Phytochemical and Biological Investigations on Clathrotropis Glaucophylla (Fabaceae), an Ingredient of Yanomamï Curare, Emphasizing on Quinolizidine Alkaloids

Research Collection

Doctoral Thesis

Phytochemical and biological investigations on Clathrotropis glaucophylla (Fabaceae), an ingredient of Yanomamï curare, emphasizing on quinolizidine alkaloids

Author(s): Sagen, Anne-Lise

Publication Date: 2002

Permanent Link: https://doi.org/10.3929/ethz-a-004438153

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For more information please consult the Terms of use.

ETH Library Diss. ETH No. 14785

PHYTOCHEMICAL AND BIOLOGICAL INVESTIGATIONS ON CLATHROTROPIS GLAUCOPHYLLA (FABACEAE), AN INGREDIENT OF YANOMAMÏ CURARE,

EMPHASIZING ON QUINOLIZIDINE ALKALOIDS

A dissertation submitted to the

SWISS FEDERAL INSTITUTE OF TECHNOLOGY ZURICH

for the degree of Doctor of Natural Sciences

Presented by ANNE-LISE SAGEN

Pharmacist (University of Oslo, Norway) born February 7, 1973 Norway

Accepted on the recommendation of

Prof. Dr. Otto Sticher, examiner Prof. Dr. Ihsan Çalis, co-examiner Prof. Dr. Gerd Folkers, co-examiner Dr. Jörg Heilmann, co-examiner

Zurich 2002 To my parents and grandparents Clathrotropis glaucophylla ACKNOWLEDGEMENTS

ACKNOWLEDGEMENTS

The present Ph. D. thesis was carried out at the Swiss Federal Institute of Technol¬ ogy (ETH) Zurich, Institute of Pharmaceutical Sciences, Section Pharmacognosy and Phytochemistry.

I would like to thank:

Prof. Dr. Otto Sticher for giving me the opportunity to work in his research group and for providing excellent working facilities.

Dr. Jörg Heilmann, my co-examiner, for general help and discussions, and for his positive spirit and catching laughter.

Prof. Dr. Ihsan Çahs, my second co-examiner, for general help and discussions,

and for being so efficient and reliable.

Prof. Dr. Gerd Folkers for accepting to act as co-examiner.

Dr. Oliver Zerbe for helpful discussions concerning structural problems, and for performing NMR experiments.

Jiirg Gertsch for the collection of the plant material and for many interesting dis¬

cussions.

Dr. Walter Amrein, Mr. Oswald Greter and Mr. Rolf Häfliger for recording

mass spectra.

Monica Baumgartner, Daniel Eichenberger and Rita Becker (diploma and

semester students) for assisting in the isolation and structure elucidation procedures of the isolated compounds.

Mr. Michael Wasescha for performing cytotoxicity assays and for general techni- cal support.

Mr. Ivo Fähnle, Mr. Jonas Friedrich, Dr. Pierre Suter and Mr. Roberto Car- anci for ail support concerning the computers.

Mrs. Beatrice Häsler for drawing the Clathrotropis glaucophylla picture.

Pmar Akbay, Fatima Hilmi, Dr. Karin Winkelmann and Dr. Conwitha Lapke,

"meine Frauen", for all help and support, and for being such good friends.

Dr. Kristian Koch and Dr. Andreas (Andy) Gerbert, my chemical expertise, for all help and discussions, and together with Mr. Magnus Karlsson, for many inter¬ esting lunch and coffee breaks.

The "Phyto-group" and all the rest of my colleagues, staff and students at the Institute of Pharmaceutical Sciences for the pleasant working atmosphere and enjoyable time we had together.

Mr. René Bemsel and his team from the "Schalter" for their always efficient and friendly support.

Mrs. Anita Caputo for keeping my lab and office as clean as possible and for her positive spirit and smile.

My family and friends for their kind support and encouragement. TABLE OF CONTENTS

TABLE OF CONTENTS

ABBREVIATIONS 1

SUMMARY 5

ZUSAMMENFASSUNG 7

RÉSUMÉ 9

SAMMENDRAG 11

RESUMEN 13

1. INTRODUCTION 15

1.1 Plants as a resource for medicinal remedies 15

1.2 Scope of the present work 17

2. THE FABACEAE FAMILY 19

2.1 Systematic classification 19

2.2 Characteristics of the family 20

2.3 Uses 20

2.4 Papilionoideae (Fabaceae) 21

2.5 The Clathrotropis genus 22

2.5.1 Characteristics of the genus 22

2.5.2 Previous phytochemical work on Clathrotropis species 22

2.5.3 Clathrotropis glaucophylla 23

3. ETHNOBOTANY 25

3.1 The Yanomam'i Indians 25 TABLE OF CONTENTS

3.2 Uses of Clathrotropis 26

3.2.1 The seeds 26

3.2.2 The bark 26

3.3 Curare 27

4. QUINOLIZIDINE ALKALOIDS (LUPINE ALKALOIDS) 31

4.1 Introduction 31

4.2 Structural types 31

4.3 Biosynthesis 32

4.4 Spectroscopy for structure determination of lupine alkaloids 36

4.5 Biological activities 37

5. RESULTS AND DISCUSSION 41

5.1 Preliminary studies 41

5.1.1 Extraction 41

5.1.2 TLC screening 42

5.1.3 Biological assays 42

5.2 Extraction 44

5.3 Fractionation and isolation 47

5.3.1 Fractionation of the alkaloid extracts 47

5.3.2 Fractionation of the DCM extract 48

5.4 Structures of the isolated compounds 52

5.5 Structure elucidation 58

5.5.1 (-)-Anagyrine (2) and (-)-thermopsine (3) 58 TABLE OF CONTENTS

5.5.2 (-)-Baptifoline (4) and (-)-epibaptifoline (5) 70

5.5.3 (-)-Clathrotropine(l) 75

5.5.4 (-)-Rhombifoline (6) and (-)-tinctorine (7) 82

5.5.5 (-)-Cytisine (8) and (-)-A/-methylcytisine (9) 89

5.5.6 (-)-Lupanine (10) 94

5.5.7 (-)-6a-Hydroxylupanine (11 ) and (+)-5,6-dehydrolupanine (12) 102

5.5.8 Betulinic acid (13), 23-0(4'-hydroxy-3'-methoxy cinnamoyl)betulinic acid (14) and 23-0(4'-hydroxy-3',5'- dimethoxy cinnamoyl)betulinic acid (15) 108

5.5.9 2'-0-Methylevernic acid (16) and confluentic acid (17) 117

5.5.10 5(S),6(S)-6(2-Hydroxy-1 -methylpropyl)-3,5-dimethyl-5,6- dihydro-2H-a-pyrone (18) 123

5.5.11 6-Hydroxy-8-methoxy-3-n-pentylisocoumarin (19) 128

5.5.12 ß-Sitosterol (20) and stigmasterol (21 ) 133

5.5.13 7ß-Hydroxysitosterol (22) and 7ß-hydroxystigmasterol (23) 134

5.5.14 ß-Sitosterol-3-O-ß-glucoside (24) and stigmasterol-3-Oß- glucoside (25) 135

5.5.15 ß-Amyrin (26) and glutinol (27) 136

5.6 Bioactivity of the isolated compounds 137

5.6.1 Quinolizidine alkaloids 137

5.6.2 Compounds isolated from the dichloromethane extract 137

5.7 Chemotaxonomic discussion 139

6. EXPERIMENTAL PART 141

6.1 Thin layer chromatography 141

6.2 Biological assays 141 TABLE OF CONTENTS

6.2.1 Brine shrimp lethality bioassay 141

6.2.2 KB cell cytotoxicity test 142

6.2.3 Antibacterial bioautographic assay 142

6.2.4 Minimum inhibitory concentration (MIC) 143

6.2.5 Antioxidant test 143

6.3 NMR spectroscopy 144

6.4 Mass spectrometry 144

6.5 Optical rotation 144

6.6 UV spectroscopy 144

7. PAPER I 145

8. PAPER II 155

9. CONCLUSIONS 163

REFERENCES 167

LIST OF FIGURES 175

LIST OF TABLES 179

PUBLICATIONS, POSTERS AND ORAL PRESENTATIONS 181

CURRICULUM VITAE 183 ABBREVIATIONS 1

ABBREVIATIONS

ACN acetonitrile

Md specific optical rotation

B.c. Bacillus cereus

brd broad doublet

brdd broad double doublet

brs broad singulet

C carbon atom

13C carbon-13

cc column chromatography CDC13 deuterated chloroform CD30D deuterated methanol

COSY correlated spectroscopy (NMR)

8 chemical shift

d doublet

2D two-dimensional

DCM dichloromethane

dd double doublet

ddd double double doublet

DEPT distortionless enhancement by polarization transfer DMSO dimethylsulphoxide DMSO-d6 deuterated dimethylsulphoxide dq double quadruplet dt double triplet

E.c. Escherichia coli

EI-MS electron impact mass spectrometry

ESI-MS electrospray mass spectrometry

EtOAc ethyl acetate

EtOH ethanol

GC gas chromatography

H hydrogen atom

lU proton 2

HR-MALDI-MS high resolution matrix assisted laser desorption/ionization

mass spectrometry HMBC heteronuclear multiple bond correlation (NMR) HPLC high performance liquid chromatography

HSQC heteronuclear single quantum coherence (NMR) IC50 inhibition concentration (50% inhibition)

J coupling constant KB cells Hela cell line (ATCC CCL 17)

m multiplet

MeOD deuterated methanol

MeOH methanol

Hg microgram

Ml Micrococcus luteus

Ml microliter MHz megaherz

MS mass spectrometry MW molecular weight m/z mass-to-charge ratio

NMR nuclear magnetic resonance

NOESY nuclear Overhauser enhancement spectroscopy (NMR) PE petroleum ether

Ps.a. Pseudomonas aeruginosa ROESY rotating frame NOESY (NMR) RP reversed phase

s singulet S.e. Staphylococcus epidermidis Si60 silica gel 60

t triplet TLC thin layer chromatography

TOCSY total correlation spectroscopy

tq triple quadruplet

UV ultraviolet ABBREVIATIONS 3

VLC vacuum liquid chromatography QA quinolizidine alkaloid 4

jSeite Leer/ I Blank leaf SUMMARY 5

SUMMARY

In the present study the bark of the Yanomamï curare plant, Clathrotropis glaucophylla Cowan, was investigated for its contents of secondary metabolites. Clathrotropis is a small genus of the Fabaceae family, with six species endemic to the tropical South

America. They are small to medium trees with very large leaves, and the fruits are oblong and flat, with large seeds. C. glaucophylla was collected in the rainforests of the upper Orinoco in Venezuela in 1999. Ethnobotanical investigation has revealed that the species C. glaucophylla and C. macrocarpa (wapu kohi) are of great economic impor¬ tance to the Yanomamï Amerindians in Venezuela, the seeds playing a significant role in alimentation, and the bark being used as ingredient of curare arrow poison.

Fractionation of the alkaloid and dichloromethane extracts of C. glaucophylla bark, by means of various chromatographic methods (VLC, CC, HPLC), led to the isolation of 27 natural compounds. From the alkaloid extract a new quinolizidine alkaloid, (-)-

13a-hydroxy-15a-(l-hydroxyethyl)-anagyrine, ((-)-clathrotropine), was isolated together with eleven known quinolizidine alkaloids: (-)-anagyrine, (-)-thermopsine, (-)- baptifoline, (-)-epibaptifoline, (-)-rhombifoline, (-)-tinctorine, (-)-cytisine, (-)-A^-methylcytisine, (-)-lupanine, (-)-6a-hydroxylupanine and (+)-5,6-dehydrolupanine. The investigation of the dichloromethane extract led to the isolation of two new betulinic acid derivatives, 23-0-(4'-hydroxy-3'-methoxy-cinnamoyl)betulinic acid and 23-0-

(4'-hydroxy-3',5'-dimethoxy-cinnamoyl)betulinic acid, one new a-pyrone, 5(5),6(5)-6-

(2-hydroxy-l-methylpropyl)-3,5-dimethyl-5,6-dihydro-2H-a-pyrone, and one new isocoumarin, 6-hydroxy-8-methoxy-3-n-pentylisocoumarin, in addition to eleven known compounds which were identified as: betulinic acid, ß-amyrin, glutinol, ß-sitosterol, stigmasterol, 7ß-hydroxysitosterol, 7ß-hydroxystigmasterol, ß-sitosterol-3-O-ß-glucoside, stigmasterol-3-O-ß-glucoside, 2'-0-methylevernic acid, and confluentic acid.

Structure elucidation of the isolated compounds was carried out by means of spec¬ troscopic, spectrometric and physical methods: ID and 2D NMR experiments (1H, 13C, COSY, HSQC, HMBC, ROESY, HSQC-TOCSY and INADEQUATE), UV, MS, and [ab-

This is the first phytochemical study on C. glaucophylla, which showed this plant to accumulate a great variety of quinolizidine alkaloids as well as triterpenes, sterols, depsides and isocoumarins. It is the first time quinolizidine alkaloids have been isolated 6

from an arrow poison ingredient, and it is also the first report on Clathrotropis species being used for arrow poison. ZUSAMMENFASSUNG 7

ZUSAMMENFASSUNG

In der vorliegenden Arbeit wurde die Rinde der Yanomamï Curarepflanze Clathrot¬ ropis glaucophylla Cowan auf ihre Inhaltsstoffe untersucht. Clathrotropis ist eine kleine

Gattung aus der Familie der Fabaceae, mit sechs einheimischen Arten im tropischen

Südamerika. Es handelt sich um kleine bis mittelgrosse Bäume mit sehr grossen Blät¬ tern. Die Früchte sind länglich und flach mit grossen Samen. Ethnobotanische Untersu¬ chungen haben gezeigt, dass die Arten C. glaucophylla und C. macrocarpa (wapu kohi) ökonomisch sehr wichtig für die Yanomamï Indianer in Venezuela sind. Die Samen spielen eine bedeutsame Rolle in der Ernährung und die Rinde wird als Pfeilgiftbe¬ standteil (Curare) verwendet.

Die Fraktionierung der Alkaloid- und Dichlormethanextrakte von C. glaucophylla, mittels verschiendener chromatographischer Methoden (VLC, CC, HPLC) führte zur

Isolierung von 27 Naturstoffen. Aus dem Alkaloidextrakt wurde ein neues Chinolizidin- Alkaloid (-)-13a-Hyroxy-15a-(l-hydroxyethyl)-anagyrin, (-)-Clathrotropin isoliert, sowie elf schon bekannte Chinolizidin-Alkaloide: (-)-Anagyrin, (-)-Thermopsin, (-)- Baptifolin, (-)-Epibaptifolin, (-)-Rhombifolin, (-)-Tinctorin, (-)-Cytisin, (-)-JV-Methyl- cytisin, (-)-Lupanin, (-)-6a-Hydroxylupanin und (+)-5,6-Dehydrolupanin. Die Untersu¬ chung des Dichlormethanextraktes führte zur Isolierung von zwei neuen Betulinsäured- erivaten: 23-0-(4'-Hydroxy-3'-methoxy-cinnamoyl)-betulinsäure und 23-0-(4'-

Hydroxy-3',5'-dimethoxy-cinnamoyl)-betulinsäure, einem neuen a-Pyron (5(S),6(S)-6-

(2-Hydroxy-l-methylpropyl)-3,5-dimethyl-5,6-dihydro-2H-a-pyron) und einem neuen Isocoumarinderivat (6-Hydroxy-8-methoxy-3-«-pentylisocoumarin), sowie elf bekannten Substanzen, die als Betulinsäure, ß-Amyrin, Glutinol, ß-Sitosterol, Stigmas- terol, 7ß-Hydroxysitosterol, 7ß-Hydroxystigmasterol, ß-Sitosterol-3-O-ß-glucosid, Stigmasterol-3-O-ß-glucosid, 2'-0-Methylevernicsäure und Confluenticsäure identifi¬ ziert wurden.

Die Strukturaufklärung erfolgte mittels ID und 2D NMR (!H, 13C, COSY, HSQC, HMBC, ROESY, HSQC-TOCSY and INADEQUATE) und UV-Spektroskopie, Mas- senspektrometrie, sowie optische Drehung.

Es handelt sich hier um die erste phytochemische Untersuchung von C. glauco¬ phylla. Sie zeigt, dass die Pflanze eine grosse Fülle verschiedener Chinolizidin-Alka¬ loide, sowie Triterpene, Sterole, Depside und Isocoumarine akkumuliert. Dies ist das 8

erste Mal, dass Chinolizidin-Alkaloide aus einer Pfeilgiftpflanze isoliert wurden.

Ferner ist es der erste Bericht, dass Clathrotropis-Arten zur Herstellung von Pfeilgiften verwendet werden. RESUME 9

RÉSUMÉ

Dans le travail qui suit, l'écorce de Clathrotropis glaucophylla Cowan est examinée pour son contenu en composés secondaires. Cette plante est utilisée par les Yanomamï pour la production du curare. Clathrotropis est un genre mineur de la famille des

Fabacées, il est représenté par six espèces endémiques des regions tropicales d'Amérique du Sud. Ce sont des arbres de hauteur petite à moyenne avec des feuilles très larges, des fruits oblongs et plats et des graines larges. Les investigations ethno-bot- aniques effectuées en 1999 dans la région d'Orinoco au Venezuela ont montré que les espèces C. glaucophylla et C. macrocarpa (wapu kohi) sont d'une grande importance

économique pour les Yanomamï amérindiens du Venezuela; les graines sont un aliment quotidien et l'écorce est l'un des ingrédients du curare, poison enduit à l'extrémité des flèches.

Les extraits d'alcaloïde et de chlorure de méthylène ont été fractionnés par dif¬ férents moyens chromatographiques (VLC, CC, et HPLC). Ainsi 27 composés naturels ont été obtenus. L'extrait alcaloïde a livré un nouvel alcaloïde quinolizidinique (-)-13a- hydroxy-15a-(l-hydroxyethyl)-anagyrine ((-)-clathrotropine), et d'autres connus: (-)- anagyrine, (-)-thermopsine, (-)-baptifoline, (-)-epibaptifoline, (-)-rhombifoline, (-)-tinctorine, (-)-cytisine, (-)-TV-methylcytisine, (-)-lupanine, (-)-6a-hydroxylupanine et (+)-

5,6-dehydrolupanine. L'extrait de chlorure de méthylène a donné deux nouveaux dérivés de l'acide bétulique, 23-0-(4'-hydroxy-3'-methoxy-cinnamoyl)-acide bétulique et 23-0-(4'-hydroxy-3',5'-dimethoxy-cinnamoyl)-acide bétulique, un nouveau oc- pyrone, 5(5),6(5)-6(2-hydroxy-l-methylpropyl)-3,5-dimethyl-5,6-dihydro-2H-a- pyrone, et un nouvel isocoumarine, 6-hydroxy-8-methoxy-3-n-pentylisocoumarine. En plus onze substances connues ont été isolées: acide bétulique, ß-amyrin, glutinol, ß-sitosterol, stigmasterol, 7ß-hydroxysitosterol, 7ß-hydroxystigmasterol, ß-sitosterol-3-O-ß- glucoside, stigmasterol-3-O-ß-glucoside, acide 0-methyl-2'-evernique, et acide conflu- entique.

La détermination des structures a été réalisée par des moyens spectroscopiques, spectrométriques et physiques: ID et 2D NMR (1H, 13C, COSY, HSQC, HMBC,

ROESY, HSQC-TOCSY et INADEQUATE), UV, MS, et [a]D.

Cette étude phytochimique constitue une première de ce genre sur C. glaucophylla.

Les résultats obtenus montrent que cette plante a la capacité d'accumuler une grande 10

variété d'alcaloïdes quinolizidiniques ainsi que des triterpènes, sterols, depsides et iso- coumarines. C'est la première fois que des alcaloïdes quinolizidiniques ont été isolés d'un ingrédient du poison de flèche, et c'est aussi le premier rapport sur une espèce du genre Clathrotropis, en tant que composant d'un poison. SAMMENDRAG 11

SAMMENDRAG

I det foreliggende arbeidet ble sekundasrstoffer i barken fra Yanomamï curare- planten Clathrotropis glaucophylla Cowan isolert og karakterisert. Clathrotropis er en liten slekt under Fabaceae familien, med seks endemiske arter i det tropiske Soramerika.

De er smâ til medium store traer med svaert store blad og avlange og flate frukter med store fro. C. glaucophylla ble samlet i regnskogene ved ovre Orinocco i Venezuela i

1999. Etnobotaniske studier har vist at artene C. glaucophylla og C. macrocarpa (wapu kohi) er okonomisk svaert viktige for Yanomamï-indianerne i Venezuela. Froene spiller en betydningsfull rolle for emaeringen, og barken blir brukt som bestanddel i curare pilgift.

Fraksjonering av alkaloid- og diklormetan-ekstraktene fra barken av C. glauco¬ phylla ble utfort ved hjelp av forskjellige kromatografiske metoder (VLC, CC, HPLC) og forte til isolering av 27 naturstoffer. Fra alkaolid-ekstraktet ble et nytt quinolizidin- alkaloid, (-)-13a-hydroxy-15a-(l-hydroxyetyl)-anagyrine ((-)-clathrotropine), isolert i tillegg til elleve kjente quinolizidin-alkaloider: (-)-anagyrine, (-)-thermopsine, (-)-baptifoline, (-)-epibaptifoline, (-)-rhombifoline, (-)-tinctorine, (-)-cytisine, (-)-JV-metyl- cytisine, (-)-lupanine, (-)-6a-hydroxylupanine and (+)-5,6-dehydrolupanine. Under- sokelsen av diklormetan-ekstraktet forte til isolering av to nye betulinsyrederivater, 23-

0-(4'-hydroksy-3'-metoksy-cinnamoyl)betulinsyre og 23-0-(4'-hydroksy-3',5'-dime- toksy-cinnamoyl)betulinsyre, et nytt a-pyron, 5(S),6(

3,5-dimethyl-5,6-dihydro-2H-a-pyron, og et nytt isocoumarin, 6-hydroxy-8-methoxy-

3-n-pentylisocoumarin, i tillegg til elleve kjente substanser identifisert som: betulinsyre, ß-amyrin, glutinol, ß-sitosterol, stigmasterol, 7ß-hydroksysitosterol, 7ß-hydroksystig- masterol, ß-sitosterol-3-O-ß-glukosid, stigmasterol-3-O-ß-glukosid, 2'-0-metylever- niksyre, og konfluentiksyre.

Slrukturoppklaringen av de isolerte substansene ble utfort ved hjelp av spektrosko- piske, spektrometriske og fysikalske metoder: ID and 2D NMR ekperimenter (*H, 13C,

COSY, HSQC, HMBC, ROESY, HSQC-TOCSY og INADEQUATE), UV, MS, og [o(b

Dette er den forste fytokjemiske Studien av C. glaucophylla. Den viser at derme planten akkumulerer en mengde forskjellige quinolizidin-alkaloider i tillegg til triter- pener, steroler, depsider og isokumariner. Det er forste gang quinolizidin-alkaloider har 12

blitt isolert fra en pilgiftplante, og det er ogsâ den forste rapporten pâ at Clathrotropis- arter blir brukt i pilgift. RESUMEN 13

RESUMEN

El présente trabajo es un estudio sobre el contenido de metabolites secundarios de la corteza de Clathrotropis glaucophylla Cowan. Es una planta utilizada par los Yanomamï en la elaboraciôn del curare. Clathrotropis es un subgénero de la familia Fabaceae, y contiene seis especies endémicas en la parte tropical d'America del Sur. Son arboles de pequefïa a media altura con hojas grandes, y cuyas frutas son oblongas y planas con hue- sos grandes. Las investigaciones ethnobotânicas han mostrado que las especies C. glau¬ cophylla y C. macrocarpa (wapu kohi) tienen una gran importancia economica para los

Yanomamï indios de Venezuela. Las semiUas tienen una importante signifïcaciôn en la alimentaciôn, y la corteza es utilizada como ingrediente de la elaboraciôn del veneno de las fléchas (curare).

El fraccionamiento del extracto de alcaloides y lo de dichlorometano, realizado con métodos de la cromatogràfïcos como VLC, CC y HPLC, permitiô en el aislamiento de

27 productos naturales. En el extracto de alcaloides, se encontraron un nuevo alcaloid quinolizidinico, (-)-13a-hidroxi-15a-( 1 -hidroxietil)-anagirine ((-)-clathrotropine); como tambien los conocidos alcaloides quinolizidinicos: (-)-anagirine, (-)-thermopsine, (-)-baptifoline, (-)-epibaptifoline, (-)-rhombifoline, (-)-tinctorine, (-)-citisine, (-)-N- methilcitisine, (-)-lupanine, (-)-6a-hidroxilupanine et (+)-5,6-dehidrolupanine. Del extracto obtenido de la planta con dichlorometanol, se encontraron dos nuevos deriva- dos del âcido betulinico, 23-0-(4'-hidroxi-3'-metoxi-cinnamoil) âcido betulinico y 23-

0-(4'-hidroxi-3',5'-dimetoxi-cinnamoil) âcido betulinico, una nueva a-pirona,

5(5),6(5)-6(2-hidroxi-l-metilpropil)-3,5-dimetil-5,6-dihidro-2H-a-pirona, y una nueva isocumarina, 6-hidroxi-8-metoxi-3-n-pentilisocumarina. Asi mismo, se aislaron once compuestos conocidos: âcido betulinico, ß-amirin, glutinol, ß-sitosterol, stigmasterol, 7ß-hidroxisitosterol, 7ß-hidroxistigmasterol, ß-sitosterol-3-O-ß-glucosid, stigmasterol-

3-O-ß-glucosid, âcido 0-metil-2'-evernico, y âcido confluentico.

La determinaciôn estructural fue realizada con métodos espectroscôpicos, espectro- metricos y fïsicos: ID et 2D NMR (1H, 13C, COSY, HSQC, HMBC, ROESY, HSQC-

TOCSY et INADEQUATE), UV, MS, y [a]D.

Es la primera vez que se realiza un estudio fitoquimico sobre C. glaucophylla. En el se muestra que esta planta acumula una gran variedad de alcaloides quinolizidinicos, asi como triterpenos, esteroles, dépsidos y isocumarinas. Es la primera vez que alcaloides 14

quinolizidinicos han sido aislados de un ingrediente de veneno de flécha, y es tambien el primer reportaje de una especie de Clathrotropis utilizado como veneno de flécha. 1 INTRODUCTION 15

1 INTRODUCTION

1.1 Plants as a resource for medicinal remedies

Human survival has always depended on plants. Throughout the ages, plants have provided mankind the essentials of life, including food, raw materials for the manufac¬ ture of clothing and shelters, poison for hunting, and medical agents for relief from ill¬ ness. It is estimated that plants are still a major source of health care for more than 80% of the worlds population (Cordell, 1995). In the industrialized world, 25% of the pre¬ scribed medicines are substances derived from higher plants, and about 120 plant- derived compounds from ca. 90 plant species are used in modern therapy. Among the estimated 400000 plant species, only a small percentage have been phytochemically investigated and the fraction submitted to biological or pharmacological screening is even smaller (Potterat and Hostettmann, 1995, Hamburger et al., 1991). Based on these precedents, plants are an important source in the search for new drugs and lead com¬ pounds. Although powerful new technologies such as high-throughput screening and combinatorial chemistry increase the possibility of drug discovery dramatically, natural products still offer unmatched structural variety. The capacity of plants to synthesize very complex molecules, which are impossible to invent by any chemist, gives reason to investigate plants for bioactivity.

The alkaloids are structurally one of the most diverse classes of secondary metabo¬ lites, and over 10000 compounds are known. Their manifold pharmacological activities have always excited man's interest, and the human recognition of alkaloids is as old as civilization. These substances have been used as drugs, in potions, medicines, teas, poultices, and poisons for 4000 years. It is likely that, in the hunt for food and dealings with enemies, particular use was made of plants containing alkaloids for arrow poisons, and this use probably preceded their medical use. Even today these poisons are still in use in Africa and South America. Toxic substances normally show pharmacological 16 CLATHROTROPIS GLAUCOPHYLLA

activities in smaller doses, and the dreaded arrow poisons have provided medicine with

effective therapy. Well known examples are ouabain and k-strophanthin for acute car¬

diac insufficiency, physostigmine for the treatment of glaucoma and myasthenia and d- tubocurarine as a muscle relaxant in anesthesia. Sooner or later, arrow poisons will dis¬

appear, and therefore there is a real need to continue to evaluate these poisons for active constituents (Roberts and Wink, 1998).

In the search for new natural compounds it is first of all necessary to know which plant to select and what type of biological activity to look for. The fact that many plants play important cultural roles and have already undergone a traditional "human screen¬

ing", favour the ethnomedical approach, which give credence to information on the medicinal and cultural uses of the plant. Certainly, the highly biologically active arrow poisons are many times more likely to possess therapeutically valuable compounds than are extracts from higher plants selected at random. On the other hand, random bio- prospecting has been employed by the industry and is largely based on a rational "hit philosophy" without caring much about the ecological and cultural context of the plants

screened.

Since indigenous plant knowledge may hold clues for curing "western diseases", it is important to regulate the access and distribution ofphytogenetic resources (including secondary metabolites) as well as the corresponding intellectual property rights (Gertsch and Sticher, 2000). 1 INTRODUCTION 17

1.2 Scope of the present work

Clathrotropis is a small genus of the Fabaceae family, with 6 species endemic to the tropical South America. C. glaucophylla was collected and its ethnobotanical properties studied as part of another Ph.D. project (Gertsch, 2002), partly carried out in the rainfor¬ ests of the upper Orinoco in Venezuela. The collection of plant material was based on a contract between the Venezuelan Ministry of Environment and the ETHZ. The ethnobo¬ tanical investigation revealed that C. glaucophylla is of great economic importance to the Venezuelan Yanomamï Indians, the seeds have a great nutrition value during the rainy season, and the bark is used as an ingredient in arrow poison (curare).

There exist only a very few publications on the Clathrotropis species, most reporting on alkaloids, and C. glaucophylla has never been subjected to phytochemical analyses before. In the preliminary studies the bark showed the most promising results, compared to the leaves, with presence of alkaloids and antibacterial activity of the dichlo¬ romethane extract. These facts, in addition to the interesting ethnobotany, made the bark of C. glaucophylla an interesting candidate for a phytochemical analysis.

The aims of this thesis were, primary, the isolation of biological active alkaloids from the bark of C. glaucophylla and thus to support the traditional use as a curare ingre¬ dient, and secondary, to carry out a phytochemical analysis of the dichloromethane extract in order to provide possible new structures for drug discovery. The results of the phytochemical investigations on the alkaloid and dichloromethane extracts of Clathrotro¬ pis glaucophylla are presented in this thesis. 18 CLATHROTROPIS GLAUCOPHYLLA

Seite Leer /

,«tti^ ^»s>-*s h usasse « ^w*^b g fl B % 8s 2 THE FABACEAE FAMILY 19

2 THE FABACEAE FAMILY

The Fabaceae (or Leguminosae) comprises 650 genera and 18000 species, which are distributed all over the world. In economic importance it is second only to the grasses,

Gramineae, and in size third only to the Orchidaceae and the Compositae. Compared with those families and many others, the Fabaceae are notably "generalists", ranging from forest giants to tiny ephemaerals with great diversity in their methods of acquiring the essentials of growth and in their modes of reproduction and defence. Although the family is almost cosmopolitan in its distribution, it has much of its diversity centred in areas of carried topography with seasonal climates (Allen and Allen, 1981 ; Polhill R. M. et al., 1981).

2.1 Systematic classification

Taxonomists are either dividing the Leguminosae family into three distinct subfam¬ ilies: Mimosoideae, Caesalpinioideae, and Papilionoideae (division being based mainly on floral difference), or full family status are accorded to each of the three subdivisions, as Caesalpiniaceae, Mimosaceae, and Fabaceae in the Order Leguminales. However, this is a matter of choice as the distinctions between the three basic groups are clear and universally accepted. The recognition of one family or three families is depending on the emphasis given to the relatively few genera that are transitional between the three groups, as against the relatively numerous genera about whose position there is no pos¬ sible difference of opinion (Hutchinson, 1964; Allen and Allen, 1981; Polhill R. M. et al., 1981).

The caesalpinioid legumes (subfamily Caesalpinioideae; or Caesalpiniaceae) are the basic, presumably most primitive, and in some ways most diversified subfamily. They are primarily tropical, woody plants and poorly known in temperate regions.

The second subfamily is the Mimosoideae (Mimosaceae), which seems to be 20 CLATHROTROPIS GLAUCOPHYLLA

derived from the caesalpinioids. These are also mostly woody plants of the tropics and warm regions and poorly known in the temperate world. The last subfamily, the papilionoid legumes (Leguminosae, subfamily Papilion- oideae or Faboideae; or Papilionaceae), constitutes the great success story of the legumes. They are inter gradient with the caesalpinioids, their evident immediate ancestor, and their derivation is possibly polyphylenic. The kinds believed to be prim¬ itive are mostly woody, tropical or warm region plants whose flowers range from nearly regular to subpapilionaceous. From this beginning they have vastly proliferated

- in habit woody to herbaceous, perennial to annual - and in ecological adaptation.

Though yet preponderantly represented in the tropics, they have invaded all the temper¬ ate and arid climates and most of the habitats of the world. They are the dominant legumes of the temperate and developed countries where they are known as the legumes by most people. Nearly all temperate American and European cultivated legumes are herbaceous papilionoids (Isley, 1982).

2.2 Characteristics of the family

The Fabaceae consists of trees, shrubs, woody vines, and annual or perennial herbs.

The leaves are usually alternate and compound - bipinnate, simply pinnate, or palmate, and rarely simple. Inflorescence variously racemose, in simple racemes, panicles, spikes, or heads. The flower structure varies to the extent that 3 subfamilies are recog¬ nized; corolla typically 5-parted; stamens 3-many, mostly 10, free, or united by their filaments in various ways; pistil single, simple, free. The most evident common char¬ acter of the legumes is the gynoecium, which consists ofbut a single carpel (a relatively uncommon character among flowering plants), bearing 2 rows of parentally placed ovules on the ventral margin. As this pistil develops into a fruit, it becomes the pod or legume either dehiscent or indéhiscent (Allen and Allen, 1981; Isley, 1982).

2.3 Uses

Many species in Mimosoideae and Caesalpinioideae are valuable for their timber, dyes, tannins, gums, insecticides, medicines, and for fibres. In addition they are among the handsomest flowering trees, vines, and shrubs in the tropical areas where they 2 THE FABACEAE FAMILY 21

abound. Numerous members of the Papilionoideae (Fabaceae) are economically impor¬ tant, especially in temperate areas. They are edible and highly nutritional crops for human and animal consumption, for forage, fodder, ground cover, green manures, and erosion control, and as a major honey source. As pioneer plants in arctic regions they form the hub of an efficient nitrogen source for the entire ecosystem (Allen and Allen, 1981).

2.4 Papilionoideae (Fabaceae)

The essential features of the Papilionoideae (Fabaceae) are the papilionoid flower

(with a calyx-tube, the adaxial petal outside in bud and forming a flag at anthesis, and the lower petals housing the fertile parts), a hilar valve in the seed, and a change in the chemical profile with the ability to synthesize quinolizidine alkaloids and isoflavones.

Most Papilionoideae have a curved radicle and many can synthesize certain nonprotein amino acids which are not found elsewhere, such as canavanine. Root nodules are reg¬ ularly formed as in most Mimosoideae and unlike most Caesalpinioideae. Bipinnate leaves, complex leaf-glands and compound pollen grains, typical of Mimosoideae and some Caesalpinioideae, are lacking. The great majority of the genera are immediately recognizable by their distinctive flowers and seeds, and for that reason the Papilion¬ oideae are often given family rank. On the other hand there are genera in Swartzieae and

Sophoreae with open radial flowers, no apparent hilar groove and a general similarity to some genera in the Caesalpinieae. Therefore the borderline between the major group¬ ings are not sharply demarcated. On these grounds, as well as the unifying features of all legumes compared to members of any other family, the papilionoid group is also often regarded as a subfamily (Polhill, 1981).

The Papilionoideae (Fabaceae) are widely distributed from rainforest to the edges of dry and cold deserts. Rainforest trees and lianes occur principally in the Amazon and around the Gulf of Guinea, with a few notable disjunctions to tropical Asia and Austra¬ lia and with some sporadic radiation there. The main diversity in growth form and sys¬ tematic composition occurs on the planalto of Brazil, the Mexican region, eastern Africa, Madagascar, and the Sino-Himalayan region. The Mediterranean, the Cape and

Australia have a notable radiation from a few basic stocks. The Swartzieae are essen- 22 CLATHROTROPIS GLAUCOPHYLLA

tially South American, the main parts of the Sophoreae and Dalbergieae are pantropical with a western bias, while the Tephrosieae and Phaseoleae are more equally pantropi¬ cal. The other tribes tend to have regional centres with only sporadic disjunctions (Pol- hill, 1981b).

2.5 The Clathrotropis genus

The genus, Clathrotropis, which contains the species C. glaucophylla which is sub¬ jected to the current phytochemical study, is classified as follows (Polhill, 1981):

Order: Rosales

Family: Leguminosae Subfamily: Papilionoideae Tribe: Sophoreae Genus: Clathrotropis

2.5.1 Characteristics of the genus

The members of this genus constituted a section of Diplotropis (Bentham, 1862) until they were segregated and elevated to generic rank (Harms, 1908). Six species are endemic to the tropical South America, principally occurring in the Amazon and

Orinoco basins. They flourish in wet tropical forests (Allen and Allen, 1981, Polhill, 1981).

They are small to medium, unarmed trees with very large, imparipinnate leaves; leaflets 5-7; stipules small. Flowers white, often with a purple blotch, fragnant, the calyx 5-toothed. The fruit is broadly oblong, flat, and woody valved with one or few large Dioclea-like seeds (Allen and Allen, 1981; Gentry, 1993).

2.5.2 Previous phytochemical work on Clathrotropis species

Only three chemical studies have been carried out on Clathrotropis species. One concerning the silica content of Clathrotropis wood (Amos, 1951), the two others 2 THE FABACEAE FAMILY 23

reporting quinolizidine alkaloids from C. brachypetala seeds (Hatfield et al., 1980) and

C. macrocarpa leaves (Ricker et al., 1994), respectively.

The seeds of C. brachypetala were thought to have toxic properties because of a very low level of prédation. This indication was confirmed when ground and defatted seeds were found to produce paralysis, convulsions, and respiratory arrest when administered orally to mice by intubation. Fractionation of an ethanolic extract revealed seven quin¬ olizidine alkaloids: anagyrine, cytisine, A5-dehydrolupanine, rhombifoline, 11-allyl- cytisine, lupanine, and N-methylcytisine. Most of these alkaloids are widely distributed and have been reported from several members of the Leguminosae. The alkaloid content of C. brachypetala seeds (1.26% of fresh weight) appears to explain their observed tox¬ icity, and thereby also the low rate of seed loss due to prédation by animals (Hatfield et al., 1980).

GC-MS on an extract of leaves of C. macrocarpa revealed 13 alkaloids, in a total concentration of 165 p:g/g dry weight sample. The identified alkaloids were: cytisine, N- methylcytisine, N-formylcytisine, oc-isosparteine, ß-isosparteine, tinctorine, thermopsine, anagyrine, 5,6-dehydrolupanine, a-isolupanine, lupanine, epibaptifoline, in addi¬ tion to a N-methylcytisine type alkaloid (Ricker et al., 1994).

2.5.3 Clathrotropis glaucophylla

Tree 30 m high, with slender twigs, slightly pubescent. Leaves 5-parted, with 4-6 cm long, sporadic pubescent leave stem. The rugose stem of the leaflet is 6-8 mm long, the lamina 7-15 cm long and 4-7 cm wide, elliptic or oblong-obovate, at the basis cuneate or obtuse, at the distal end acute or obtuse, upper side slightly shiny, glabrous, at the under surface slightly pressed pilose and glaucescent, with the midrib slightly protrude on the upper side and prominent on the under surface, 8-11 primary veins are slightly protrude on the upper side, on the under surface prominent, the secondary veins lacking on the upper side and slightly protrude on the undersurface. Terminal inflorescence,

23.5 cm or longer, golden-brownish pilose, lanceolate bracts, 2-2.5 mm long, and linear fusiformed bracteoles. Pedicels ca. 1 mm long, with golden hairs. Calyx campanulate,

7.5 mm long and 5 mm wide, inside and outside golden sericeous, 5-toothed; the two upper more connate and 1.3-2.5 mm long, 2 mm wide. Corolla glabrous, vexillum 10 24 CLATHROTROPIS GLAUCOPHYLLA

mm long and 7 mm wide, roundish, claw 4 mm long. Wing and carina approximately of equal length, lanceolate-ovoid, ca. 9 mm long, 1.5 mm wide, at the base twofold auriculate. Stamina glabrous, on short filaments, stigma terminal, style glabrous, fila¬ mentous, ovary very short, tightly laniferous-setose; fruit not observed (Cowan, 1954). 3 ETHNOBOTANY 25

3 ETHNOBOTANY

3.1 The Yanomamï Indians

The Guyana shield, watershed of the Orinoco and Amazon basins, is almost entirely covered in tropical rain forests with occasional clearings, small savannas of uncertain origin. From the mountain ridges down to the plains that spread towards the big rivers, the deceptive uniformity of the landscape hides a great variety of natural resources.

Throughout the centuries the Yanomamï, originating from the Parima range, have migrated toward the river valleys, on the plains both to the south in Brazil, and to the north in Venezuela. Today their territory covers a 80,000 square kilometre area between

Venezuela and Brazil. The Yanomamï can be subdivided into four linguistic groups: the

Sanema to the north of Brazil, in the upper Auaris and Erebata region, the Ninam in the upper Paragua region, north to the Sierra Parima and around the Apiau and Mucajai riv¬ ers, the Yanomam in the watershed of the Catrimani and Demini rivers, as well as in the

Sierra Parima and finally the largest group, the Yanomamï who inhabit the upper

Orinoco and its big rivers Ocamo, Mavaca and Siapa. Of a total population to be esti¬ mated around 17,000 people, the Venezuelan Yanomamï number about 11,000. The

Yanomamï live in communities of 20 to 200 individuals in round houses (shapono or yano) with an open space in the middle. More than half of the about 110 communities in

Venezuela are still semi-nomadic and move their village every three to seven years.

Because of warfare they sometimes split up and start new communities. Yanomamï vil¬ lages are scattered irregularly, but usually thinly, over a vast area of tropical landscape.

Distances between villages can be as short as a few hours walk to as much as a week or

10 days, depending on the political relationship. Despite of the fact that the Yanomamï are hunters and gatherers, they also are shifting cultivators. About 60% of their food originates from cultivated plants. Each village has a big garden outside the community house, which is much bigger than the village itself. Virtually all groups cultivate certain 26 CLATHROTROPIS GLAUCOPHYLLA

plant species, such as plantains, bananas, corn, magie plants, and tobacco. Plantains are the daily bread during the dry season and it is always eaten together with the main pro¬ tein sources, such as meat, fish, and maggots. Palm fruits also play an important role

during that period. However, during the rainy season there is often a lack of plantains

and most Yanomamï groups gather the substitute wabu, which they often eat in great amounts. From the villages of the upper Ocamo to the remote groups of the upper

Mavaca and Sierra Unturan, wabu is known and appreciated as substitute of plantains

(Gertsch and Sticher, 2000; Gertsch et al., 2002).

3.2 Uses of Clathrotropis

3.2.1 The seeds

The important plantains substitute, wabu, mentioned above, is the Yanomamï name

for the fruits of Clathrotropis glaucophylla and C. macrocarpa. They ripe from Septem¬ ber to December, and during this period the Indians regularly collect them for food. In the Parima mountains there are forests of C. glaucophylla spp. that bear fruits also in

January and February. The normal way of preparing the seeds is to soak them in water,

after having cut them into smaller pieces (usually 2-4 pieces). The seeds are normally put into a basket which is left in water, preferably in a small stream or pond so that the water is changed continually, for 3-4 days. After the soaking time, the seeds are shortly boiled before eaten. The Yanomamï know that the seeds are toxic if not treated, and even if they sometimes chose a shorter preparation procedure, they never eat it untreated (Gertsch, 2000, pers.com.).

3.2.2 The bark

The bark of Clathrotropis is also known by the Yanomamï to be toxic, and they make use of it in arrow poison (curare). The most important constituent in the

Yanomamï curare is the bark of a small shrub from the genus Strychnos (Loganiaceae).

The bark is extracted with hot water until a black mass is obtained. To this mass, other constituents, like the bark of Clathrotropis, are added. Other constituents are, for exam¬ ple, Tabernaemontana (Apocynaceae), which in lack of Strychnos is used as the main 3 ETHNOBOTANY 27 constituent, Virola (Myristicaceae) and Psychotria (Rubiaceae). Before admixture of

Clathrotropis, the bark is cut into very small pieces, then it is allowed to macerate in water together with the Strychnos extract. The mixture is all the time being kept warm, but not boiling. After some time the mixture is filtrated and evaporated into a thick, tough mass. The arrows are dipped into this mass and left drying on the heat, before they are wrapped in leaves. The arrows can be kept in this form for a long time as long as water is avoided. The Yanomamï Indians use their poisonous arrows to kill smaller ani¬ mals like monkeys and sloths, and in warfare (Gertsch, 2000, pers.com.).

3.3 Curare

Since the discovery of the Americas in the sixteenth century the Old World knows about the existence of very potent arrow poisons known as "curare". This interest-rous¬ ing and feared material - the only new secret weapon the Aztecs, Toltecs, Incas and other native inhabitants of Central and South America were able to use against the armory and horses of the Spanish conquistadores - was very soon investigated by sci¬ ence-minded padres and educated followers of the invaders. "Curare" and "curarizing" are phenomenological terms describing neuromuscular block of impulse transmission of the motor end plates as a result of inhibition of acetylcholine (ACh) with the conse¬ quence of complete paralysis of skeletal musculature (Waser, 1972; Neuwinger, 1998).

In 1895, Rudolf Boehm in Germany classified curare poisons, more ethnographi- cally than chemically. He distinguished different types of curare preparations by the container in which they were packed:

(a) Tubo-, bamboo- or para-curare, packed in bamboo tubes. They are derived from Menispermaceae (chiefly Chondodendron and Curarea species) and originate mostly from the lower Amazonas region.

(b) Calabash-curare, the most active type, is packed in small gourds or cala¬ bashes. It is usually obtained from Loganiaceae {Strychnos species) and is mainly pro¬ duced in the region of the great rivers Orinoco and Amazonas.

(c) Pot-curare, which is contained in small earthenware pots, are mixed Loga- niaceae/Menispermaceae products and are found mainly in the area covered by the mid¬ dle reaches of the Amazon and to a lesser extent also in Guyana. 28 CLATHROTROPIS GLAUCOPHYLLA

Most curares comprise several components, which may be added for various rea¬ sons and which differ widely in accordance with local flora and tribal superstitions. The scientific investigation of the complex curare produced about 70 alkaloids, and the dif¬ ferent poisons differ very much in the composition of the alkaloids. However, all curares always have in common the presence of dimeric quartemary alkaloids, although of different structures, as the main curarizing principles. A good curare con¬ tains up to 12% quartemary alkaloids. The distance between the two protonated N for maximum curare activity is about 0.85-0.9 nm. The main bases of calabash or Strych¬ nos curare are easily water-soluble bis-quaternary dimeric indole alkaloids formed by the union of two equal or different monomelic molecules of strychnine type and con¬ taining two quartemary ammonium groups. The major alkaloid of the tubo or Chondo¬ dendron curare is (+)-tubocurarine.

When introduced into the bloodstream, all the dimeric quartemary curare alkaloids are highly active muscle relaxants, acting by nondepolarizing and competitive mecha¬ nism at the neuromuscular junction, competing with ACh for the active surface of the receptor and thus blocking the nerve-muscular transmission. Blockade of the postsyn¬ aptic nicotinic acetylcholine receptors causes progressive paralysis of voluntary move¬ ment and, as final result, complete paralysis of the skeletal or striated muscle apparatus.

The muscle involvement follows a definite order: eye and ear muscles are the first to be affected, then the neck, limb, trunk, intercostal muscles, and diaphragm; death results

from anoxia caused by respiratory failure. Artificial respiration prevents death. Physos- tigmine from the African arrow poison plant Physostigma venenosum, immediately given, antagonizes all paralytic symptoms.

Curare poisons are very stable, once prepared. Although more than 140 years old, curare samples present in various European museums were shown to have remained very effective.

The transformation of curare from deadly poison to a valuable medical agent is one of the best examples of the refinement of the crude natural product to an indispensable aid in modem medicine. In 1935, King announced the isolation of a pure crystalline alkaloid, (+)-tubocurarine chloride, from a museum sample of tubo-curare of unknown botanical derivation. Seven years later, the same compound was found in a crude curare known to have been prepared from the single plant Chondodendron tomentosum. In 3 ETHNOBOTANY 29

1939, curare was introduced in clinical medicine, first as a preventive of traumatic com¬ plications in metrazol convulsion shock therapy and electric shock therapy. Since 1942 the pure alkaloid (+)-tubocurarine has become a useful neuromuscular relaxant in all forms of anaesthesia.

Two other curare alkaloids, C-calebassin and C-toxiferine from calabash curare, are also used in clinical medicine, e.g., in Switzerland, especially for tetanic convulsions; their duration of effect is longer-lasting than that of (+)-tubocurarine with fewer effects on respiration (Waser, 1972; Neuwinger, 1998). 30 CLATHROTROPIS GLAUCOPHYLLA

lui li\ iwUi 4 QUINOLIZIDINE ALKALOIDS (LUPINE ALKALOIDS) 31

4 QUINOLIZIDINE ALKALOIDS

(LUPINE ALKALOIDS)

4.1 Introduction

More than 200 naturally occurring quinolizidine (lupine) alkaloids are known.

These represent about 2% of the 10000 known alkaloids in nature (Saito and Murakoshi,

1995). Alkaloids of the quinolizidine class are found in bacteria, fungi, higher plants, invertebrates and vertebrates; and both terrestrial and marine sources are represented.

Simple bicyclic compounds form a rather small subset of the lupine alkaloids, the overwhelming majority of which have tricyclic or tetracyclic structures based on the quinolizidine motif. They are characteristic secondary metabolites of the Fabaceae and are especially abundant in the tribes Genisteae, Sophoreae and Thermopsideae, although representative examples have also been isolated from several other plant fam¬ ilies (Wink et al, 1995; Michael, 2001).

These alkaloids are of importance to mankind because they are toxic to human and livestock, and some of them show pharmacological activities.

4.2 Structural types

It is reasonable to classify the quinolizidine alkaloids (QA) into four groups: (/) lupinine, (if) sparteine/lupanine, (Hi) a-pyridone (anagyrine/cytisine), and (iv) matrine.

According to this classification, the lupine plants may then be divided into three major categories: (i) species containing alkaloids of the lupinine group, (if) species containing alkaloids of the sparteine and anagyrine groups, but no matrine alkaloids, and (Hi) spe¬ cies containing alkaloids of the matrine group. The genera containing matrine are the most primitive, followed by the genera accumulating sparteine- and oc-pyridone-type 32 CLATHROTROPIS GLAUCOPHYLLA

alkaloids, but no matrine bases. The most advanced genera are those containing lupinine-type alkaloids (Fig. 4.1) (Saito and Murakoshi, 1995).

4.3 Biosynthesis

In the I960's it was established by Schütte, Mothes and coworkers that lysine and its decarboxylation product cadaverine serve as the only precursors for the bi-and tet¬ racyclic QA. This has been confirmed by Spenser and Robins and their respective coworkers in the 70s and 80s, using deuterium-, carbon 13- and nitrogen 15-labelled cadaverines and NMR techniques for analyzing the products. However, details of the processes involved in the construction of lupinine and the tetracyclic quinolizidine

alkaloids are poorly understood (intermediates may well be enzyme-bound), and theo¬ ries are being continually revised (Wink, 1987).

The alkaloids of lupinine-type are formed with two units of cadaverine; whereas the

alkaloids of sparteine-, anagyrine-, and matrine-types are constructed with three units

of cadaverine. The diiminium cation is proposed as a most likely intermediate in the biosynthesis of the tetracyclic alkaloids (Fig. 4.2) (Saito and Murakoshi, 1995).

There is no agreement whether sparteine or lupanine is the first labelled tetracyclic

alkaloid. Since sparteine is, chemically, the more "primitive" alkaloid, many authors

assume that it is also the first alkaloid and that lupanine and the other alkaloids derive

from it. However, it is lupanine which is most widely distributed in the plant kingdom

and there is a number of species which do not contain sparteine (Wink, 1987).

Tracer studies with QA have revealed a number of sterochemical details and have

established some of the later steps in the formation of pyridones in tetracyclic and tri¬ cyclic QA. For the biosynthesis of the cytisine-type alkaloids it has been postulated that lupanine is converted into 5,6-dehydrolupanine, via 6-hydroxylupanine. 5,6-dehydro-

lupanine is further oxidized to anagyrine, rhombifoline, cytisine and N-methylcytisine (Fig. 4.2).

Labeling studies for (+)-sparteine and (-)-N-methylcytisine suggest that it is ring A

of a tetracyclic precursor that must be degraded and ring D that is converted into a pyri-

done. Comparison of the stmcture and sterochemistry of (-)-anagyrine with those of

(+)-sparteine suggests that if they are formed from the same tetracyclic intermediate GENISTEAE Lupinus CHoOH Cytisus

.Ns Lupininetype

THERMOPSIPEAE Thermopsis Baptisia

q a-Pyridone type EUCHRESTEAE Euchresta

SOPHOREAE Sophora Echinosophora Maackia

Matrine type

Papilionoideae

Figure 4.1 Phylogenicrelationshipsof tribes and genera containinglupinealkaloids in the Papilionoideae(Fabaceae) (7R:9R) series

(-)-Tetrahydrocytisine

£ (-)-Rhombifoline Ö (-)-Cytisine O (-)-N-methylcytisine

Figure 4.2 Possible biosyntheticpathway of lupinealkaloids (Saitoand Murakoshi, 1995) 4 QUINOLIZIDINE ALKALOIDS (LUPINE ALKALOIDS) 35 with identical absolute configuration at C-6 and C-l 1, then it is likely that ring A of the tetracyclic intermediate would be converted into a pyridone in order to form anagyrine.

However, the theory was tested and it was clear from the labeling patterns that if (+)- sparteine and (-)-anagyrine are formed from the same tetracyclic intermediate, then it must be ring D that is converted into a pyridone (this is the same orientation as that required for the formation of the pyridone in (-)-N-methylcytisine (Wink, 1987; Ohmiya et al., 1995; Robins, 1995).

Some interesting relations are observed in the absolute configuration and biosynthe¬ sis of the QA. The absolute configuration of C-6, C-7, C-9 and C-l 1 are determined at the ring-cyclization steps and is enzyme dependent. The sparteine-lupanine-type alka¬ loids accumulate in plants as two series of enantiomers, the (7S:9S) series and the

(7R:9R) series. The oc-pyridone-type alkaloids occur only as (7R:9R) enantiomers. No (7S:9S)-a-pyridone-type alkaloids have been found, suggesting lack of enzymatic activity of a-pyridone formation for (7S:9S)-series alkaloids (Fig. 4.2) (Ohmiya et al., 1995; Saito and Murakoshi, 1995).

QA formation is observed in the aerial parts of legumes only, although the whole plant accumulates these alkaloids. The alkaloids are produced by leaf chloroplasts, and are then distributed all over the plant via phloem and stored in epidermal cells and in seeds. As a consequence, alkaloid profiles are more diverse in leaves than in seeds, but the highest concentrations are usually reached in the mature seeds, which can contain up to 5% alkaloid per dry weight (equivalent to 200 mmol/kg). Legumes with many small seeds have a lower, species with few and big seeds generally the highest alkaloid con¬ tent (Wink, 1987; Wink et al, 1995).

The formation of cadaverine is catalyzed by lysine decarboxylase, and this enzyme is localized in chloroplast stroma. The activity of this enzyme and the alkaloid content are related to leaf re-greening. Lysine decarboxylase is present also in alkaloid free higher plants. During seedling development, the formation and accumulation of lupine alkaloids decreases initially and then increases, corresponding to the development of chloroplasts (Ohmiya et al., 1995). 36 CLATHROTROPIS GLAUCOPHYLLA

4.4 Spectroscopy for structure determination of lupine alkaloids

For the identification of QA IR-spectroscopy, which had been used as the main

method a few decades ago, has been widely substituted by NMR spectroscopy and

mass spectrometry. Often mass spectrometry has been the method of choice in phy¬ tochemical studies of QA, especially in combination with GC (GC-MS), when the

identification of known stmctures is demanded. In most cases, the molecular ions of

lupine alkaloids are detectable in electron impact (EI) ionization technique, and, there¬

fore, useful for determination of molecular mass and composition by the combination with high resolution mass spectrometry. (In the EI mass spectra of a-pyridone alka¬

loids, the fragment ions at m/z 146 and 160 appear generally, see Fig. 5.13) (Saito and Murakoshi, 1995).

Proton and carbon-13 nuclear magnetic resonance (NMR) techniques are essential

and extremely powerful methods for the stmcture elucidation of new QA and are often used in combination with MS techniques.

Since, in the ^-NMR spectra of lupine alkaloids, most signals of alicyclic hydro¬ gens appear in the range of Ö 1-2.5, it is not easy to assign all overlapping signals. How¬

ever, several characteristic features in the spectra are helpful for structural elucidation.

In the ^-NMR spectra of lupanine-type alkaloids, only the signal of H-10a (equato¬ rial) resonates downfield ca. 54.5, because of the deshielding effects of the amidocar- bonyl residue. This downfield shift is even more remarkable in the signals of the both

H-10a and H-10ß of a-pyridone alkaloids. These two methylene hydrogens appear as the AB part ofABX-type system by spin-coupling with H-9. This signal pattern is com¬ mon in the spectra of a-pyridone alkaloids and can be ascribed to the steric relation of the C-l 0 hydrogens and the carbonyl residue, which is positioned at the center of the angle between H-10a and H-10ß (Saito and Murakoshi, 1995).

For 13C-NMR spectroscopy, consideration of the y-effects gives useful information on the stereochemistry of a molecule. The up-field shift of the C-8 signal of sparteine compared with that of a-isosparteine is explained by the close approach of C-8 to the N-16 electron pair that arises from the boat form of the C-ring. A similar high-field shift of the C-8 signal of (-)-anagyrine having the C/D c/s-quinolizidine stmcture was also observed, but this was attributed to the y-interaction of C-l2 and C-l7 and C-8

(vide supra) (Ohmiya et al., 1995). 4 QUINOLIZIDINE ALKALOIDS (LUPINE ALKALOIDS) 37

The substitution effects of hydroxy and N-oxide groups are observed in chemical shifts of carbon signals, and they are, therefore, useful for determination of the positions of substitutions. Among the 2D-techniques, NOESY is quite useful for the determina¬ tion of configuration and conformation of lupine alkaloids (Saito and Murakoshi, 1995).

4.5 Biological activities

QA have a wide variety of biological activities. They are toxic or inhibitory for most organisms. Sparteine is used therapeutically as an antiarrythmic drug and in obstetrics, whereas lupinine and matrine have been in use in folk medicine in Eastern Asia (Wink,

1987). A summary of reported pharmacological activities is given in Table 4.1

Stockmen in the western U.S. have recognized the toxicity of lupines since the late

1800s, and large livestock loses, particularly in sheep, were reported in Montana and other western states. Signs include central and peripheral nervous system stimulation followed by depression, frequent urination and defecation, dilated pupils, trembling, incoordination, excessive salivation and nervousness. These progress to muscular weak¬ ness recumbency, and eventually death from respiratory paralysis. Signs may appear as early as one hour after ingestion and progressively get worse over the course of 24 to 48 hours even if further ingestion does not occur. Generally, if death does not occur within this time frame, the animal recovers completely. The clinical signs of poisoning are pre¬ sumed to be due to quinolizidine and piperidine alkaloids (Panter et al., 1999).

The pyridone alkaloids such as cytisine and anagyrine are more acute toxic than the corresponding saturated alkaloids such as sparteine and lupanine. Both cytisine and anagyrine have been implicated as teratogens in higher animals, and it has been shown by Keeler (Keeler, 1973) that anagyrine has mutagenic properties and produces malfor¬ mations ("crooked calf disease") in early foetal stages. Common molecular targets for

QA are nicotinic and muscarinic acetylcholine receptors and Na+, K+ -ion channels, besides protein biosynthesis. Variation in structural diversity is thought to enhance tox¬ icity (i.e. enabling the attack at several targets) and to reduce the chance that a herbivore will develop resistance towards alkaloids (Kinghom and Balandrin, 1984; Wink et al., 1995). Little is known about individual alkaloid toxicity, however, 14 alkaloids isolated 38 CLATHROTROPIS GLAUCOPHYLLA

from Lupinus albus, L. mutabilis, and Anagyris foetida were analyzed for their affinity

to nicotinic and/or muscarinic acetylcholine receptors (Schmeller et al., 1994). Of the 14 compounds tested, the a-pyridones (N-methylcytisine and cytisine) showed the highest affinities at the nicotinic receptor (IC50: 0.05 and 0.14 jjM, respectively), as measured by displacement of radio-labeled ligand H-nicotine, while several quinoliz¬ idine alkaloid types including the teratogen anagyrine showed weak activity at the mus¬

carinic receptor (IC50: 132 p:M). Lupanine which is widely distributed in legumes as a major alkaloid, displayed an IC50 of 5 |im at the nicotinic receptor and is 100 times more active than hydroxylated lupanines or alkaloids of the multiflorine series.

Because nicotinic and muscarinic receptors are widely distributed within the body, a number of tissues and organs will be affected, to a certain degree, if an animal con¬

sumes these plants. Additionally, QA such as lupanine and sparteine inhibit Na+- and

K+ -channels, thus blocking the signal transduction in nerve cells at a second critical point. This might potentiate the toxicity caused by binding of the alkaloids to the ACh- receptors (Schmeller et al., 1994; Wink, 1998).

Alkaloids such as sparteine have long been known to possess antiarrythmic ("car¬ diotonic") and uterotonic properties in both experimental animals and in humans.

Experimental evidence indicates that the pharmacological actions of sparteine, espe¬ cially on the utems, are mediated via prostaglandins. Its actions have also been postu¬ lated to be related to their ability to bind or chelate divalent cations such as calcium.

Diamines such as sparteine and its steroisomers can act as ligands for divalent cations

such as magnesium and calcium. In addition, sparteine has been implicated in mem¬ brane processes involving potassium conductance. Studies with quinolizidine alkaloids

such as sparteine, multiflorine, and lupinine have established that they have a depres¬

sant effect on the central nervous system of experimental animals. In addition, human

subjects who cannot rapidly metabolize sparteine show signs of CNS activity, such as diplopia, blurred vision, dizziness, and headache (Kinghom and Balandrin, 1984).

The syndrome known as "crooked calf disease" is associated with lupine ingestion and was first reported in the late 1950s and is associated with various skeletal contrac- ture-type birth defects and occasionally cleft palate. Through epidemiologic evidence and chemical comparison of teratogenic and non-teratogenic lupines, the QA anagyrine 4 QUINOLIZIDINE ALKALOIDS (LUPINE ALKALOIDS) 39

was determined to be the teratogen (Keeler, 1973). If one compares binding affinities of anagyrine in nicotinic versus muscarinic receptors, there is 16 times greater binding affinity to muscarinic receptors. Perhaps information about the maternal or fetal mech¬ anism of toxicity or teratogenicity of anagyrine can be applied from this comparison

(Panter et al., 1999). 40 CLATHROTROPIS GLAUCOPHYLLA

Table 4.1 Examples ofbiological activity of lupine alkaloids (Kinghorn and Balandrin, 1984; Ohmiya et al, 1995).

Alkaloid Type of activity

Spartein Oxytocic; uterotonic, antiarrythmic, diuretic, respiratory depressant/stimulant, hypoglycemic, inhibition of natural killer cell growth, bacteriostatic, chemical defense, inhibition of seed germination

Lupanine Antiarrythmic, hypotensive, hypoglycemic, inhibi¬ tion of ß-glucosidase, bacteriostatic, chemical defense, inhibition of seed germination

Cytisine Teratogenic, respiratory stimulant (nicotine-like activity), hallucinogenic, oxytocic; uterotonic, inhi¬ bition of edema, myopathy, nematocidal, chemical defense

Anagyrine Teratogenic, inhibition of acetylcholinesterase, myopathy, nematocidal

N-Methylcytisine Hypoglycemic, inhibition of edema, myopathy

Retamine Uterotonic, hypotensive, diuretic

13 -Hydroxylupanine Antiarrythmic, hypotensive, bacteriostatic

Calpurine Ichthyotoxic, antiarrythmic, hypotensive

Multiflorine CNS depressant, antidiabetic

Matrine Antipyretic, cardiotonic, antiulcerogenic

5,6-Dehydrolupanine Myopathy

Thermopsine Myopathy

Sophoridine, Allomatrine, Cardiotonic Sophoramine

Sophoramine N-oxide, Epilupinine Hypoglycemic N-oxide

17-Oxosparteine Inhibition of edema

Ammodendrine Inhibition of acetylcholinesterase

Matrine N-oxide Antiulcerogenic

Augustifoline Bacteriostatic 5 RESULTS AND DISCUSSION 41

5 RESULTS AND DISCUSSION

5.1 Preliminary studies

5.1.1 Extraction

Air-dried and ground material, 13 g bark and 11 g leaves, were extracted 3 times with petroleum ether (150 ml) by shaking the suspension for 24 hours at room tempera¬ ture. Then the plant material was extracted with dichloromethane (3 x 150 ml) for 24 hours and finally with methanol (3 x 150 ml) for 24 hours. Leaves (10 g) and bark (10 g) were also extracted separately with water (3 x 100 ml) for 24 hours. The organic extracts were evaporated to dryness on rotavapor at 30 °C. The water extracts were evaporated on rotavapor at 30 °C and subsequently lyophilised.

A specific alkaloid extraction was carried out from a MeOH extract (2.3 g) of the bark, as the literature had revealed alkaloids from this genus.

The yields are given in Table 5.1.

Table 5.1 Preliminary extraction yields from 13 g bark and 11 g leaves

Extract C. glaucophylla bark C. gl'lucophylla leaves

Petroleum ether 37.9 mg 148.1 mg

Dichloromethane 71.3 mg 214.4 mg

Methanol 2457.8 mg 2679.9 mg

Water 780.0 mg 910.2 mg

Alkaloid 12 mg 42 CLATHROTROPIS GLAUCOPHYLLA

5.1.2 TLC screening

The extracts were analyzed by Thin Layer Chromatography (TLC) in order to get

more information about the chemical composition of the extracts (for details see "Thin

layer chromatography" on page 141).

Before detection only chlorophyll could be seen at the front. By observing the TLC plates under UV 254 and 366, especially the apolar extracts showed many bands. The

MeOH extract of the leaves also showed quite a few bands, whereas the bark MeOH

extract and the water extracts showed very few.

After detection with vanillin/H2S04 the apolar extracts, in addition to the MeOH

extract of the leaves, showed the most bands. The polar extracts of the bark didn't show

many, except some purple and brown/grey bands in the more polar region. The brown/

grey bands close to the start were probably sugars. The petroleum ether extract and the

dichloromethane extracts showed a very similar partem with mostly purple and blue-

green bands. These were thought to be terpenoids and steroids. The MeOH extract of

the leaves showed some purple and yellow to brown bands in the apolar region and a

similar partem as the MeOH extract from the bark in the polar region.

The methanol extracts gave some positive results by observing the TLC plates under UV 366 after spraying with NST/PEG. The bark extract revealed some yellow/

orange bands in the more polar region, and from the leaves extract a very fluorescence

blue band was observed in addition to a yellow/orange band. These were also in the polar region. The yellow bands were probably flavonoids, whereas the blue band could

be a phenolic acid. From the apolar extracts, especially the dichloromethane extract,

also some yellow and blue bands were observed, mostly in the apolar region and at the

start point.

After spraying with Dragendorff reagent, orange bands were observed in all the

bark extracts indicating the presence of alkaloids. The specific alkaloid extract showed 5-6 bands after Dragendorff detection.

5.1.3 Biological assays

All extracts were tested for brine shrimp lethality-, KB cell cytotoxicity-, antibac¬

terial- and antioxidant-activitity (for details see "Biological assays" on page 141). 5 RESULTS AND DISCUSSION 43

None of the extracts showed significant activity in the brine shrimp and the KB cell test systems. The results of the antibacterial and antioxidant tests are given in Table 5.2 and

Table 5.3.

Table 5.2 Antibacterial activities of the tested extracts

Part Extract ug B.c. E.c. M.l. Ps.a. St.e.

500 ++ PE 200 ++.++. ++

500 +++ - +++ - +++ DCM

200 ++ - +++ - +++ Bark

500 + - MeOH 200

500 - + - + H,0 200

500 - PE

200 -

500 - - + DCM

200 - - + Leaves

500 - MeOH

200 -

500 - H,0

200 -

Antibacterial activity: size of the inhibition zone on the agar plate (-: no inhibition; +: 1 mm; ++: 2 mm; +++: >2 mm). Tested organisms: Bacillus cereus (B.c.), Escherichia coli (E.c), Micrococcus luteus (M.L), Pseudomonas aeruginosa (Ps.a.), Staphylococcus epidermidis (S.e.) 44 CLATHROTROPIS GLAUCOPHYLLA

Table 5.3 Antioxidant activity of the tested extracts

Part Extract (ig Activity PE 100 Bark no

DCM 100 yes

MeOH 100 yes

H20 100 yes

Alkaloid 50 yes PE 100 Leaves no

DCM 100 no

MeOH 100 yes

H20 100 yes

5.2 Extraction

Normally when one want to study the alkaloids of a plant, a specific alkaloid extraction, making use of acid and base, is carried out on an EtOH extract of the plant or directly on the plant material. However, this would possibly destroy other substances than the alkaloids and make a general phytochemical analysis impossible. As the pre¬ liminary studies had shown the presence of alkaloids in all the bark extracts, it was decided to leave 250 g of the total 1 kg bark, for an EtOH (75%) extraction and subse¬ quently an alkaloid extraction. This yielded 750 mg alkaloid extract. The other 750 g of the bark from C. glaucophylla were successively macerated to exhaustion with DCM,

MeOH and MeOH-H20 (80:20) at room temperature. These extractions offered a pre¬ liminary fractionation of the compounds in the plant. The results are given in Fig. 5.1.

The dichloromethane extract of the bark had shown the most interesting biological results in the preliminary studies and was chosen for a general phytochemical investi¬ gation. As the methanol extract and the methanol/water extract had not revealed many interesting spots on TLC, except for the alkaloids, and as the results from the isolation procedure of the first alkaloid extract were rather scarce, due to small amounts, it was decided to carry out an alkaloid extraction on these extracts as well. The resulting alka- 5 RESULTS AND DISCUSSION 45

loid extract was then added to the remaining fractions of the first alkaloid extract to give a second extract of 1.5 g. The extraction and isolation procedures are shown in Fig. 5.1, Fig. 5.2, Fig. 5.3, Fig. 5.4, Fig. 5.5 and Fig. 5.6.

f Dried bark of "\ c Dried bark of ^\ C. glaucophylla

EtOh1 75% } \ 32.I39 g DC MeOH MeOH 80% Alkaloid 7.1 3g 71.(35 g 11.44 g extraction 1 j

0.73g AIkaloid r exti action 2 Compounds Comp)ounds r \ - 2 and 10 13 -27 1.5 Og ^ I

(*omp

Figure 5.1 Extraction scheme of C. glaucophylla 46 CLATHROTROPIS GLAUCOPHYLLA

Extracts from C. glaucophylla

1) 0.1 N HCl 2) extracted with DCM (3x)

f } Organic phase Water phase

1)topH9withNH4OH 2) extracted with DCM (4x)

f \ Water extract DCM extract

1)topH11 withNH4OH 2) extracted with DCM (4x)

Alkaloid extracts

Water extract*

* Negative Meyer test

Figure 5.2 Extraction ofalkaloids from MeOH-, MeOH 80%-, and EtOH 75% -extracts of C. glaucophylla bark 5 RESULTS AND DISCUSSION 47

5.3 Fractionation and isolation

Fractionation and isolation were carried out from the alkaloid extracts and dichlo¬ romethane extract with vacuum liquid chromatography (VLC), open column chroma¬ tography (CC) and HPLC.

The phytochemical investigation of these extracts yielded 27 natural products, 5 of which have not been reported in the literature.

Fractionation processes were conducted by the help of TLC and *H NMR spectros¬ copy. For the DCM extract the first fractionation step was bioactivity guided, the result showing activity against gram positive bacteria in the more polar fractions. Due to small quantities, the further isolation process was carried out without TLC bioautographic assays. The solvent systems for column chromatography were optimized using TLC.

5.3.1 Fractionation of the alkaloid extracts

The fractionation and isolation procedure of the two alkaloid extracts are described in paper I. Details, including exact mobile phases and sample amounts, are shown in

Fig. 5.3 and Fig. 5.4. As a primary fractionation the alkaloid extract 1 (730 mg) was sub¬ mitted to CC on silica gel, using mixtures of dichloromethane and methanol of increas¬ ing polarity as mobile phase. The further fractionation afforded two alkaloids (2 and

10). The remaining fractions of alkaloid extract 1 were added to the alkaloid extracts from the MeOH- and MeOH 80% -extracts to give alkaloid extract 2 (1.5 g). This was submitted to CC on silica gel, using DCM-MeOH-NH3 of increasing polarity as mobile phase. In total 12 quinolizidine alkaloids were isolated from C. glaucophylla in this study. 48 CLATHROTROPIS GLAUCOPHYLLA

Alkaloid extract 1 0.73 g

CC Si60 DCM-MeOH 98:2-9:1

Fr. 1-7 Fr. 8 Fr. 9 Fr. 10 Fr. 11-14 (96 mg) (71 mg)

CC Al203 60 CC Al203 60 Hex-EtOAc-EtOH Hex 100%- Hex-EtOAc-EtOH 6:3:1 w 97:6:4-6:3:1

é 10 (3rng) (9mg)

Figure 5.3 Isolation scheme of the alkaloid extract 1 of C. glaucophylla

5.3.2 Fractionation of the DCM extract

As a crude separation procedure VLC using silica gel as stationary phase and a sol¬ vent gradient of hexane and ethyl acetate and subsequently ethyl acetate and methanol was employed. The further fractionation has been carried out using either VLC or open column chromatography. The solvent systems were optimized using *H NMR and TLC

(for details see "Thin layer chromatography" on page 141). The isolation and purifica¬ tion procedures of compounds 14 and 15 are described in paper II. The complete isola¬ tion procedure is shown in Fig. 5.5 and Fig. 5.6 Alkaloid extract 2 1.5g

CC Si60 DCM-MeOH-*, 98:2:* - MeOH 100% J Fr.1 Fr.2 Fr.3 Fr.4 2 Fr.6 Fr.7 Fr.8 (35 mg) (17 mg) (10 mg) (16 mg) (324 mg) (72 mg) HPLC A HPLC A HPLC A T t t CC Si60 6 7 3 CC Si60 DCM-MeOH-* DCM-MeOH-* (9mg) (1.2mg) (3mg) 99:1:*-MeOH 100% 99:1/*-MeOH 100%

Fr.6.4 Fr.6.1 Fr.6.2 Fr.6.3 Fr.6.5 Fr.6.6 Fr.6.7 Fr.6.8 Fr.7.1 Fr.7.2 Fr.7.3 (46 mg) (41 mg) (100 mg) (37 mg) (30 mg) HPLCB HPLCB HPLCB HPLCB 94:6:* 94:6.* 85:15:* CC 60 85:15:* Al203 y Hex-EtOAc-EtOH 12(18mg) 110:7:3-100:7:4 1 (9 mg) 5 (9 mg) 2 (20 mg) 8 (7.5mg) 4 (3.5mg) v HPLC A 9 11 Si60 8 mm HPLCB Si6016mm (15 mg) (8mg) DCM-MeOH-* M.21T1INH3/ 97:3:* DCM-MeOH-* 500 ml DCM-MeOH

Figure 5.4 Isolation scheme of the alkaloid extract 2 of C. glaucophylla oOl Parti DCM<î extract 7.13

VLC Si60: Hex-EtOAc 8:2 - EtOAc 100% EtOAc-MeOH8:2-1:1

Fr. Fr. 1 Fr. 3 Fr. 4 Fr.5 Fr. 6 7 Fr. 8 Fr. 2 Si60 Si60 Si60 Hex-EtOAc 7:3 - 3:7 Hex-EtOAc 7:3 - 4:6 Hex-EtOAc 95:5 - DCM-MeOH 8:2 DCM-MeOH 9:1 - MeOH 100% EtOAc 100%

Fr. 8-10 Fr. 1-2 Fr. 3 Fr. 4-6 Fr. 1-3 Fr. 4-5 Fr. 6-12 Fr. 1 Fr. 2 Fr. 1 Fr. 2-6 19 27 Si60 (4.8mg) (22 mg) I DCM-MeOH Sièo 99:1-9:1 Hex-DCM 7:3 13 (7.5mg) DCM 100% Si60: DCM-Hex1:1 -8:2 Fr. 2-3 Fr. 4-5 Fr. 6-12 ; Hex-DCM-EtOAc 5:5:2 Fr. 2-4 Hex-EtOAc 1:1-EtOAc 100% Fr. 1-2 Fr. 3 Fr. 4-6 Fr. 7 26 RP-18 Si60 MeOH-H20 9:1 (20 mg) RP-18 DCM 100%- I H20-ACN 10:0-0:10 Hex-DCM 7:3 Fr. 1-2

Fr. 1-6 Fr. 8-12 Fr. 9 r—[ Fr. 1-2 Fr. 3 Fr. 4-5 20+21 i Sephadex (45 mg) precipitation Sephadex Cyclohex-DCM-MeOH Cyclohex-DCM-MeOI- 7:4:1 13 7:4:1 Fr. 1 Fr. 3 L ï (11 mg) 17 22+23 14 (2.3mg) 15(1-1mg) (15 mg) (5mg)

Figure 5.5 Isolation scheme, partI,of the DCM extract of C. glaucophylla.Fractions with antibacterialactivityare double underlined ? /DCM extract7.13 g ^\ Part II

VLC Si60: Hex-EtOAc 8:2 - EtOAc 100% EtOAc-MeOH8:2-1:1

Fr. 9 Fr. 10 Fr. 11

' VLC Si60: Hex-EtOAc 1:1 - EtOAc 100% DCM-MeOH 9:1 - MeOH 100% VLC Si60: Hex-EtOAc 1:1 - EtOAc 100% DCM-MeOH 9:1 - MeOH 100% Fr. 1 Fr. 2 Fr. 3-4 Fr. 5-6 Fr. 7

Si60: Si60: 6:4 Hex-EtOAc DCM-MeOH 98:2 - 8:2 Fr. 1-6 Fr. 7 Fr. 8 16 Fr. 12 (51 mg) Sephadex LH-20 Fr. 1-4 Fr. 5-7 Fr. 1-6 i Fr. 7-8 cyclohex-DCM-MeOH 16 7:4:1 RP-18 I H20-MeOH3:7-9:1 1 (7 mg) 18 24+25 (5 mg) (6.5mg)

Figure 5.6 Isolationscheme,partII,of the DCM extract of C. glaucophylla.Fractions with antibacterialac¬

tivityare double underlined 52 CLATHROTROPIS GLAUCOPHYLLA

5.4 Structures of the isolated compounds

In this study 27 compounds were isolated. From the alkaloid extracts 12 quinolizi¬ dine alkaloids (1-12) were obtained, while the DCM extract afforded betulinic acid

(13) and two derivatives of betulinic acid (14,15), two depsides (16,17), one a-pyrone

(18), one isocoumarin (19), six sterols (20-25) and two triterpènes (26, 27).

Table 5.4 Compounds isolated from Clathrotropis glaucophylla in this study

Comp. Name mg Isolated from

i (-)-clathrotropine 9 mg Alkaloid extract 2 (-)-anagyrine 39 mg Alkaloid extract 3 (-)-thermopsine 3 mg Alkaloid extract 4 (-)-baptifoline 3.5 mg Alkaloid extract

5 (-)-epibaptifoline 9 mg Alkaloid extract 6 (-)-rhombifoline 9 mg Alkaloid extract 7 (-)-tinctorine 1.2 mg Alkaloid extract 8 (-)-cytisine 7.5 mg Alkaloid extract 9 (-)-N-methylcytisine 15 mg Alkaloid extract 10 (-)-lupanine 9 mg Alkaloid extract 11 (-)-6a-hydroxylupanine 8 mg Alkaloid extract 12 (+)-5,6-dehydrolupanine 18 mg Alkaloid extract 13 betulinic acid 11mg DCM extract

' ' 14 23 -0-(4 -hydroxy-3 -methoxy-cinnamoyl)- DCM extract 2.3 mg betulinic acid

' ' ' 15 23 -0-(4 -hydroxy-3 ,5 -dimethoxy-cin- DCM extract 1.1mg namoyl)betulinic acid 16 2'-0-methylevernic acid 51mg DCM extract 17 confluentic acid 15 mg DCM extract 18 5(5),6(5)-6(2-hydroxy-1 -methylpropyl)- DCM extract 5 mg 3,5-dimethyl-5,6-dihydro-2H-a-pyrone 19 6-hydroxy-8-methoxy-3-«-pentylisocou- 4.8 mg DCM extract marin 20 ß-sitosterol DCM extract 45 mg 21 stigmasterol DCM extract 22 7ß-hydroxysitosterol DCM extract 5 mg 23 7 ß-hydroxystigmasterol DCM extract 24 ß-sitosterol-3-Oß-glucoside DCM extract 6.5 mg 25 stigmasterol-3-O-ß-glucoside DCM extract 26 ß-amyrin 20 mg DCM extract 27 glutinol 22 mg DCM extract 5 RESULTS AND DISCUSSION 53

19 HO^^CH3

= 18

9 CH3

Figure 5.7 Structures of the isolated compounds 1-12 54 CLATHROTROPIS GLAUCOPHYLLA

rCOOH

29 29

rCOOH

24' \ 23 CH2 23 CH2

9' 9'

•x 8X

,>\ ^s2

*OCH3 H3CO 4' OCH?

OH OH

14 15

Figure 5.8 Structures of the isolated compounds 13-15 5 RESULTS AND DISCUSSION 55

H3CO

H3C

5'CH3 HO' "CH3 9

Figure 5.9 Structures of the isolated compounds 16-19 56 CLATHROTROPIS GLAUCOPHYLLA

Figure 5.10 Structures of the isolated compounds 20-25 5 RESULTS AND DISCUSSION 57

Figure 5.11 Structures of the isolated compounds 26-27 58 CLATHROTROPIS GLAUCOPHYLLA

5.5 Structure elucidation

5.5.1 (-)-Anagyrine (2) and (-)-thermopsine (3)

Figure 5.12 Structures of (-)-anagyrine (2) and (-)-thermopsine (3)

The molecular mass of (-)-anagyrine (2) ([a]ff -133° (c = 0.1, EtOH)) was obtained from the EI-MS spectrum. In addition to the molecular peak at m/z 244 (36), it also showed other fragment ions typical for anagyrine-type alkaloids (m/z 98, 146 and 160) (Fig. 5.13) (Ohmiya et al., 1995). The 13C NMR spectrum (Fig. 5.14) revealed the signals for 15 carbon atoms. Of these, two could be distinguished by the HSQC experiment (Fig. 5.16) as quaternary carbon atoms, three as unsaturated CH-groups, three as aliphatic CH-groups and seven as aliphatic CH2-groups. The chemical shift (Ö

165.5) of the quaternary carbon atom (C-2) indicated a carbonyl group. Based on this information, the molecular formula of (-)-anagyrine (2) was established as C15H20N2O. 5 RESULTS AND DISCUSSION 59

The signals in the H NMR spectrum (Fig. 5.15) were well dispersed. However, they showed overlaps between H-12ß and H-14ß at 5 1.21 ppm, and H-8oc and H-14a at 5

1.72 ppm. All the other multiplets were clearly distinguished. The spectrum revealed the signals for three unsaturated proton doublet-doublets at 5 6.41 (./= 1.2, 9.0 Hz, H-3), 5

7.45 (J =1.1, 9.0 Hz, H-4), and 5 6.27 (J= 1.2, 7.1 Hz, H-5). In the COSY spectrum (Fig. 5.17) H-4 showed correlation with both H-5 and H-3, thus confirming its position. The chemical shifts of C-10 (550.1), C-l 1 (564.5), C-15 (555.3), and C-17 (553.7)

(see Table 5.5) indicated that these carbons are lying next to a nitrogen atom. The exact positions of the carbons and protons in the aliphatic part of the molecule were deter¬ mined by the 2D-experiments COSY (Fig. 5.17) and HMBC (Fig. 5.18), and these allowed the planar structure of the molecule to be established. The COSY spectrum revealed 2 spin systems, one for the unsaturated protons and one for the aliphatic ones.

The HMBC correlations between C-6 and H-4, H2-10 and H2-17, as well as between C-

2 and H2-10, assigned the connections of the two spin systems.

= The EI-MS spectrum for (-)-thermopsine (3) ([a]§ - 43° (c 0.1, EtOH)) showed a molecular peak at m/z 244 (44). Together with the information from the 13C NMR spec¬ trum, the molecular formula could be established as C15H20N2O, the same as for (-)- anagyrine (2). Although the *H NMR and 13C NMR spectra of (-)-anagyrine (2) and (-

)-thermopsine (3) were clearly different (Fig. 5.15 and Fig. 5.14), the same planar struc¬ tures were obtained for the two molecules, after consulting the 2D NMR spectra.

(-)-Anagyrine (2) and (-)-thermopsine (3) differ only in the sterochemistry of C-l 1.

Literature has presented evidence that the configuration of the electron pair of the ter¬ tiary amine nitrogen in anagyrine is ß, that the rings C and D have chair conformation and that the ring junction between them is eis. For thermopsine (3) the C/D ring junction is trans (Rycroft and Robins, 1991; Robins and Rycroft, 1992; Mikhova and Duddeck, 1998).

This was also confirmed by ROESY experiments (Fig. 5.19 and Fig. 5.20), which showed correlations between H-ll (52.97) and H-10oc (53.92) and H-10ß (54.01), for (-)-anagyrine (2), and between H-12o/ß (5 1.52) and H-10a (5 3.66) and H-10ß (5 4.25), for (-)-thermopsine (3) (The correlation of H-l 1 and H-12 with both H-10oc and 60 CLATHROTROPIS GLAUCOPHYLLA

H-10ß, is probably due to the strong mutual coupling of H-10a and H-10ß). Hence the configurations were determined to be 1R,9R,\\R and 1R,9R,\\S for (-)-anagyrine (2) and (-)-thermopsine (3), respectively (or their mirror images 1S,9S,\ lSand 1S,9S,\\R).

In this case, H-17ß and H-8a, as well as H-10ß and H-8ß, form a W configuration via C-l7, C-7, C-8 and via C-10, C-9, C-8, respectively. The COSY spectrum of (-)- anagyrine (2) showed weak signals for the correlations between H-17ß and H-8a and between H-10ß and H-8ß. The V-coupling-constant (1.8 Hz) in the doublet-doublet- doublet of H-17ß also referred to a long-range coupling. The 1,3-diaxial relationship between H-17a and H-8ß, and between H-10a and H-8a, for both (-)-anagyrine (2) and (-)-thermopsine (3), was proved by signals in the ROESY spectra. In (-)-anagyrine

(2) H-17a also showed NOE-correlation to H-12a, while in (-)-thermopsine (3) it showed correlation to H-ll (Fig. 5.19 and Fig. 5.20). These results agree with the assigned configurations.

r2.4«7 L2.2E7

12.117

L2.0B7 and/or so. 1.JI7

75: .1.917

70: Ll.717

<5i Ll.5«7

«oj .1.417

55: .1.317

50. .1.2*7 45. [M]+ 1.117 40. .9.4M

35. :«.»<

30. 7. IB«

25. J.«6

20. 4.71«

.3.5««

L2.4I6

ll.2M !||, 215.1 »H I . 172.1 "J"1201.1 :a.QKO iwJ Bt,, ,,^,.... ,|lfc...... Itl,... ..It.,,. I|.. viM,,,. .11., 150 ISO 170 100 MO 200 210 220 230 240 250 2(0 '27Ö' 2*0 M/I

Figure 5.13 EI-MS spectrum of (-)-anagyrine (2) 5 RESULTS AND DISCUSSION 61

Table 5.5 C NMR spectral data of (-)-anagyrine (2) and (-)-thermopsine (3)

Isolated 2a Isolated 2b Lit. values0 2 Isolated 3a Carbon 5 ppm 5 ppm 5 ppm 5 ppm

2(Q 165.5 163.5 163.6 165.8

3(CH) 116.6 116.5 116.6 116.6

4(CH) 141.3 138.6 138.7 141.4

5(CH) 107.8 104.5 104.5 107.9

6(C) 154.0 151.9 152.0 153.7

7(CH) 36.8 35.4 35.6 36.5

8 (CH2) 21.3 20.7 20.8 28.3

9(CH) 33.9 32.5 32.6 34.2

10(CH2) 53.1 51.4 51.5 46.3

11 (CH) 64.5 63.0 63.1 67.4

12(CH2) 23.4 22.5 22.7 31.0

13 (CH2) 26.6 25.5 25.6 25.4

14(CH2) 19.9 19.1 19.2 26.4

15(CH2) 55.3 54.3 54.4 57.4

17(CH2) 53.7 52.9 53.0 64.6

a 75.5 MHz, 295 K, in CD3OD,b 125.8 MHz, 295 K, in CDC13,c Asres et al., 1986 (62.5 MHz, CDC13) b) (-)-thermopsine4

17 15

i^fWf*M^ 111ri AiâàilMà

a)(-)-anagyrine CD3OD 17 13 8 10I 12 14

ii-tii.. »^1 frAoÊîîllrtt^ \jj|*iL)i. n II lullifa Ly| imLUM

" I I I I I I I I I I I I I I I I I I l'f 1"! I'lriTlTTTT l'I'l I I | 1 TT TT1 I I t ITTITTI I""!'"! rTTl "TT T1 I I I I TTT I1 I T I T I I r I I I I I t 1 1 I 't J I I I I I I I I I I r'I I I I I I I 1 I I I I I I 1 I I I I I I 1 I I 1 I I I I I I I I I I I I I I I I I I I I I 1 I I T I I I I I I I I" 1 1 160 150 140 130 120 110 100 90 80 70 60 50 40 30 ppm

Figure 5.14 13CNMR spectraof (-)-anagyrine(2)and (-)-thermopsine(3)(5ppm, 75.5 MHz, 295 K, CD3OD) 5 RESULTS AND DISCUSSION 63

Table 5.6 H NMR spectral data of (-)-anagyrine (2) and (-)-thermopsine (3)

2a 3a

5 ppm, 7 in Hz 5 ppm, 7 in Hz

3 6.41 (dd, 7=1.2,9.0) 6.42 (dd, 7= 1.2, 7.0)

4 7.45 (dd, 7= 7.1,9.0) 7.46 (dd, 7= 7.0, 9.0)

5 6.27 (dd, 7=1.2,7.1) 6.28 (dd, 7= 1.2, 9.0)

7 3.10 (brd, 7= 2.3) 3.08 (m)

8a 1.72 (m*) 2.01 (m*)

8ß 2.10 (brd, 7= 13.2) 1.92 (m*)

9 2.22 (m) 2.17 (m)

10a 3.92 (dd, .7=6.6, 15.5) 3.66 (dd, 7= 6.7, 15.9)

10ß 4.01 (d, 7 =15.5) 4.25 (d, 7= 15.9)

11 2.97 (brd, 7 =12.3) 2.10 (m)

12a 2.01 (dq, 7=4.0,12.9) 1.52 (m*)

12ß 1.21 (m*) 1.52 (m*)

13a 1.91 (m) 1.79 (m)

13ß 1.56 (tq, 7=3.9, 13.0) 1.34 (m*)

14a 1.72 (m*) 1.55 (m*)+

14ß 1.21 (m*) 1.46 (m*)+

15a 2.68 (m) 1.95 (m*)

15ß 2.79 (dt, 7= 3.0,13.6) 2.63 (brd, 7= 11.4)

17a 3.49 (dd, 7= 2.7,11.1) 2.42 (dd, 7= 2.6,11.3)

17ß 2.50 (ddd,7= 1.8,2.8,11.1) 2.83 (m, 7= 11.3)

* 500.1 MHz, 295 K, CD3OD, overlap,+ interchangeable H90 CD3OD b) (-)-thermopsine

l \ 10ß 35 10a

> - .... jJ- Ji*-*J

a)(-)-anagyrine H90 CD30D

10ß 10a 3 5 Ret 17a 7 }îg 15ß I7ß 8ß 14a 0 H 15a 12a 13ß l?ß 0 n ' 9 13(

JLjJ Jul ' « A_ jll i-^i -n. —1—1—1—1—1—1—1—1—1—1—1—1—i—r— ' ' ' ' ' ' ' ' ' 1 " 1 " 1 1 " 1 " " " " l " " l I " " " " I " r r , 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm

Figure 5.15 lU NMR spectraof a)(-)-anagyrine(2)and b) (-)-thermopsine(3)(5 ppm, 500.1 MHz, 295 K, CD3OD) 10ß 3 5 10a 15ß 8ß 17a 17ß 14a 14ß I1115a 9 lV3ß 12ß 11 l A_ I ppm JJLILLM— i i i ; r; —\ i i ii _i_ ' _i_ _'. . 4*- - L i*. . I + - ±- -i- x- -J» | 20 :i2- 13 9 7 - 40

""+- - 60 11

100

3 120

-140

t—i—|—i—i—i—i—i—i—i—i—i—i—i—r—i—i—i—i—i—i—i—i—i—i—i—i—i—r—i—i—i—I—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—i—r~ 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm

Figure 5.16 HSQC spectrum of (-)-anagyrine(2)(5ppm, 500.1 MHz, 295 K, CD3OD) 66 CLATHROTROPIS GLAUCOPHYLLA

UV^ lOß 10a 3 5 12a i 17a Li UlilMiALÂ ppm 20

h 40

60 11

h 80

100

-120

-140

6 160 2

180

• •

i i i i i i i i i i i i i 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm

Figure 5.18 HMBC spectrum of (-)-anagyrine(2)(5 ppm, 500.1 MHz, 295 K, CD3OD) 68 CLATHROTROPIS GLAUCOPHYLLA

O Oh p O Oh (N en m' I Oh

. . . I

»Sri **

Q O Q U

m ON CN

N p

O 4* O >n

(U C

o (L)

w o

"^ Öl «Q. o O 17a

ppm

12a/ß 1.5

-2.0

-2.5

-3.0

-3.5 *J&HH -4.0 O * H

-4.5

2.5 ppm

Figure 5.20 ROESY spectrum of (-)-thermopsine(3)(5 ppm, 500.1 MHz, 295 K, CD3OD) 70 CLATHROTROPIS GLAUCOPHYLLA

5.5.2 (-)-Baptifoline (4) and (-)-epibaptifoline (5)

(-)-baptifoline (4) C15H20N2O2 MW:260

(-)-epibaptifoline (5) C15H20N2O2 MW: 260

Figure 5.21 Structures of (-)-baptifoline (4) and (-)-epibaptifoline (5)

= (-)-Baptifoline (4), [a]§ -67° (c 0.1, EtOH), and (-)-epibaptifoline (5), [a]§3 -

84° (c = 0.1, EtOH), were identified by comparing the ID and 2D NMR spectra with those of the other a-pyridone alkaloids and with literature data (Greinwald et al., 1990;

Kennelly et al., 1999) (In the literature the assignments of C-l 1 and C-l3 have to be changed for compound 4, and the chemical shift of C-l 7 in compound 5 is assumed to be false, see Table 5.7). The 13C NMR (Fig. 5.22) and lB. NMR (Fig. 5.23) spectra were very similar to each other and, except for the C-l3, to those of (-)-anagyrine (2) (Fig. 5.14), for instance. The down field shift of C-l 3 in (-)-baptifoline (4) (565.7) and

(-)-epibaptifoline (5) (5 70.6) suggested the substitution of an hydroxyl group at this position. This was confirmed by the EI-MS spectra. For both substances an M+ peak appeared at m/z 260, consistent with their molecular formula C15H20N2O2. The two substances have the same planar structure and differ only in the sterochemistry of carbon 13. The stereochemical position of the 13-OH groups was decided from the C 5 RESULTS AND DISCUSSION 71 chemical shifts of C-l3 and from the ROESY experiments. The a-effects of equatorial hydroxyl groups are larger than those of the axial groups (Mikhova and Duddeck,

1998). This suggested that (-)-baptifoline (4) has an axial hydroxyl group (ß position), while in (-)-epibaptifoline (5) it is equatorial (a position). This was also confirmed by the ROESY experiments, which showed a cross peak between H-11 and H-13 in 5, but not in 4. Except for the hydroxyl group the substances are identical to (-)-anagyrine (2), also in the stereochemistry. Hence, the configurations were determined to be 7R,9R,llR,l3R and 1R,9R,UR,13S for (-)-baptifoline (4) and (-)-epibaptifoline (5), respectively (or their mirror images 1S,9S,1 \S,\3S and 7S,9S,l 15,13R).

Table 5.7 liC NMR spectral data of (-)-baptifoline (4) and (-)-epibaptifoline (5)

Lit. values Lit. values Isolated 4a Isolated 5a Isolated 5C Carbon 4b 5d 5 ppm 5 ppm 5 ppm 5 ppm 5 ppm

2(C) 165.7 163.6 165.5 163.5 163.7

3(CH) 117.0 116.5 116.6 116.7 116.8

4(CH) 141.0 138.9 141.3 138.7 138.8

5(CH) 108.0 105.0 107.8 104.6 104.7

6(C) 153.1 151.8 153.8 151.5 151.7

7(CH) 36.3 35.2 36.7 35.4 35.6

8 (CH2) 20.8 20.4 21.2 20.7 20.8

9(CH) 33.3 31.8 34.4 32.1 32.3

10(CH2) 52.9 51.5 53.1 51.5* 51.6*

11 (CH) 57.6 65.3 62.7 61.2 61.3

12 (CH2) 30.0 29.0 32.6 31.8 32.1

13 (CH) 65.7 55.8 70.6 70.0 70.1

14(CH2) 26.1 25.7 29.2 28.5 28.7

* 15 (CH2) 48.9 47.8 53.1 52.4* 52.7

* # 17(CH2) 52.8 52.1 53.4 52.1 62.2

a 75.5 MHz, 295 K, CD3OD,b Kennelly et al., 1999 (100.6 MHz, CDC13),c 125.8 MHz, * 295 K, CDCI3,d Greinwald et al., 1990 (100.6 MHz, CDC13), interchangeable b) (-)-epibaptifoline

3 5 9 14 8 13 11

2 6

wfr«<# »<>iiii^Miiti^»iAu^»^wwhii^tv%wi•»^w«i^'ww^^wMi***W**f >^

a)(-)-baptifoline

cd3od

>^^4vi##NfN»NW»^ MkiflkÉikiJkiiuJülJt«.m]ItÉkJj^

I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 1 1 1 1 1 1 1 1 1 r 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I I I 'i I i'" I .|MM...,.|.,.,.,,,,|.. 160 150 140 130 120 110 100 90 80 70 60 50 40 30 ppm

Figure 5.22 13CNMR spectraof a)(-)-baptifoline(4)and b) (-)-epibaptifoline(5)(5 ppm, 75.5 MHz, 295 K, CD3OD) 5 RESULTS AND DISCUSSION 73

Table 5.8 lH NMR spectral data of (-)-baptifoline (4) and (-)-epibaptifoline (5)

Ô ppm, J in Hz ô ppm, J in Hz

3 6.43 (dd, .7=1.2, 9.0) 6.41 (dd,J= 1.2,8.9)

4 7.47 (dd, 7=7.0, 9.0) 7.46 (dd, 7=7.0,8.9)

5 6.29 (dd, 7= 1.2, 7.0) 6.27 (dd, 7= 1.2, 7.0)

7 3.16 (bs) 3.10 (m)

8a 1.78 (brd, 7=13.4) 1.77 (m*)

8ß 2.12 (brd, J= 13.4) 2.08 (brd,/=13.2)

9 2.21 (m) 2.27 (m)

10a 3.96 (dd, 7= 6.7 15.5) 3.98 (dd,J= 6.6, 15.5)

10ß 4.05 (d,J= 15.5) 4.03 (d, 7=15.5)

11 3.52 (m) 2.97 (brd, 7= 12.4)

12a 2.27 (m) 1.84 (m*)

12ß 1.37 (m*) 1.54 (m*)

13 4.20 (m) (a) 3.66 (m) (ß)

14a 1.97 (m) 1.52 (m*)

14ß 1.33 (m*) 1.52 (m*)

15a 3.31 (m) 2.74 (m*)

15b 2.51 (m) 2.74 (m*)

17a 3.55 (m*) 3.33 (m)

17ß 2.60 (bd, 7 =10.9) 2.51 (m*)

* 500.1 MHz, 295 K, CD3OD, overlap H20 17a b) (-)-epibaptifoline

CD3OD 15a/b 14a/ß 3 5 10ß 10a 17ß 12a 12ß H 8a 9 8ß

JL _MM_ lluL^

a)(-)-baptifoline CD3OD H90

10ß 15a 10a 13 17a 12ß 11 12a08ß 14ß UaSa 17ßl5b JU JLJUL

' ' ' ' 1 ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' ' 1 l 1 1 1 1 1 1 1 1 1 1 1 1 1 7.5 7.0 6.5 6.0 5.5 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 ppm

Figure 5.23 lH NMR spectraof a)(-)-baptifoline(4)and b) (-)-epibaptifoline(5)(5ppm, 500.1 MHz, 295 K, CD3OD) 5 RESULTS AND DISCUSSION 75

5.5.3 (-)-Clathrotropine(l)

(-)-clathrotropine (1) C17H24N203 MW304

Figure 5.24 Structureof(-)-13a-hydroxy-15a-(l-hydroxyethyl)-anagyrine, (-)-clathrotropine (1)

(-)-Clathrotropine (1), [a]$ -146° (c = 0.1, EtOH), UVmax nm; MeOH (log e): 205

(2.67), 234 (2.79), 309 (2.87), was obtained as crystals. The molecular mass was estab¬ lished by HR-MALDI mass spectrometry, m/z 305.1845 [M+H]+ (Fig. 5.25). With this information and the carbon numbers from the 13C NMR spectrum (Fig. 5.26), the molecular formula was determined to be C^24^03.

1 ^ The C NMR spectrum (Fig. 5.26) showed the presence of 17 carbon atoms which could be assigned as shown in Table 5.9. The typical signals for 2 quartemary carbon atoms and 3 CH groups between 5100 and 5170 present in a-pyridone alkaloids could be observed for (-)-clathrotropine (1). In addition the DEPT 135 NMR experiment (Fig.

5.26) revealed the signals for 5 CH2 groups, of which two had chemical shifts indicating a position next to a nitrogen atom (C-10, 5 53.2 and C-l7, 5 47.4). It also showed the signals for 2 CH groups at 5 33.4 (C-9) and 5 36.5 (C-7), and 4 CH groups between 5

64.0 and ô 70.7. The chemical shifts of the latter indicated that these were bounded to either a nitrogen- or an oxygen atom.

Except for two more carbon atom signals and a down field shifted value for C-15 (5 76 CLATHROTROPIS GLAUCOPHYLLA

1 ^ 67.6), the C NMR spectrum of (-)-clathrotropine (1) was very similar to that of (-)-

epibaptifoline (5) (Fig. 5.22). This suggested a substitution of an a hydroxy at carbon

13 and a C2 group at carbon 15.

The signals in the H NMR spectrum (Fig. 5.27) were well dispersed showing no

overlap. At 5 0.99 a large doublet integrated for 3 protons appeared. The signal was

assigned to CH3-19, which showed correlation to the CH-18 at 5 67.5/5 3.57 in the

HSQC-TOCSY and COSY spectra. H-18 further correlated with H-15 in the COSY

spectrum (Fig. 5.28) and with C-l4 in the HMBC spectrum. This information con¬

firmed the substitution of a 1-hydroxyethyl group at carbon 15.

The configuration was determined to be 1R, 9R,\\R,\'$S,\5S (or the mirror image)

after the observation of cross peaks between H-11, H-13 and H-15, between H-11 and H-10a/ß, and between H-17a, H-8ß, H-12a and H-14a in the ROESY experiment

(Fig. 5.29). Hence the substance was identified as (-)-13a-hydroxy-15a-(l-hydroxyethyl)-anagyrine (1), which is new in nature, and was given the name (-)-clathrotropine.

+ 30S.2 100- [M+H]

90-

80-

70-

60-

50-

273.0 40-

30-

327.2 20- 277.2

10- 245.2 152.6 303.2 189.1 203^ 259.1 76.3 287.2 1330.3 362.0 101.7 136.5 | 163.6 I I 243.2 i ! ' ' II. I- - < J313.0 I i i i | i i i i I i i i i | I I I I I I I I I I I I I I I I I 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I ' I 100 150 200 250 300 350 Mass/Charge

Figure 5.25 MALDI spectrum of (-)-clathtrotropine (1) 5 RESULTS AND DISCUSSION 77

Table 5.9 13C and *H NMR spectral data of (-)-clathrotropine (1)

la lb Carbon Proton 5 ppm 5 ppm, J in Hz

2(C) 165.3

3(CH) 116.5 3 6.41 (dd,J= 1.2, 9.0)

4(CH) 141.3 4 7.47 (dd, .7=7.0, 9.0)

5(CH) 107.6 5 6.30 (dd, .7=1.2, 7.0)

6(C) 153.6

7(CH) 36.5 7 3.12 (bs)

8 (CH2) 8a 1.85 (brd, J= 13.2) 22.0 8ß 2.15 (brd, /= 13.2)

9(CH) 33.4 9 2.29 (m)

10 (CH2) 10a 4.13 (d,J= 15.4) 53.2 10ß 3.92 (dd, .7= 6.3,15.4)