.; Giumanini, A. G.; Cauci, S.: Acetic formic anhydride: areview. Tetrahedron 1990, 46, 1081–1118.URL http://dx.doi.org/10.1016/S0040-4020(01)86676-X
[280] Reber, F.; Lardon, A.; Reichstein, T.: Teilsynthese von 11-Epi-corticosteron. Bestandteile der Nebennierenrinde und verwandte Stoffe,86. Mitteilung. Helv. Chim. Acta 1954, 37, 45–58. See compounds XIIand XVIII.URL http://dx.doi.org/10.1002/hlca.19540370107
[281] Oliveto, E. P.; Gerold, C.; Rausser, R.; Hershberg, E. B.: 11-OxygenatedSteroids. XII. The Preparation of 17α-Hydroxycorticosterone 21-Acetate(Kendall’s Compound F Acetate) via 11β-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 anHydroxyalkyl-3-morpholinonen. Liebigs Ann. Chem. 1989, pp. 665–669.
[284] Eckenberg, P.; Groth, U.; Huhn, T.; Richter, N.; Schmeck, C.: A usefulapplication of benzyl trichloroacetimidate for the benzylation of alcohols.Tetrahedron 1993, 49, 1619–1624.URL http://dx.doi.org/10.1016/S0040-4020(01)80349-5
[285] Tanabe, M.; Peters, R. H.: (R,S)-Mevalonolactone-2-13C. 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 mono-protection of symmetrical diols. Tetrahedron Lett. 1997, 38, 5945–5948.URL http://dx.doi.org/10.1016/S0040-4039(97)01328-2
[287] Charette, A. B.: Tetra-n-butylammonium iodide. In Encyclopedia ofreagents for organic synthesis, L. A. Paquette, ed., Wiley, New York1995.URL http://dx.doi.org/10.1002/047084289X.rt018
[288] Kraus, G. A.; Frazier, K. A.; Roth, B. D.; Taschner, M. J.; Neuen-schwander, K.: Conversion of lactones into ethers. J. Org. Chem. 1981,46, 2417–2419.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. 501–531 2001.
[290] Hansen, M. C.; Verdaguer, X.; Buchwald, S. L.: Convenient Two-StepConversion of Lactones into Cyclic Ethers. J. Org. Chem. 1998, 63,2360–2361.URL http://dx.doi.org/10.1021/jo9716082
[291] Kawamura, K.; Hinou, H.; Matsuo, G.; Nakata, T.: Efficient strategyfor convergent synthesis of trans-fused polycyclic ethers based on an in-tramolecular SmI2-promoted cyclization of iodo ester. Tetrahedron Lett.2003, 44, 5259–5261. 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 ofnaphtho[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 Syn-thesis of 4-Deazaformycin A. J. Org. Chem. 2003, 68, 6466–6469. Seecompound 7.URL http://dx.doi.org/10.1021/jo026715x
BIBLIOGRAPHY 267
[294] Barbier, M.; Devys, M.; Parisot, D.: Synthesis of anhydronectri-achrysone, an extensively conjugated γ-pyrone. Synth. Commun. 1993,23, 1481–1488.
[295] Ikawa, T.; Sajiki, H.; Hirota, K.: Unexpected deprotection of silyl andTHP ethers induced by serious disparity in the quality of Pd/C catalystsand elucidation of the mechanism. Tetrahedron 2004, 60, 6189–6195.URL http://dx.doi.org/10.1016/j.tet.2004.05.040
[296] Sajiki, H.; Hattori, K.; Hirota, K.: Highly chemoselective hydrogenationwith retention of the epoxide function using a heterogeneous Pd/C-ethylenediamine catalyst and THF. Chem. Eur. J. 2000, 6, 2200–2204.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 effects).
[298] Bruno, M.; Rosselli, S.; Pibiri, I.; Piozzi, F.; Simmonds, M.: Hydro-genation derivatives of neo-clerodanes. Heterocycles 2000, 55, 599–612.See hydrogenations of montanin C (4) and teucrin A (6). Site accessedFebruary 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 ofhardwood extractives. XXXIV. Constituents of Guarea species. J. Chem.Soc. 1962, pp. 5095–5104. 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, 789–791.
[301] Brieskorn, C. H.; Pfeuffer, T.: Labiatenbitterstoffe: Pikropolin und ähn-liche 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 penduliflaworosin, a new furanoidditerpene from Croton penduliflorus. J. Chem. Soc., Perkin Trans. 11981, pp. 1151–1153.URL http://dx.doi.org/10.1039/P19810001151
[303] Bui, A. M.; Parello, J.; Potier, P.; Janot, M. M.: Sur la structuredu collybolide, nouvelle substance sesquiterpénique extraite du Colly-bia maculata Alb. et Sch. ex Fries (Basidiomycètes). C. R. Acad. Sci.,C. 1970, 270, 1022–1026. See 2 and 2’.
268 BIBLIOGRAPHY
[304] Chan, W. R.; Taylor, D. R.; Willis, C. R.: Terpenoids from the Eu-phorbiaceae. I. The structure of crotonin, a norditerpene from Crotonlucidus. J. Chem. Soc., C 1968, pp. 2781–2785.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 andstereochemistry of aphanamixin and aphanamixinin. Tetrahedron 1970,26, 1859–1867. 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 oflimonin. J. Am. Chem. Soc. 1952, 74, 688–693.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 antifeedantagents against Spodoptera frugiperda larvae. A structure-activity rela-tionship study. J. Agric. Food Chem. 2002, 50, 6766–6774.URL http://dx.doi.org/10.1021/jf020607u
[308] De Alvarenga, M. A.; Gottlieb, H. E.; Gottlieb, O. R.; Magalhães, M. T.;Da Silva, V. O.: Diasin, a diterpene from Croton diasii. Phytochemistry1978, 17, 1773–1776.URL http://dx.doi.org/10.1016/S0031-9422(00)88692-6
[309] Rodriguez-Hahn, L.; O’Reilly, R.; Esquivel, B.; Maldonado, E.; Or-tega, A.; Cardenas, J.; Toscano, R. A.; Chan, T.-M.: Tilifodiolide,tetraline-type diterpenoid of clerodanic origin from Salvia tiliaefolia. J.Org. Chem. 1990, 55, 3522–3525.URL http://dx.doi.org/10.1021/jo00298a026
[310] Hori, T.; Kiang, A. K.; Nakanishi, K.; Sasaki, S.; Woods, M. C.: Struc-tures of fibraurin and a minor product from Fibraurea chloroleuca. Tetra-hedron 1967, 23, 2649–2656.URL http://dx.doi.org/10.1016/0040-4020(67)85129-9
[311] Melera, A.; Schaffner, K.; Arigoni, D.; Jeger, O.: Zur Konstitution desLimonins I. Über den Verlauf der alkalischen Hydrolyse von Limonin undLimonol. Helv. Chim. Acta 1957, 40, 1420–1437. See hydrogenations ofI 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, 2491–2493.URL http://dx.doi.org/10.1021/ja01150a024
[313] Ida, Y.; Kubo, S.; Fujita, M.; Komori, T.; Kawasaki, T.: Struktur derDiosbulbine-D, -E, -F, -G und H. Justus Liebigs Ann. Chem. 1978, pp.818–833. See hydrogenation of compounds D and F.
BIBLIOGRAPHY 269
[314] Popa, D. P.; Reinbol’d, A. M.: Structure of teucrin A. Chem. Nat. Comp.1975, 9, 27–30. 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 desFraxinellons 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 witha norbornane part skeleton. Tetrahedron 1972, 28, 2333–2340.URL http://dx.doi.org/10.1016/S0040-4020(01)93576-8
[317] Kouzi, S. A.; McMurtry, R. J.; Nelson, S. D.: Hepatotoxicity of German-der (Teucrium chamaedrys L.) and One of Its Constituent NeoclerodaneDiterpenes 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.: AntimalarialActivity of Tropical Meliaceae Extracts and Gedunin Derivatives. J. Nat.Prod. 1997, 60, 336–341. See compound 10.URL http://dx.doi.org/10.1021/np9605394
[319] Siegel, S.: Rhodium on Alumina. In Encyclopedia of reagents for organicsynthesis, 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 fluoride dopedalumina: an effective reagent for ester hydrolysis under solvent free con-ditions. Green Chem. 2001, 3, 261–262.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 typographicalerror, being smaller than several daughter ions. The correct value is 365.
[322] Presser, A.; Hüfner, A.: Trimethylsilyldiazomethane - A Mild and Effi-cient Reagent for the Methylation of Carboxylic Acids and Alcohols inNatural Products. Monatsh. Chem. 2004, 135, 1015–1022.URL http://dx.doi.org/10.1007/s00706-004-0188-4
[323] Ref 60, p. 153 (cyclic α-diones).
[324] Raymond, S.: The Structure of Methyl Acetylacrylate. J. Am. Chem.Soc. 1950, 72, 4304–4306.URL http://dx.doi.org/10.1021/ja01165a537
270 BIBLIOGRAPHY
[325] Caton, M. P. L.; Darnbrough, G.; Parker, T.: Prostaglandins VI: Base-Catalyzed Autoxidation of α-Hydroxycyclopentanones and the Synthesisof 9,10-Diketoprostanoic Acids. Tetrahedron Lett. 1980, 21, 1685–1686.URL http://dx.doi.org/10.1016/S0040-4039(00)77786-0
[326] Sundt, E.; Jeger, O.; Prelog, V.: Conversion of Cevagenin into a Diosphe-nol. Chem. Ind. 1953, pp. 1365–1366.
[327] Lavie, D.; Shlomo, S.: The Constituents of Ecballium elaterium L. II.α-Elaterin. J. Am. Chem. Soc. 1958, 80, 707–710.URL http://dx.doi.org/10.1021/ja01536a046
[328] Lack, R. E.; Ridley, A. B.: Autoxidation of 2α-Hydroxy-5α-Cholestan-3-One in Methanolic Potassium Hydroxide. J. Chem. Soc., C 1968, pp.3017–3020.URL http://dx.doi.org/10.1039/J39680003017
[329] Weissberger, A.; Schwarze, W.: Über die Autoxydation des 1-Oxy-α-tetralons. Justus Liebigs Ann. Chem. 1931, 487, 53–61.
[330] Slavikova, B.; Kasal, A.; Budesinsky, M.: Autoxidation vs hydrolysis in16α-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 autox-idation of cyclic 1,4-diketones to enediones. J. Org. Chem. 1960, 25,1853–1855.URL http://dx.doi.org/10.1021/jo01081a006
[332] Ref 260, p. 82 (PCC oxidation).
[333] González, A. G.; Aguiar, Z. E.; Luis, J. G.; Ravelo, A. G.: New diter-penes from Salvia texana. Chemical and biogenetic aspects. Tetrahedron1989, 45, 5203–5214. 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 ofring A in lupeol. Tetrahedron 1966, 22, 3597–3599.URL http://dx.doi.org/10.1016/S0040-4020(01)92548-7
[335] Marchand, A. P.; Reddy, G. M.: Ultrasound-promoted sodium boro-hydride 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-PCUD-8,11-dione. Org. Prep. Proced. Int. 1990, 22, 528–531. 59 was sonicatedat 35 kHz, 320 W output.
[336] Gemal, A. L.; Luche, J. L.: Lanthanoids in organic synthesis. 6. Reduc-tion of α-enones by sodium borohydride in the presence of lanthanoidchlorides: synthetic and mechanistic aspects. J. Am. Chem. Soc. 1981,
BIBLIOGRAPHY 271
103, 5454–5459. On decomposition, see p. 5456 c. 2, esp. fig. 3.URL http://dx.doi.org/10.1021/ja00408a029
[337] Marchand, A. P.; LaRoe, W. D.; Sharma, G. V. M.; Suri, S. C.; Reddy,D. S.: Facile stereoselective reductions of enediones and cage diketonesusing NaBH4-CeCl3. J. Org. Chem. 1986, 51, 1622–1625. See attemptedreduction of 12 and 16.URL http://dx.doi.org/10.1021/jo00359a054
[338] Dimitrov, V.; Kostova, K.; Genov, M.: Anhydrous cerium(III) chloride -effect of the drying process on activity and efficiency. Tetrahedron Lett.1996, 37, 6787–6790.URL http://dx.doi.org/10.1016/S0040-4039(96)01479-7
[339] Zsakó, J.; László, T.: Complex compounds of some diosphenolic typecarotenoids. Astacene complexes. Spectrophotometric study of the sys-tem iron(III)-astacene. Rev. Roum. Chim. 1981, 26, 237–243.
[340] Rubin, Y.; Knobler, C. B.; Diederich, F.: Precursors to the Cy-clo[n]carbons: From 3,4-Dialkynyl-3-cyclobutene-1,2-diones and 3,4-Dialkynyl-3-cyclobutene-1,2-diols to Cyclobutenodehydroannulenes andHigher Oxides of Carbon. J. Am. Chem. Soc. 1990, 112, 1607–1617. SeeScheme V.URL http://dx.doi.org/10.1021/ja00160a047
[341] McMurry, J. E.: Ester cleavages via SN2-type dealkylation. Org. React.1976, 24, 187–224.URL http://dx.doi.org/10.1002/0471264180.or024.02
[342] Salomon, C. J.; Mata, E. G.; Mascaretti, O. A.: Recent developments inchemical deprotection of ester functional groups. Tetrahedron 1993, 49,3691–3734.URL http://dx.doi.org/10.1016/S0040-4020(01)90225-X
[343] Müller, P.; Siegfried, B.: SN2 reactions with carboxylic esters. Selectivecleavage of methyl esters. Helv. Chim. Acta 1974, 57, 987–994.URL http://dx.doi.org/10.1002/hlca.19740570404
[344] Elsinger, F.; Schreiber, J.; Eschenmoser, A.: Notiz über die Selektiv-ität der Spaltung von Carbonsäuremethylestern mit Lithiumjodid. Helv.Chim. Acta 1960, 43, 113–118.URL http://dx.doi.org/10.1002/hlca.19600430116
[345] Dean, P. D. G.: Halogenolysis of methyl glycyrrhetate with lithiumiodide-dimethylformamide. J. Chem. Soc. 1965, p. 6655.URL http://dx.doi.org/10.1039/JR9650006655
[346] Müller, P.; Siegfried, B.: Decarboxylation of β-keto esters in hexam-ethylphosphoric triamide. Tetrahedron Lett. 1973, pp. 3565–3568.URL http://dx.doi.org/10.1016/S0040-4039(01)86971-9
272 BIBLIOGRAPHY
[347] Dykstra, R. R.: Hexamethylphosphoric Triamide. In Encyclopedia ofreagents for organic synthesis, L. A. Paquette, ed., Wiley, New York1995.URL http://dx.doi.org/10.1002/047084289X.rh020
[348] Elsinger, F.: 2-Benzylcyclopentanone. Org. Synth. 1965, 45, 7. AlsoColl. 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 ofmethyl esters. Synth. Commun. 1972, 2, 389–394.
[350] Beck, A. K.; Seebach, D.: N,N -Dimethylpropyleneurea. In Encyclopediaof reagents for organic synthesis, L. A. Paquette, ed., Wiley, New York1995.URL http://dx.doi.org/10.1002/047084289X.rd366
[351] Sharma, L.; Nayak, M. K.; Chakraborti, A. K.: A mild and chemoselec-tive method for ester O-alkyl cleavage using in situ generated potassiumthiophenoxide from catalytic quantities of base. Tetrahedron 1999, 55,9595–9600.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 highlyselective demethylation of a quaternized ergoline into a one-pot synthesisof pergolide. Org. Process Res. Dev. 1997, 1, 77–80.URL http://dx.doi.org/10.1021/op9600015
[353] Bartlett, P. A.; Johnson, W. S.: Improved reagent for the O-alkyl cleav-age of methyl esters by nucleophilic displacement. Tetrahedron Lett.1970, 11, 4459–4462.URL http://dx.doi.org/10.1016/S0040-4039(01)83950-2
[354] Kelly, T. R.; Dali, H. M.; Tsang, W. G.: Lithium thiomethoxide: aconvenient mercaptide reagent. Tetrahedron Lett. 1977, 18, 3859–3860.URL http://dx.doi.org/10.1016/S0040-4039(01)83373-6
[355] Feutrill, G. I.; Mirrington, R. N.: Reactions with thioethoxide ion indimethylformamide. II. O-Alkyl cleavage of methylene ethers, methylesters, and aryloxyacetates. Aust. J. Chem. 1972, 25, 1731–1735.URL http://dx.doi.org/10.1071/CH9721731
[356] Vaughan, W. R.; Baumann, J. B.: Reactions of Alkyl Carboxylic Esterswith Mercaptides. J. Org. Chem. 1962, 27, 739–744.URL http://dx.doi.org/10.1021/jo01050a010
[357] Ireland, R. E.; Thompson, W. J.: An approach to the total synthesis ofchlorothricolide: the synthesis of the top half. J. Org. Chem. 1979, 44,3041–3052. 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.; Safina, B. S.: AMild and Selective Method for the Hydrolysis of Esters with TrimethyltinHydroxide. Angew. Chem. Int. Ed. 2005, 44, 1378–1382.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.: Trimethyltinsyndrome as a hippocampal degeneration model: temporal changes andneurochemical features of seizure susceptibility and learning impairment.Neuroscience 1997, 81, 1183–1191.URL http://dx.doi.org/10.1016/S0306-4522(97)00220-0
[360] Mascaretti, O. A.; Furlan, R. L. E.: Esterifications, transesterifica-tions, and deesterifications mediated by organotin oxides, hydroxides,and alkoxides. Aldrichimica Acta 1997, 30, 55–68.
[361] Lee, D. Y.; He, M.; Kondaveti, L.; Liu-Chen, L. Y.; Ma, Z.; Wang, Y.;Chen, Y.; Li, J. G.; Béguin, C.; Carlezon, W. A.; Cohen, B.: Synthesisand in vitro pharmacological studies of C(4) modified salvinorin A ana-logues. Bioorg. Med. Chem. Lett. 2005, 15, 4169–4173.URL http://dx.doi.org/10.1016/j.bmcl.2005.06.092
[362] Oikawa, M.; Wada, A.; Okazaki, F.; Kusumoto, S.: Acidic, SelectiveMonoacylation of vic-Diols. J. Org. Chem. 1996, 61, 4469–4471. Seecompound 38.URL http://dx.doi.org/10.1021/jo960081a
[363] Burchat, A. F.; Chong, J. M.; Nielsen, N.: Titration of alkyllithiumswith a simple reagent to a blue endpoint. J. Organomet. Chem. 1997,542, 281–283.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 withborane-tetrahydrofuran. Remarkably convenient procedure for the selec-tive conversion of carboxylic acids to the corresponding alcohols in thepresence of other functional groups. J. Org. Chem. 1973, 38, 2786–2792.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, auseful reagent for the Eschenmoser α,β cleavage of α,β-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 syn-thesis of 3-chlorogentisyl alcohol–a metabolite of Penicillium canadense.Tetrahedron 1972, 28, 1107–1111. 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 9,10-dihydro-9,10-ethanoanthracenes. Tetrahedron 1972, 28, 1435–1467. Seecompound 5. BH3 was generated in situ as per reference 5 therein. Stoi-chiometry 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. Re-action of diborane in tetrahydrofuran with selected organic compoundscontaining 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 en-hances 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. Hemispherandswith four self-organizing units. J. Am. Chem. Soc. 1984, pp. 2160–2171.See compounds 32 and 37.URL http://dx.doi.org/10.1021/ja00319a042
[372] Dias, J. R.; Pettit, G. R.: Reduction of δ-lactones and hindered esterswith diborane. J. Org. Chem. 1971, 36, 3485–3489. See compounds 7and 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 inthe controlled reaction of carboxylic acids with borane-tetrahydrofuran.Acyloxyboranes as intermediates in the fast reduction of carboxylic acidsby 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 ofreagents for organic synthesis, L. A. Paquette, ed., Wiley, New York1995.URL http://dx.doi.org/10.1002/047084289X.rm273
BIBLIOGRAPHY 275
[376] Pine, S. H.; Shen, G. S.; Hoang, H.: Ketone methylenation using theTebbe 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 titanium-based reagents. Org. React. 1993, 43, 1–91.URL http://dx.doi.org/10.1002/0471264180.or043.01
[378] Lombardo, L.: Methylenation of carbonyl compounds with zinc-dibromomethane-titanium tetrachloride. Applications to gibberellins.Tetrahedron Lett. 1982, 23, 4293–4296. Note: due to the thick con-sistency of this reagent, transfer was very difficult using 0.5 mm I.D. (21gauge) 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-methylene-cis-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 sol-vents. Org. React. 1975, 22, 401–422. See table p. 418 (eliminations).URL http://dx.doi.org/10.1002/0471264180.or022.03
[381] Yamamura, S.; Toda, M.; Hirata, Y.: Modified Clemmensen Reduc-tion: Cholestane. Org. Synth. 1973, 53, 289. See discussion re α-acetoxycholestanone.URL http://www.orgsyn.org/orgsyn/pdfs/CV6P0289.pdf
[382] Chamberlin, A. R.; Sheppeck, II, J. E.: p-Toluenesulfonylhydrazide. InEncyclopedia of reagents for organic synthesis, L. A. Paquette, ed., Wi-ley, New York 1995.URL http://dx.doi.org/10.1002/047084289X.rt137
[383] Caglioti, L.: Reduction of ketones by use of the tosylhydrazone deriva-tives: androstan-17β-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 recrys-tallisation.URL http://www.orgsyn.org/orgsyn/pdfs/CV5P1055.pdf
[385] Bandgar, B. P.; Sadavarte, V. S.; Uppalla, L. S.; Govande, R.: Chemose-lective preparation of oximes, semicarbazones, and tosylhydrazones with-out catalyst and solvent. Monatsh. Chem. 2001, 132, 403–406.URL http://dx.doi.org/10.1007/s007060170126
276 BIBLIOGRAPHY
[386] Ješelnik, M.; Varma, R. S.; Polanc, S.; Kočevar, M.: Solid-state syn-thesis of heterocyclic hydrazones using microwaves under catalyst-freeconditions. Green Chem. 2002, 4, 35–38.URL http://dx.doi.org/10.1039/b108029f
[387] Hajipour, A. R.; Mallakpour, S. E.; Imanzadeh, G.: A Rapid and Con-venient Synthesis of Oximes in Dry Media under Microwave Irradiation.J. Chem. Res. 1999, pp. 228–229.URL http://dx.doi.org/10.1039/a806359a
[388] Bertz, S. H.; Dabbag, G.: Improved preparations of some arenesulfonyl-hydrazones. J. Org. Chem. 1983, 48, 116–119.URL http://dx.doi.org/10.1021/jo00149a022
[389] Iida, T.; Tamura, T.; Matsumoto, T.; Chang, F.: Improved Conditionsfor Preparation and Reductive Cleavage of Steroidal Ketone Tosylhydra-zones. Synthesis 1984, pp. 957–959.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 arylhy-drazones at ambient conditions. Ultrason. Sonochem. 2003, 10, 45–48.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 oftosylhydrazone derivatives from keto esters. Synth. Commun. 1984, 14,281–288.
[392] Hertz, H. S.; Coxon, B.; Siedle, A. R.: Disproportionation and pyrolysisof p-toluenesulfonylhydrazine. J. Org. Chem. 1977, 42, 2508–2509.URL http://dx.doi.org/10.1021/jo00434a036
[393] Pouchert, C. J.; Behnke, J.: The Aldrich library of 13C and 1H FT-NMRspectra, vol. 2. Aldrich Chemical Co., Milwaukee 1993. See p. 1634A.
[394] El-Sayed, R. A.: Review on the chemistry of sulfonohydrazides and sul-fonoazides. Phosphorus, Sulfur Silicon Relat. Elem. 2004, 179, 237–266.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 Or-ganic Functional Group Transformations, A. R. Katritzky; O. Meth-Cohn; C. W. Rees, eds., Pergamon, Oxford, vol. 1, pp. 1–26, 1st ed.1995.
[396] Barton, D. H. R.; Ferreira, J. A.; Jaszberenyi, J. C.: Free radical de-oxygenation of thiocarbonyl derivatives of alcohols. In Preparative Car-bohydrate Chemistry, S. Hanessian, ed., Marcel Dekker, New York, pp.151–172 1997.
BIBLIOGRAPHY 277
[397] Salomon, C. J.; Danelon, G. O.; Mascaretti, O. A.: A Practical Methodfor the Disposal of Organotin Residues from Reaction Mixtures. J. Org.Chem. 2000, 65, 9220–9222. 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 rad-ical reactions. 32. Radical deoxygenations, dehalogenations, and deami-nations with dialkyl phosphites and hypophosphorous acid as hydrogensources. J. Org. Chem. 1993, 58, 6838–6842.URL http://dx.doi.org/10.1021/jo00076a054
[399] Takamatsu, S.; Katayama, S.; Hirose, N.; Naito, M.; Izawa, K.: Radi-cal deoxygenation and dehalogenation of nucleoside derivatives with hy-pophosphorous acid and dialkyl phosphites. Tetrahedron Lett. 2001, 42,7605–7608.URL http://dx.doi.org/10.1016/S0040-4039(01)01617-3
[400] Rho, H.-S.; Ko, B.-S.: Regioselective deoxygenation of the cyclic thiono-carbonates of 2,3-dihydroxy esters with magnesium in methanol. Synth.Commun. 1999, 29, 2875–2880.
[401] Brown, A. C.; Carpino, L. A.: Magnesium in methanol: substitute forsodium amalgam in desulfonylation reactions. J. Org. Chem. 1985, 50,1749–1750.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 trib-utyltin hydride. Chem. Pharm. Bull. 1983, 31, 1842–1847.
[403] Patroni, J. J.; Stick, R. V.; Engelhardt, L. M.; White, A. H.: The deoxy-genation of some carbohydrate diols via the derived cyclic thiocarbonate.Aust. J. Chem. 1986, 39, 699–711.URL http://dx.doi.org/10.1071/CH9860699
[404] Rho, H.-S.: Tributyltin hydride-induced free radical deoxygenation ofthe cyclic thionocarbonates of threo-2,3-dihydroxy esters and ketones.Synth. Commun. 1997, 27, 3887–3893.
[405] Kanemitsu, K.; Tsuda, Y.; Haque, M. E.; Tsubono, K.; Kikuchi, T.: Re-action of cyclic thioxocarbonates with tributyltin hydride. Chem. Pharm.Bull. 1987, 35, 3874–3879.
[406] Tsuda, Y.; Kanemitsu, K.; Kakimoto, K.; Kikuchi, T.: Utilization ofsugars in organic synthesis. XIX. Trialkyltin radical used to catalyzethe O,S -rearrangement of cyclic thionocarbonates. A new entry to thiosugars. Chem. Pharm. Bull. 1987, 35, 2148–2150.
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 de-rived from some carbohydrate 1,2-diols. Aust. J. Chem. 1989, 42, 2127–2141.URL http://dx.doi.org/10.1071/CH9892127
[408] Szammer, J.; Ötvös, L.: A convenient preparation of tri-n-butyltin hy-dride. Chem. Ind. 1988, p. 764.
[409] McAlonan, H.; Stevenson, P. J.: Convenient One-Pot Synthesis of Hexa-n-butylditin from Bis(tri-n-butyltin) Oxide. Organometallics 1995, 14,4021–4022.URL http://dx.doi.org/10.1021/om00008a058
[410] Giese, B.; Gröninger, K. S.: 1,3,4,6-Tetra-O-acetyl-2-deoxy-α-D-glucopyranose. Org. Synth. 1990, 69, 66–71. 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 forreductions 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 purification oforganic stannanes. J. Org. Chem. 1991, 56, 4985–4987.URL http://dx.doi.org/10.1021/jo00016a037
[413] Crich, D.; Sun, S.: A Practical Method for the Removal of OrganotinResidues from Reaction Mixtures. J. Org. Chem. 1996, 61, 7200–7201.URL http://dx.doi.org/10.1021/jo960751c
[414] Rajan Babu, T. B. V.; Bulman Page, P. C.; Buckley, B. R.: Tri-n-butylstannane. 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 or-ganic 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 deoxygenationof alcohols via their S -methyl dithiocarbonate derivatives with di-n-butylphosphine oxide as hydrogen-atom donor. Synlett 1998, pp. 39–40.URL http://dx.doi.org/10.1055/s-1998-1582
[417] Jang, D. O.; Cho, D. H.: Radical deoxygenation of alcohols and vicinaldiols with N -ethylpiperidine hypophosphite in water. Tetrahedron Lett.2002, 43, 5921–5924.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, 247–249.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 un-natural (R)-malic acid derivatives with phenylsilanes. Synth. Commun.2003, 33, 515–519.URL http://dx.doi.org/10.1081/SCC-120015803
[420] Klein Gebbinck, E. A.; Jansen, B. J. M.; de Groot, A.: Insect antifeedantactivity of clerodane diterpenes and related model compounds. Phyto-chemistry 2002, 61, 737–770. 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.; Jáuregui, E. A.; Sosa,M. E.; Tonn, C. E.; Luco, J. M.; Gordaliza, M.: Structure-AntifeedantActivity Relationship of Clerodane Diterpenoids. Comparative Studywith Withanolides and Azadirachtin. J. Agric. Food Chem. 2000, 48,1384–1392.URL http://dx.doi.org/10.1021/jf990006b
[422] Simmonds, M. S. J.; Blaney, W. M.; Esquivel, B.; Rodriguez-Hahn, L.:Effect of clerodane-type diterpenoids isolated from Salvia on the feedingbehavior of Spodoptera littoralis. Pestic. Sci. 1996, 47, 17–23.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 elec-trophysiological study of antifeedant mechanisms associated with poly-hydroxy alkaloids. J. Chem. Ecol. 1990, 16, 3167–3196. See p. 3171(choice assay).URL http://dx.doi.org/10.1007/BF00979618
[424] Novac, O.; Guenier, A. S.; Pelletier, J.: Inhibitors of protein synthesisidentified by a high throughput multiplexed translation screen. NucleicAcids Res. 2004, 32, 902–915. See p. 904 under “dual-luciferase reporterassay.”.URL http://dx.doi.org/10.1093/nar/gkh235
[425] Reference Method for Broth Dilution Antifungal Susceptibility Testingof Yeasts; Approved Standard M27-A2. National Committee for ClinicalLaboratory Standards, Wayne, PA, 2nd ed. 2002.
[426] Methods for Dilution Antimicrobial Susceptibility Tests for Bacteria thatGrow Aerobically; Approved Standard M7-A6. National Committee forClinical Laboratory Standards, Wayne, PA, 6th ed. 2003.
[427] Performance Standards for Antimicrobial Disk Susceptibility Tests; Ap-proved Standard M2-A8. National Committee for Clinical LaboratoryStandards, Wayne, PA, 8th ed. 2003.
280 BIBLIOGRAPHY
[428] Photos by Andrea Bigham, University of Melbourne.
[429] Rovinsky, S. A.; Cizadlo, G. R.: Salvia divinorum Epling et Játiva-M.(Labiatae): An Ethnopharmacological Investigation. The McNair Schol-arly Review 1998, 3, 142–156. College of St Scholastica (Duluth, MN).Unofficial 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.: κ-opioid receptor agonist suppres-sion of HIV-1 expression in CD4+ lymphocytes. Biochem. Pharmacol.2001, 61, 1145–1151. 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.: κ-opioid receptor ago-nist inhibition of HIV-1 envelope glycoprotein-mediated membrane fu-sion and CXCR4 expression on CD4+ lymphocytes. Biochem. Pharma-col. 2002, 63, 1037–1041.URL http://dx.doi.org/10.1016/S0006-2952(02)00875-4
[432] Reeves, J. D.; Piefer, A. J.: Emerging drug targets for antiretroviraltherapy. Drugs 2005, 65, 1747–1766.URL http://www.ingentaconnect.com/content/adis/dgs/2005/00000065/00000013/art00002
[433] Gekker, G.; Lokensgard, J. R.; Peterson, P. K.: Naltrexone potentiatesanti-HIV-1 activity of antiretroviral drugs in CD4+ lymphocyte cultures.Drug Alcohol Depend. 2001, 64, 257–263.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.: U50,488inhibits HIV-1 Tat-induced monocyte chemoattractant protein-1 (CCL2)production by human astrocytes. Biochem. Pharmacol. 2003, 65, 9–14.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.; Pe-terson, P. K.: U50,488 protection against HIV-1-related neurotoxicity:involvement of quinolinic acid suppression. Neuropharmacology 2000,39, 150–160.URL http://dx.doi.org/10.1016/S0028-3908(99)00063-5
[436] O’Brien, M. E.; Clark, R. A.; Besch, C. L.; Myers, L.; Kissinger, P.:Patterns and correlates of discontinuation of the initial HAART regimenin an urban outpatient cohort. J. Acquir. Immune Defic. Syndr. 2003,34, 407–414.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 andantiemetic effects of opioids in the dog. Eur. J. Pharmacol. 1986, 128,143–150.URL http://dx.doi.org/10.1016/0014-2999(86)90760-0
[438] Kohl, R. L.; MacDonald, S.: New pharmacologic approaches to the pre-vention of space/motion sickness. J. Clin. Pharmacol. 1991, 31, 934–946.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 molecularbeacons to detect virus replication in human immunodeficiency virustype 1-infected individuals on prolonged effective antiretroviral therapy.J. Virol. 1999, 73, 6099–6103.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 fromindividuals infected with drug-resistant virus and declining CD4+ T-cellcounts. J. Acquir. Immune Defic. Syndr. 2005, 40, 140–148.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 NI-AID Compound (9966) Tested in PBMC Cell-Based Assays. SouthernResearch Institute, Frederick, Maryland 2004. Available from the au-thor.
[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.: κ opioid receptors in humanmicroglia downregulate human immunodeficiency virus 1 expression.Proc. Natl. Acad. Sci. U. S. A. 1996, 93, 8051–8056.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.: κ-opioidpotentiation of tumor necrosis factor-α-induced anti-HIV-1 activity inacutely infected human brain cell cultures. Biochem. Pharmacol. 1998,56, 397–404.URL http://dx.doi.org/10.1016/S0006-2952(98)00161-0
[444] Chou, J. Z.; Chait, B. T.; Wang, R.; Kreek, M. J.: Differential biotrans-formation of dynorphin A (1-17) and dynorphin A (1-13) peptides inhuman blood, ex vivo. Peptides 1996, 17, 983–990.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 κ2-selective activation of G-proteins: κ opioid receptor stimulation of [35S]GTPγS binding in guinea pig brain. Biochem. Pharmacol. 1998, 56,
282 BIBLIOGRAPHY
113–120.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, 785–793.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.; Holling-shead, M.; Sausville, E. A.: Relationships between drug activity in NCIpreclinical in vitro and in vivo models and early clinical trials. Br. J.Cancer 2001, 84, 1424–1431.URL http://dx.doi.org/10.1054/bjoc.2001.1796
[449] US National Cancer Institute, Developmental Therapeutics ProgramData 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 (P388protocol).URL http://dtp.nci.nih.gov/branches/btb/pdf/In_Vivo_Cancer_Models.pdf
[451] Schultz, R. J. Drug Synthesis and Chemistry Branch, National CancerInstitute. 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.; Sheffler, D. J.; Roth, B. L.: Identification of theMolecular Mechanisms by Which the Diterpenoid Salvinorin A Binds toκ-Opioid Receptors. Biochemistry 2005, 44, 8643–8651.URL http://dx.doi.org/10.1021/bi050490d
[454] Rovner, S. L.: Chemical ‘Naming’ Method Unveiled. Chem. Eng. News2005, 83, 39–40. (accessed Jan. 2006).URL http://pubs.acs.org/isubscribe/journals/cen/83/i34/html/8334sci1.html
[455] IUPAC International Chemical Identifier (accessed Oct. 2005).URL http://www.iupac.org/inchi
[456] Hoye, T. R.; Hanson, P. R.; Vyvyan, J. R.: A Practical Guide to First-Order Multiplet Analysis in 1H NMR Spectroscopy. J. Org. Chem. 1994,59, 4096–4103.URL http://dx.doi.org/10.1021/jo00094a018
BIBLIOGRAPHY 283
[457] Hoye, T. R.; Zhao, H.: A Method for Easily Determining Coupling Con-stant Values: An Addendum to “A Practical Guide to First-Order Multi-plet Analysis in 1H 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 ofCommon Laboratory Solvents as Trace Impurities. J. Org. Chem. 1997,62, 7512–7515.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.: 2-Substituted tryptamines: agents with selectivity for 5-HT6 serotoninreceptors. J. Med. Chem. 2000, 43, 1011–1018.URL http://dx.doi.org/10.1021/jm990550b
[460] Pauwels, P. J.; Colpaert, F. C.: Heterogeneous ligand-mediated Ca++
responses at wt and mutant α2A-adrenoceptors suggest multiple ligandactivation binding sites at the α2A-adrenoceptor. Neuropharmacology2000, 39, 2101–2111.URL http://dx.doi.org/10.1016/S0028-3908(00)00040-X