©2011 John Peter Munafo, Jr ALL RIGHTS RESERVED

John Peter Munafo, Jr

NATURAL PRODUCTS CHEMISTRY OF LILIUM LONGIFLORUM:

STRUCTURAL ELUCIDATION, QUANTIFICATION, BIOLOGICAL ACTIVITY

AND FUNGAL METABOLISM OF STERODAL GLYCOSIDES

JOHN PETER MUNAFO JR

A Dissertation submitted to the

Graduate School-New Brunswick

Rutgers, The State University of New Jersey

in partial fulfillment of the requirements

for the degree of

Doctor of Philosophy

Graduate Program in Plant Biology

written under the direction of

Professor Thomas J. Gianfagna

and approved by

______

New Brunswick, New Jersey

May, 2011 ABSTRACT OF THE DISSERTATION

Natural Products Chemistry of Lilium longiflorum: Structural Elucidation, Quantification,

Biological Activity and Fungal Metabolism of Steroidal Glycosides

By JOHN PETER MUNAFO JR

Dissertation Director:

Professor Thomas J. Gianfagna

The Easter lily (Lilium longiflorum Thunb., Liliaceae) has beautiful white flowers and a delicate aroma and is appreciated worldwide as an attractive ornamental plant. In addition to its economic importance and popularity in horticulture, lily bulbs are regularly consumed in Asia, as both food and medicine. The Easter lily is a rich source of steroidal glycosides, a group of compounds that may be responsible for some of the traditional medicinal uses of lilies and may play a role in the pant-pathogen interaction. This research project was designed to: 1) Isolate and characterize new steroidal glycosides from the bulbs of L. longiflorum, 2) quantify their contents in all of the organs of L. longiflorum, and 3) perform studies on the antifungal activity and fungal metabolism of the compounds.

A phytochemical investigation conducted on the bulbs resulted in the discovery of several novel steroidal glycosides. A novel acetylated steroidal glycoalkaloid and two novel steroidal furostanol saponins, along with three other steroidal glycosides were isolated from the bulbs of L. longiflorum for the first time. A LC-MS/MS method performed in multiple reaction monitoring (MRM) mode was developed for the

ii simultaneous quantitative analysis of the five steroidal glycosides in the different organs of L. longiflorum. The highest concentrations of total steroidal glycosides were detected in flower buds, lower stems, and leaves. The steroidal glycoalkaloids were detected in higher concentrations as compared to the furostanol saponins in all of the plant organs except for the fibrous and fleshy roots. The proportions of steroidal glycoalkaloids to furostanol saponins were higher in the plant organs exposed to light and decreased in proportion from the aboveground organs to the underground organs. The highest concentrations of the steroidal glycoalkaloids were detected in flower buds, leaves, and bulbs.

Purified steroidal glycosides were evaluated for fungal growth inhibition activity against the plant pathogenic fungus, Botrytis cinerea. All of the compounds showed weak fungal growth inhibition activity; however, the natural acetylation of C-6′′′ of the terminal glucose in the acetylated steroidal glycoalkaloid, increased the antifungal activity by inhibiting the rate of metabolism of the compound by the fungus. A model system was developed to generate fungal metabolites of the steroidal glycoalkaloids and this system led to the discovery of several new fungal metabolites. The fungal metabolites characterized from the model system were subsequently identified by LC-MS and found to naturally occur in Easter lily tissues infected with the fungus.

iii

ACKNOWLEDGEMENTS

I would acknowledge my major advisor, Professor Thomas Gianfagna, for his guidance and helpful insight throughout this research. I would also like to thank my faculty committee members, Professor Richard Merritt and Professor Chee-Kok Chin and my outside member Dr. John Didzbalis. I would like to acknowledge my collaborators, colleagues and friends; Professor Leslie Jimenez, Professor Edward Durner, Dr. Ahalya

Ramanthan, Dr. Marshall Bergen, Dr. Christopher Johnson, Dr. Mark Kelm, Dr.

Catherine Kwik-Uribe, Thomas Collins, Jeanne Peters, Nimmi Rajmohan, Dr. Mahdu

Aneja, Bob Carhart, Jadwiga Leonczak, Professor Ilya Raskin, Dr. Slavko Komarnytsky, and Debroa Esposito. I would like to give a special thanks to my family and especially my wife, Kristin, for their constant encouragement and support. Most of all, I would like to thank God, for creating such a wonderful Universe for us to explore and ponder.

TABLE OF CONTENTS

ABSTRACT OF THE DISSERTATION ...... II

ACKNOWLEDGEMENTS ...... IV

TABLE OF CONTENTS ...... V

LIST OF TABLES ...... X

LIST OF FIGURES ...... XI

CHAPTER 1: GENERAL INTRODUCTION ...... 1

1.1. INTRODUCTION ...... 1

1.2. BOTANICAL CLASSIFICATION ...... 2

1.3. BOTANICAL DESCRIPTION ...... 2

1.4. NATURAL PRODUCTS FROM LILIACEAE ...... 5

1.5. SAPONINS IN GENERAL ...... 6

1.5.1. Steroidal Saponins ...... 9

1.5.1.1. Commercially Important Steroidal Saponins ...... 12

1.5.1.2. Dietary Sources of Steroidal Saponins ...... 16

1.5.1.3. Steroidal saponins isolated from Lilium ...... 18

1.6 STEROIDAL ALKALOIDS ...... 27

1.6.1 Steroidal Alkaloids in Liliaceae ...... 27

1.6.1.1. Classification of Isosteroidal Alkaloids of Liliaceae ...... 28

1.6.1.2. Classification of Steroidal Alkaloids of Liliaceae ...... 30

1.6.1.3. Steroidal Alkaloids in Lilium ...... 31

1.6.2. Steroidal Glycoalkaloids ...... 32

1.6.2.1. Dietary Sources of Steroidal Glycoalkaloids ...... 35

1.6.2.2. Steroidal Glycoalkaloids in Lilium ...... 37

1.7. PLANT ORGAN DISTRIBUTION OF STEROIDAL GLYCOSIDES ...... 40

1.8. STEROIDAL GLYCOSIDES IN PLANT DEFENSE ...... 42

1.9 DETOXIFICATION OF STEROIDAL GLYCOSIDES ...... 45

CHAPTER 2: ISOLATION AND STRUCTURAL DETERMINATION OF

STEROIDAL GLYCOSIDES FROM THE BULBS OF EASTER LILY (LILIUM

LONGIFLORUM THUNB.) ...... 48

2.1. ABSTRACT ...... 48

2.2. INTRODUCTION ...... 49

2.3. MATERIALS AND METHODS ...... 51

2.3.1. Plant Material...... 51

2.3.2. Chemicals...... 52

2.3.3. Isolation and Purification of Steroidal Glycosides 1 – 5 from L. longiflorum.53

2.3.3.1. Sequential Solvent Extraction of Lyophilized L. longiflorum Bulbs ...... 53

2.3.3.2. Gel Permeation Chromatography (GPC) ...... 54

2.3.3.3. Semipreparative Reverse-Phase High-Performance Liquid

Chromatography (RP-HPLC) ...... 56

2.3.4. Structural Elucidation ...... 59

2.3.4.1. Acid Hydrolysis of Compounds 1 – 5...... 61

2.3.4.2. Aglycone Analysis ...... 61

2.3.4.3. Sugar Composition Analysis ...... 62

2.3.4.4. Determination of Sugar Absolute Configurations ...... 62

2.3.4.5. Thin Layer Chromatography (TLC) ...... 63

2.4. RESULTS AND DISCUSSION ...... 64

2.4.1 Structure Elucidation of Compounds 1 – 5...... 64

2.3.4.1. Structure Elucidation of Compound 1 ...... 67

2.3.4.2. Structure Elucidation of Compound 2 ...... 77

2.3.4.3. Structure Elucidation of Compound 3 ...... 84

2.3.4.3. Structure Elucidation of Compound 4 ...... 93

2.3.4.4. Structure Elucidation of Compound 5 ...... 102

2.5. CONCLUSION ...... 109

CHAPTER 3: QUANTITATIVE ANALYSIS OF STEROIDAL GLYCOSIDES IN

DIFFERENT ORGANS OF EASTER LILY (LILIUM LONGIFLORUM THUNB.)

BY LC-MS/MS ...... 114

3.1. ABSTRACT ...... 114

3.3. MATERIALS AND METHODS ...... 120

3.3.1. Plant material...... 120

3.3.2. Chemicals ...... 123

3.3.3. Histology and Microscopy...... 123

3.3.4. Purification and Confirmation of Analytical Standards...... 124

3.3.4.1. Nuclear Magnetic Resonance Spectroscopy (NMR)...... 126

3.3.5. Quantitative Analysis of Steroidal Glycosides in Lilium longiflorum...... 126

3.3.5.1. Sample Preparation ...... 126

3.3.5.2. Analytical Standard Preparation ...... 127

3.3.5.3. Liquid Chromatography-Mass Spectrometry (LC-MS/MS)...... 127

3.3.5.4. Recovery ...... 132

vii

3.3.5.5. Thin Layer Chromatography (TLC)...... 132

3.3.5.6. Statistical Analysis...... 133

3.4. RESULTS AND DISCUSSION ...... 133

3.4.1. Quantification of steroidal glycosides in the different organs of L. longiflorum.

...... 133

3.4.2. Histological visualization of furostanol localization in bulb scale sections of L.

longiflorum...... 152

3.4.3. Quantification of steroidal glycosides within bulb scales of L. longiflorum. 156

3.5. CONCLUSION ...... 160

CHAPTER 4: ANTIFUNGAL ACTIVITY AND FUNGAL METABOLISM OF

STEROIDAL GLYCOSIDES OF EASTER LILY (LILIUM LONGIFLORUM) BY

THE PLANT PATHOGENIC FUNGUS, BOTRYTIS CINEREA ...... 163

4.1. ABSTRACT ...... 163

4.2. INTRODUCTION ...... 164

4.3. MATERIALS AND METHODS ...... 168

4.3.1. Plant material...... 168

4.3.2. Fungal cultures...... 169

4.3.3. Chemicals...... 169

4.3.4. Isolation and Purification of Steroidal Glycosides 1 – 5 from Lilium

longiflorum...... 170

4.3.4.1. Nuclear Magnetic Resonance Spectroscopy (NMR)...... 171

4.3.4.2. Liquid Chromatography-Mass Spectrometry (LC-MS)...... 171

4.3.4.3. Partial acid hydrolysis of compound 1...... 172

viii

4.3.5. B. cinerea growth inhibition assay...... 173

4.3.6. In vitro fungal metabolism of compounds 1 and 2...... 174

4.3.7. Scale-up fungal metabolism of compound 1...... 175

4.3.7.1. Semi-preparative RP-HPLC isolation of the fungal metabolites of

compound 1...... 180

4.3.8. Isolation and Purification of Compound 6 from Lilium longiflorum bulbs. . 183

4.3.8.1. Sequential Solvent Extraction of Lyophilized L. longiflorum Bulbs...... 183

4.3.8.2. Gel Permeation Chromatography (GPC)...... 184

4.3.9. Infection of L. Longiflorum tissue and sample preparation for LC-MS analysis.

...... 188

4.3.10. Statistical Analysis...... 188

4.4. RESULTS AND DISCUSSION ...... 189

4.4.1. Fungal growth inhibition assay...... 189

4.4.2. Metabolism of compound 1 and 2 by B. cinerea...... 193

4.4.3. In planta identification of compounds 6 – 10 by LC-MS...... 208

4.4.4. Isolation and identification of compound 6 from L. Longiflorum bulbs...... 214

4.5. CONCLUSION ...... 215

SUMMARY AND CONCLUDING REMARKS ...... 216

LITERATURE CITED ...... 220

CURRICULUM VITA………………………………………………………………..238

LIST OF TABLES

Table 1.1. Steroidal saponins found in Lilium…………………………………………..19

13 Table 2.1. C NMR spectral data of compounds 1 – 5 in pryridine-d5…………….…..113

Table 3.1. ANOVA for concentrations of compounds 1 – 5 in the different organs of L. longiflorum………………………………………………………………..140

Table 3.2. Concentrations of compounds 1 – 5 in the different organs of L. longiflorum………………………………………………………………..141

Table 3.3. Concentrations of compounds 1 – 5 in whole bulb scale, bulb epidermis, and bulb mesophyll……………………………………………….…...157

Table 4.1. ANOVA for treatment, rate, and the interaction between treatment and rate…………………………………………………………………….…189

LIST OF FIGURES

Figure 1.1. Image of L. longiflorum in full bloom…………………………………….4

Figure 1.2. Structures of 2,3-oxidosqualene, a triterpenoid sapogenin

(quillaic acid), and a steroidal sapogenin (diosgenin)…………………………….7

Figure 1.3. Examples of monodesmosidic and didesmosidic saponins………...... 8

Figure 1.4. Structures of a basic steroidal backbone, a spirostane

backbone, and a furostane backbone………………………………………...…..10

Figure 1.5. Examples of different carbohydrate linkages……………………………11

Figure 1.6. Structures of Dioscin and Protodioscin………………………………….14

Figure 1.7. Molecular structures of steroidal saponins from Lilium……...... 20

Figure 1.8. Isosteroidal alkaloids of Liliaceae: Representative

examples of cevanine type, veratramine type, and jervine type

isosteroidal alkaloids……………………………………………………………..29

Figure 1.9. Steroidal alkaloids of Liliaceae: Representative examples

of solanidine type and verazine type steroidal alkaloids………………………...30

Figure 1.10. Structures of the steroidal alkaloids etioline and

teiemine isolated from the bulbs of L. candidum………………………………...32

Figure 1.11. Structures of the most common aglycones of the

steroidal glycoalkaloids………………………………………………………….34

Figure 1.12. Structures of steroidal glycoalkaloids isolated from

C. cordatum, L. philippinense, L. brownii, and L. mackliniae…………..……….39

Figure 2.1. Isolation scheme for compounds 1 – 5 purified

from the bulbs of L. longiflorum…………………………………………………54

Figure 2.2. RP-HPLC chromatogram of 1 – 5 isolated from

L. longiflorum……………………………………………………………………57

Figure 2.3. Total ion chromatogram of L. longiflorum extract

and of compounds 1 – 5 isolated by RP-HPLC……………………………….…58

Figure 2.4. Structures of compounds 1 – 5 isolated from

L. longiflorum bulbs……………………………………………………………...60

Figure 2.5. High resolution mass spectrum of compound 1…………………………68

Figure 2.6. GC-MS chromatogram of the TMSi derivatives of

(22R, 25R)-spirosol-5-en-3 -ol.…………………………………………………69

Figure 2.7. GCMS mass spectra of TMSi derivatives of

(22R, 25R)-spirosol-5-en-3 -l……………………………………………...……70

Figure 2.8. GCMS mass spectra of TMSi derivatives of the

aglycone of compound 1…………………………………………………………71

Figure 2.9. Total ion chromatogram of (22R, 25R)-spirosol-5-en-3 -ol

and the aglycone of compound 1 generated by LC-MS………………………....72

Figure 2.10. HMBC long-range correlations for the interglycosidic

linkages for the carbohydrate moiety of compound 1……………………….…..73

Figure 2.11. ESI+–MS mass spectrum of compound 1……………………………...74

Figure 2.12. 1H NMR spectrum and 13C NMR spectrum of compound 1…………..75

Figure 2.13. High resolution mass spectrum of compound 2……………………...... 78

xii

Figure 2.14. HMBC long-range correlations for the interglycosidic

linkages for the carbohydrate moiety of compound 2…………………………...79

Figure 2.15. ESI+–MS mass spectrum of compound 2………………………………80

Figure 2.16. 1H NMR spectrum and 13C NMR spectrum of compound 2…………...82

Figure 2.17. High resolution mass spectrum of compound 3………………………..85

Figure 2.18. LRMS- mass spectrum of compound 3…………………………...... 86

Figure 2.19. GCMS spectra of TMSi derivatives of the aglycone of

compound 3 and (25R)-spirost-5-en-3 -ol………………………………………88

Figure 2.20. HMBC long-range correlations for the interglycosidic

linkages for the carbohydrate moiety of compound 3……………………….….89

Figure 2.21. ESI+–MS mass spectrum of compound 3…………………………...….90

Figure 2.22. 1H NMR spectrum and 13C NMR spectrum of compound 3………...…91

Figure 2.23. High resolution mass spectrum of compound 4………………………..94

Figure 2.24. LRMS- mass spectrum of compound 4………………………..……….95

Figure 2.25. Partial HMBC spectrum of compound 4……………………………….96

Figure 2.26. HMBC long-range correlations for the interglycosidic

linkages for the carbohydrate moiety of compound 4……………………..……98

Figure 2.27. ESI+–MS mass spectrum of compound 4………………………….…..99

Figure 2.28. 1H NMR spectrum and 13C NMR spectrum of compound 4………….100

Figure 2.29. High resolution mass spectrum of compound 5………………………103

Figure 2.30. LRMS- mass spectrum of compound 5……………………………….104

Figure 2.31. HMBC long-range correlations for the interglycosidic

linkages for the carbohydrate moiety of compound 5…………………………105

xiii

Figure 2.32. ESI+–MS mass spectrum of compound 5………………………..……106

Figure 2.33. 1H NMR spectrum and 13C NMR spectrum of compound 5………….107

Figure 2.34. ESI+–MS mass spectra of compounds 1 – 5………………………..…112

Figure 3.1. Plant organs of L. longiflorum analyzed in this study…………….……122

Figure 3.2. Structures of steroidal glycoalkaloids 1 – 2 and furostanol

saponins 3 – 5 quantified in the various L. longiflorum organs…………….…...126

Figure 3.3. MS2 product ion spectra of steroidal glycoalkaloids 1 – 2………..……129

Figure 3.4. MS2 product ion spectra of furostanol saponins 3 – 5………………….130

Figure 3.5. MS/MS chromatograms for the quantitative analysis

of compounds 1 – 5 in a L. longiflorum bulb scale…………………………..…131

Figure 3.6. Calibration equation for compound 1………………………………..…134

Figure 3.7. Calibration equation for compound 2……………………………….….135

Figure 3.8. Calibration equation for compound 3……………………………….….136

Figure 3.9. Calibration equation for compound 4……………………………….….137

Figure 3.10. Calibration equation for compound 5…………………………………138

Figure 3.11. Proportions of steroidal glycoalkaloids 1 – 2 to furostanol

saponins 3 – 5 in the different organs of L. longiflorum…………………….….142

Figure 3.12. Concentrations of steroidal glycoalkaloid 1 in the

different organs of L. longiflorum………………………………………………143

Figure 3.13. Concentrations of steroidal glycoalkaloid 2 in the

different organs of L. longiflorum………………………………………………144

Figure 3.14. Concentrations of compound 3 in the different

organs of L. longiflorum………………………………………………………..147

xiv

Figure 3.15. Differences in saccharide composition and

interglycosidic linkages of compounds 3 – 5………………………………..…148

Figure 3.16. Concentrations of compound 4 in the different organs of L.

longiflorum……………………………………………………………………..150

Figure 3.17. Concentrations of compound 5 in the different

organs of L. longiflorum………………………………………………………..151

Figure 3.18. Histochemical staining of a bulb scale section…………………….…153

Figure 3.19. Histochemical analysis of bulb basal plate and bulb scale sections….154

Figure 3.20. Histochemical analysis of bulb basal plate and bulb scale sections….155

Figure 3.21. Proportions of compounds 1 – 5 in whole bulb scale,

bulb epidermis, and bulb mesophyll……………………………………………158

Figure 3.22. Proportions of compounds 1 – 5 in different organs

of L. longiflorum……………………………………………………………..…159

Figure 4.1. Total ion chromatogram of the partial acid-catalyzed

hydrolysis products of compound 1………………………………………….…173

Figure 4.2. Extracted ion chromatograms of m/z 885 taken every 24 hours

over the course of 96 hours………………………………………………….…176

Figure 4.3. Extracted ion chromatograms of m/z 723 taken every 24 hours over the course of 96 hours…………………………………………….…….….…177

Figure 4.4. Extracted ion chromatograms of m/z 577 taken every 24 hours

over the course of 96 hours……………………………………………….……178

Figure 4.5. Extracted ion chromatograms of m/z 414.6 taken every 24 hours

over the course of 96 hours……………………………………………….……179

Figure 4.6. RP-HPLC chromatogram of compounds 6, 7, and 10…………….……181

Figure 4.7. Total ion chromatograms of compounds 6, 7, and 10 isolated

by RP-HPLC……………………………………………………………………182

Figure 4.8. Isolation scheme for compound 6 purified from the bulbs of L.

longiflorum...... 185

Figure 4.9. RP-HPLC chromatogram of compound 6 isolated from

L. longiflorum bulbs………………………………………………………...... 186

Figure 4.10. Structures of compounds 1 – 10...... 187

Figure 4.11. Growth inhibition activity of compounds 1 – 5 on the radial

mycelia growth of B. cinerea…………………………………………...………190

Figure 4.12. Growth inhibition activity of compounds 1 and 2 on the

radial mycelia growth of B. cinerea……………………………………….……192

Figure 4.13. ESI+–MS mass spectra of steroidal glycoalkaloids 1 and 2……..……193

Figure 4.14. Metabolism of compound 1 by B. cinerea……………………………194

Figure 4.15. ESI+–MS mass spectra of fungal metabolite 6…………………..……195

Figure 4.16. Molecular structure and fragmentation of compound 6………………196

Figure 4.17. ESI+–MS mass spectra of fungal metabolite 7………………………..196

Figure 4.18. Molecular structure and fragmentation of compound 7………………197

Figure 4.19. ESI+–MS mass spectra of fungal metabolite 10………………………198

Figure 4.20. Molecular structure and fragmentation of compound 10……….…….199

Figure 4.21. Metabolism of compound 2 by B. cinerea……………………………200

Figure 4.22. ESI+–MS mass spectra of fungal metabolite 8…………………….….201

Figure 4.23. Proposed molecular structure and fragmentation of compound 8….…202

xvi

Figure 4.24. ESI+–MS mass spectra of fungal metabolite 9…………………….….203

Figure 4.25. TIC of the metabolites of compound 2, extracted ion

chromatogram of m/z 738.8, and the TIC of the partial acid-catalyzed

hydrolysis products of compound 1………………………………………….…204

Figure 4.26. Proposed molecular structure and fragmentation of compound 9…….205

Figure 4.27. Proposed partial metabolic pathways for compounds 1 and 2………..207

Figure 4.28. Extracted ion chromatograms (EIC) for compound 8

(m/z 780.5) of control plant tissue and plant tissue infected with B. cinerea…..209

Figure 4.29. Extracted ion chromatograms (EIC) for compound 9

(m/z 738.8) of control plant tissue and plant tissue infected with B. cinerea…..210

Figure 4.30. Extracted ion chromatograms (EIC) for compound 7

(m/z 576.7) of control plant tissue and plant tissue infected with B. cinerea…..211

Figure 4.31. Extracted ion chromatograms (EIC) for compound 10

(m/z 414.6) of control plant tissue and plant tissue infected with B. cinerea…..212

Figure 4.32. Extracted ion chromatograms (EIC) for compound 6

(m/z 722.8) of control plant tissue and plant tissue infected with B. cinerea…..213

Figure 4.33. Total ion chromatogram (TIC) of compound 6 isolated by RP-HPLC from L. longiflorum bulbs……………………………..……214

xvii

Chapter 1: General Introduction

1.1. Introduction

The Easter Lily, (Lilium longiflorum Thunb., family Liliaceae), has beautiful white flowers and a delicate aroma and is appreciated worldwide as an attractive ornamental plant. Easter lilies are most commonly seen as indoor potted plants or floral arrangements around the Easter holidays; however, they are also often planted outdoors as bedding plants in flower gardens. In addition to their esthetic value, lily bulbs and flower buds are regularly consumed as a food in Asia for their distinctive bitter taste and have long historical use in traditional Chinese medicine.

L. longiflorum is native to the Ryuckyu archipelago of Japan and the islands of eastern Taiwan (Wilson, 1925; Hiramatsu et al., 2001). The archipelago located between

Ryukyu and Taiwan consists of approximately 200 islands located between southwestern mainland Japan and the southeastern China. Geological studies suggest that the archipelago was formed from a section of the Asian continent and it has been proposed that the biota native to the islands evolved from common ancestors located on the adjacent mainland (Kizaki and Oshiro, 1977; Kimura, 1996). Consistent with the hypothesis, L. longiflorum is genetically similar to other Lilium species in the geographic region, such as L. formosanum, and is readily capable of producing fertile interspecific hybrids with other members of the genus (Hiramatsu, et al., 2001; Preston, 1933). In addition, cytological analysis of forty-six species revealed very little variation in the

2 karyotypes within the genus; thus, interspecific hybridization is widely employed for the development of new Lilium cultivars (Stewart, 1947).

1.2. Botanical Classification

L. longiflorum is classified under the family Liliaceae in the order Liliales.

According to the Germplasm Resources Information Network (GRIN), the Liliaceae family contains 16 accepted genera and the genus Lilium contains 110 accepted species

(GRIN, 2011). The most common vernacular names for L. longiflorum include Easter lily, Bermuda Easter lily, Bermuda lily, trumpet lily, and white trumpet lily.

1.3. Botanical Description

The following botanical description for L. longiflorum is based on wild specimens endemic to Taiwan. It is important to note that commercial varieties of L. longiflorum have undergone extensive selection and hybridization and as a result the commonly available cultivars are variable in morphological characteristics. Nevertheless, wild specimens are the original breeding stock of today‘s modern cultivars. The bulbs of L. longiflorum are pale yellow to white in color, globose to obovoid in shape and approximately 4.0 – 6.0 cm in diameter. Bulb scales are imbricate in morphology. The stem is green in color, erect to ascending in habit, and approximately 45.0 – 90.0 cm long. The surface of the stem is scabrid and pubescent. The plant has entire alternate leaves and are lanceolate to falcate-lanceolate in morphology. Typical leaves are

3 approximately 15.0 – 25.0 cm long and 1.0 – 2.5 cm wide. The leaf apex is acute and the base is amplexicual. The leaves are chartaceous and glabrous on both surfaces. Leaves have three veins that are slightly impressed on upper surface and are elevated on the lower surface. Easter lily flowers are terminal and are pure white in appearance. Flowers range from solitary to several, have a fragrant aroma, and are typically 12.0 – 17.0 cm long and 5.0 – 7.0 cm in diameter. The flower habit is horizontal or nodding. Bracts are lanceolate and are approximately 3.0 – 5.0 cm long with an acute apex. The pedicel is green in appearance, 2.5 – 6.0 cm in length, and glabrous. Six tepals occur in two series.

The tepals are white in appearance and are slightly tinged green toward base abaxially.

The outer tepals are oblanceolate to broadly oblanceolate and 2.5 – 3.0 cm wide. The inner tepals are somewhat broader, ranging from 3.5 – 4.5 cm in width, and have an obtuse apex. Filaments are 7.0 – 9.0 cm long and glabrous. Anthers are cylindric and range from 2.0 – 3.0 cm in length. The ovaries are green in appearance, elongate, cylindric, and 2.5 – 4.0 cm in length. The ovaries are glabrous. The style is 7.0 – 9.0 cm long and pale yellow in appearance. The stigma is dark green and trifid. The fruit is a loculicidal capsule that is cylindric and 3.0 – 5.0 cm in length and 2.5 – 3.5 cm in width

(Ying, 2000; Wu and Raven, 2000).

Figure 1.1. Image of L. longiflorum in full bloom.

1.4. Natural products from Liliaceae

The Liliaceae family is a rich source of natural products displaying a vast range of structural diversity. A multitude of natural products have been isolated and characterized from Liliaceae including, dimeric ent-kaurane diterpenes (Kitajima et al., 1982; Wu et al.,

1995), flavonoid glycosides (Fattorusso et al., 2002; Francis et al., 2004), anthocyanins

(Takeda et al., 1986; Nørbæk and Kondo, 1999), stilbenes (Zhou et al., 1999), phenolics

(Tai et al. 1981), phenolic glucosides (Shoyama et al., 1987), phenolic amides (Park,

2009) carotenoids (Tsukida et al., 1965), sterols (Itoh, et al., 1977), alkaloids

(Shimomura, H, 1987), and sulfur-containing compounds (Lanzotti, 2006). Most notably, there has been extensive work done on the isolation and characterization of steroidal glycosides including steroidal saponins (Harmatha et al., 1987; Matsuura et al.,

1989; Yang, et al., 2004) and steroidal glycoalkaloids (Mimaki and Sashida, 1990;

Sashida et al., 1990) from within the Liliaceae family.

Steroidal glycosides have been reported to exhibit a wide range of biological activities including antifungal (Sautour et al., 2005; Zhou et al., 2003), platelet aggregation inhibition (Zhang et al., 1999; Huang et al., 2006), anti-cholinergic (Gilani et al., 1997) anti-diabetic (Nakashima et al., 1993), anti-hypertensive (Oh et al., 2003), cholesterol lowering (Matsuura et al., 2001), anti-inflammatory (Shao et al., 2007), antiviral (Gosse et al., 2002), and anticancer (Acharya et al., 2009; Pettit et al., 2005;

Mimaki et al., 1999; Jiang et al., 2005). Additionally, steroidal glycosides have a wide variety of commercial uses including surfactants (Yamanaka et al., 2008), foaming agents

(Singh et al., 2003), vaccine adjuvants (Rajput et al., 2007), and serve as precursors for the industrial production of pharmaceutical steroids (Hansen, 2007).

Steroidal saponins have been found in over 100 plant families and in some marine organisms such as starfish and sea cucumber (Güçlü-Üstündağ and Mazza, 2006). They are characterized by a steroid type skeleton glycosidically linked to carbohydrate moieties. Steroidal glycoalkaloids are characterized by a nitrogen containing steroid type skeleton glycosidically linked to carbohydrate moieties. In contrast to steroidal saponins, the occurrences of steroidal glycoalkaloids are, thus far, limited to the members of the plant families Solanaceae and Liliaceae (Li et al., 2006; Ghisalberti, 2006).

1.5. Saponins in general

Saponins are a structurally diverse class of natural products that are characterized by a non-polar sapogenin moiety glycosidically linked to one or more polar carbohydrate moieties. Based on the composition of sapogenin skeleton, they are generally classified into two major categories, triterpenoid saponins and steroidal saponins (Abe et al., 1993).

Triterpenoid saponins contain a thirty carbon aglycone and steroidal saponins contain a twenty-seven carbon aglycone. Both classes are derived from the thirty carbon precursor

2,3-oxidosqualene (Haralampidis et al., 2000). Isopentenyl pyrophosphate synthesized via the mevalonate pathway is the five carbon donor for the biosynthesis of terpenes in plants. Triterpenoid and steroidal sapogenins are synthesized from the thirty carbon hydrocarbon squalene, which is subsequently oxidized to squalene 2,3-epoxide, and then

7 converted to tetra- or pentacyclic triterpenes by a family of 2,3-oxidosqualene cyclases.

Following cyclization, the sapogenin moiety is subsequently mono- or

Figure 1.2. Structures of (A) 2,3-oxidosqualene, (B) a triterpenoid sapogenin (quillaic acid), and (C) a steroidal sapogenin (diosgenin) (Haralampidis et al., 2000; Kuljanabhagavad, et al. 2008; Espejo, et al., 1982).

29 30

20 19 21

12 18 22

11 13 17 25 COOH 9 14 16 1 15 2 10 8 OH

3 5 7 27 6 RO 4 B

24 CHO

21 25 O 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 2 10 8

3 5 7 6 C RO 4

poly- glycosylated by a wide variety of glycosyltransferase enzymes. In addition to the structure of the sapogenin, saponins are also classified according to the number of carbohydrate moieties that are glycosidically linked to the aglycone. Accordingly, they

8 are referred to as mono-, di-, or tridesmosidic, based on the number of carbohydrate moieties linked to the sapogenin skeleton.

Figure 1.3. Examples of (A) monodesmosidic and (B) didesmosidic saponins isolated from Quillaja Saponaria and Chenopodium quinoa, respectively (Kuljanabhagavad, et al. 2008; Guo, et al., 1998).

COOH OH HOOC O HO O A HO CHO OH O HO O HO OH

COO HO OH OH OH OH OH O O O OH HO O HO O OH CHO OH B

1.5.1. Steroidal Saponins

Steroidal saponins are widely found throughout the plant kingdom and have been reported in a broad range of orders including Solanales (Ferro, et al., 2005), Ranunculales

(Braca, et al., 2004), Sapindales (Achenbach, et al., 1994), Fabales (Murakami et al.,

2000), Cyperales (Osbourn, et al., 2000), Liliales (Debella, et al., 1999), Dioscoreales

(Haraguchi et al., 1994), Aspargales (Sautour et al., 2007), and Zingiberales (Lin, et al.,

1996). In addition, steroidal saponins have been documented in over 100 plant families and in many marine organisms (Güçlü-Üstündağ and Mazza, 2006). Steroidal saponins are divided into two main classes, spirostanols and furostanols, based on structural differences in the aglycone. Spirostanols have a six-ring structure (A – F rings), referred to as a spirostane skeleton, and are monodesmosidic, typically having a carbohydrate moiety -glycosodically attached by an ether linkage to the C-3 carbon of the aglycone.

Furostanols have a pentacyclic aglycone (A – E rings), referred to as a furostane skeleton, and are bidesmosidic with one carbohydrate moiety attached through an ether linkage at

C-3 carbon and a second carbohydrate moiety attached by an ether linkage at the C-26 carbon. The most common furostanols have a single glucose linked at the C-26 position; however, multiple sugars can be attached, but this is less common.

Figure 1.4. Structures of (A) a basic steroidal backbone (A-F rings), (B) a spirostane backbone, and (C) a furostane backbone.

C D A

A B

21 25 O 26 18 23 24 27 20 22 12 17 19 11 13 16 O B 9 14 15 1 2 10 8

3 5 6 7 RO 4

27 21 OH 18 20 23 25 OR2 22 24 26 12 17 19 11 13 16 O 9 14 15 1 2 10 8

3 5 7 C 6 RO 4

Figure 1.5. Examples of different carbohydrate linkages: (A) linear arrangement ( -D- glu-(1→4)- -D-glu), (B) branched arrangement ( -L-rha-(1→2)- -L-xyl-(1→3)- -D- glu), and (C) branched arrangement ( -L-rha-(1→2)- -D-glu-(1→4)- -D-glu).

OH OH O 4 HO 4' O R -D-Glc -D-Glc A HO O O OH HO OH

OH -D-Xly O O R HO HO O 3 HO O 2' -D-Glc B OH 3' O 2 1'' -L-Rha H3C O HO OH OH

OH OH O HO 4' O R -D-Glc HO O O 4 OH HO 2' -D-Glc C O 1'' -L-Rha 2 H3C O HO OH OH

The sugar composition of the carbohydrate moiety of steroidal saponins most commonly include, D-glucose, D-galactose, D-glucuronic acid, D-galacturonic acid, L-rhamnose,

L-arabinose, D-xylose, and D-fucose. All dextrorotatory form sugars are linked via a - glycosidic linkage and all levorotatory form sugars are linked via an -glycosidic linkage. The composition of the carbohydrate moiety can range from one sugar to multiple sugars and can be linked in a linear or branched arrangement. In addition, sugars can be attached by different interglycosidic linkages resulting in a vast number of

12 possible structural arrangements. Structural differences in the carbohydrate moiety have been shown to play a role in the biological activity of the molecules and differential biological activity of steroidal saponins containing the same aglycone but differing only in carbohydrate composition has been previously reported (Mimaki et al., 2001).

1.5.1.1. Commercially Important Steroidal Saponins

Steroidal saponins have been reported to exhibit a wide range of biological activities including antifungal (Sautour et al., 2005), platelet aggregation inhibition

(Zhang et al., 1999; Huang et al., 2006), anti-diabetic (Nakashima et al., 1993), cholesterol lowering (Matsuura, 2001), anti-inflammatory (Shao et al., 2007), antiviral

(Gosse et al., 2002) and anticancer (Acharya et al., 2009; Pettit et al., 2005; Mimaki et al.,

1999). Additionally, steroidal saponins have wide a variety of commercial uses including surfactants (Yamanaka et al., 2008), foaming agents (Singh et al., 2003), vaccine adjuvants (Rajput et al., 2007), and serve as precursors for the industrial production of pharmaceutical steroids (Hansen, 2007).

The pioneering work published by Russell Marker in the 1940s resulted in the semi-synthetic preparation of progesterone from the sapogenin, diosgenin, obtained by the hydrolysis of steroidal saponins extracted from the Japanese yam, Dioscorea tokoro

(Marker, et al., 1940). The Dioscoreaceae family is a rich source of steroidal saponins, and through further investigations with the goal of discovering an even richer source of precursors for steroid synthesis, other species such as D. Mexicana, D. composita, and D.

13 floribunda were identified. Diosgenin and similar steroid-based sapogenins are still an important intermediate for the industrial preparation of pharmaceutical steroids including anti-inflammatory, androgenic, estrogenic, and contraceptive drugs. In addition to a source of pharmaceutical precursors, steroidal saponin rich tubers and roots from members of the Dioscorea genus are commonly consumed as both food and medicine in much of Africa, Asia, and Tropical America (Sautour, et al. 2007).

Steroidal saponins have been isolated and characterized from many Dioscorea species including D. bulbifera var. sativa (Teponno, et al., 2006), D. cayenensis

(Sautour, et al. 2004), D. collettii var. hypoglauca (Hu, et al., 2003), D. composita

(Espejo, et al., 1982), D. deltoidea var. orbiculata (Shen, et al., 2002), D. futschauensis

(Liu, et al., 2003), D. nipponica (Cui, et al., 2004), D. olfersiana (Haraguchi, et al.,

1994), D. panthaica (Dong, et al., 2004), D. parviflora (Yang, et al., 2005), D. polygonoides (Osorio, et al. 2003), D. prazeri (Wij, et al., 1977), D. pseudojaponica

(Yang, et al., 2003), D. spongiosa (Yin, et al., 2003), D. villosa (Sautour, et al., 2006), D. zingiberensis (Sun, et al., 2003). The high levels of steroidal saponins in these species may contribute to the reported medicinal properties of these plants. The tubers and roots several Dioscorea species including D. colletii var. hypoglauca, D. panthaica , D. nipponica, and D. futschauensis have been traditionally used in China for various medicinal uses including anticancer, cardiovascular, rheumatism, and a general tonic

(Lacaille-Dubois, 2002; Li and Zhou, 1994; Li, et al., 2000). In fact, a crude drug used in traditional Chinese medicine that is prepared with D. panthaica is regularly used for the prevention and treatment of cardiovascular diseases in China today (Li and Zhou, 1994).

In addition, another crude drug made from a mixture of steroidal saponins extracted from

D. nipponica is used to treat rheumatism (Li, et al., 2000).

Steroidal saponins isolated from Dioscorea have shown various biological activities including cytotoxic activity, immunomodulating activity, antimicrobial activity, hormonal activity (Lacaille-Dubois, 2002), anti-osteoporotic activity (Yin et al., 2003), anti-inflammatory activity (Tewtrakul et al., 2007), and anti-allergic activity (Tewtrakul et al., 2006).

Figure 1.6. Structures of Dioscin and Protodioscin isolated from D. collettii var. hypoglauca (Hu, et al., 2003)

21 25 O 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 H3C OH 2 10 8 O 4' 3 5 7 HO 1''' O 6 HO O O 4 OH 2' HO 1' O 1'' Dioscin H3C O HO OH OH

OH O HO O HO 1'''' OH 21 OH 26 25 18 23 24 27 20 22 12 17 19 11 13 16 O 9 14 15 H3C 1 O OH 2 10 8 HO 1''' 4' 3 5 7 HO O O 4 6 OH O HO 2' O 1' 1'' H3C O HO OHOH Protodioscin

In addition to the commercial use of steroidal saponins as precursors for the synthesis of pharmaceutical steroids, steroidal saponins serve as an important raw material in the food, pharmaceutical, cosmetic and agricultural industries. The most common commercial applications of steroidal saponins include surfactants (Yamanaka et al., 2008), foaming agents (Singh et al., 2003), vaccine adjuvants (Rajput et al., 2007), and feed additives (Anthony et al., 1994; Balog et al., 1994).

Yucca schidigera, a desert plant from the Agavaceae family, is one of the most important commercial sources of steroidal saponins. Y. schidigera, commonly referred to as yucca, is native to the southwestern United States and Mexico where it has a long historical use as a medicine to treat ailments including inflammation, headaches, gonorrhea, and arthritis (Cheeke, 1998). The primary commercial raw material use of yucca extract is as a foaming agent for beverage manufactures, food manufacturers, and cosmetic companies. The foaming activity of yucca extract is due to the high steroidal saponin content (Oleszek et al., 2001).

In the agricultural feed industry, yucca extract is used as a livestock feed supplement. It has been reported to increase livestock growth rates (Mader and Brumm,

1987; Anthony et al., 1994), increase feed efficiency (Mader and Brumm, 1987) and improve general health of livestock (Anthony et al., 1994; Balog et al., 1994). In addition, yucca extract utilized as an animal feed supplement has been reported to reduce malodorous aromas associated livestock waste (Cheeke, 2000). Steroidal saponins have also been isolated and characterized in other Yucca species including Y. aloifolia

(Bahuguna, et al., 1991), Y. elephantipes (Zhang, et al., 2008), Y. filamentosa (Dragalin, et al., 1975), Y. glauca (Stohs and Obrist, 1975), and Y. gloriosa (Nakano, et al., 1991).

1.5.1.2. Dietary Sources of Steroidal Saponins

Plants from the Allium genus are of great agricultural importance and have a long history of use as both food and medicine. In particular, garlic, A. sativum, and onion, A. cepa, have been used in traditional medicine since ancient times (Block, 1985). The

Allium genus belongs to the Amaryllidaceae family which is closely related to the

Liliaceae family. The Allium genus is a rich source of steroidal saponins, a group of compounds that may play a role in the traditional medicinal uses of Allium species.

Steroidal saponins isolated from Allium species exhibit various biological activities including cytotoxicity (Mimaki et al., 1999a), antifungal activity (Morita et al., 1988), anti-blood coagulation activity (Peng, et al., 1996), antispasmodic activity (Corea et al.,

2005), anti-tumor activity (Sang et al., 2003), anti-platelet aggregating activity (Peng, et al., 1996), cholesterol lowering activity (Matsuura et al., 2001), and insecticidal activity

(Harmatha et al., 1987). Steroidal saponins with biological activity have been isolated and characterized in over thirty species of the Allium genus (Lanzotti, 2005). Some other

Allium species that are commonly consumed as food and have steroidal saponins with biological activity include shallots, A. ascalonicum (Fattorusso et al., 2002), leeks, A. porrum (Harmatha et al., 1987; Fattorusso et al., 2000), and elephant garlic, A. ampeloprasum (Morita et al., 1988; Mimaki et al., 1999b).

Plants from the Asparagus genus which are also high in steroidal saponins, are of great agronomic importance and have a long history of use as both food and medicine. In particular, garden asparagus, A. officinalis, is consumed worldwide and is a rich source of steroidal saponins (Shimoyamada, et al., 1990; Shimoyamada, et al., 1996; Huang and

Kong, 2006). In addition to use as a food, many Asparagus species are used in traditional medicine. In fact, the root of A. filicinus is used in traditional Chinese medicine as a treatment for colds, coughs, and pneumonia (Zhou et al., 2007) and A. racemosus is used in traditional Indian medicine for the treatment of spasm, chronic fevers, and rheumatism

(Hayes et al., 2008). The Asparagus genus belongs to the Asparagaceae family which is closely related to the Liliaceae family. Steroidal saponins isolated from Asparagus species exhibit putative biological activities including antifungal (Shimoyamada et al.,

1996), antiprotozoal (Oketch-Rabah and Dossaji, 1997) and cytotoxic activity (Zhang et al., 2004). In addition, steroidal saponins with biological activities have been identified and characterized in other members of the genus including A. acutifolius (Sautour, et al.,

2007), A. adscendens (Sharma, et al., 1982), A. africanus (Debella, et al.,1999), A. cochinchinensis (Zhang, et al., 2004), A. dumosus (Ahmad, et al., 1998), A. filicinus

(Sharma, et al., 1996), A. gobicus (Yang, et al., 2004), A. oligoclonos (Kim, et al.,

2005), and A. plumosus (Sati, et al., 1985).

1.5.1.3. Steroidal saponins isolated from Lilium

Extensive work has been done on the isolation and characterization of steroidal saponins in Lilium. Steroidal saponins have been reported in L. brownii (Mimaki and

Sashida 1990a; Mimaki and Sashida 1990b; Hou and Chen, 1998), L. candidum (Mimaki et al., 1998; Mimaki et al., 1999; Eisenreichova, et al., 2000), L. hansonii (Ori et al.,

1992), L. henryi Baker (Mimaki et al., 1993), L. longiflorum (Mimaki et al., 1994), L. mackliniae Sealy (Sashida et al., 1991), L. martagon L. (Satou et al., 1996), L. pardalinum Kellogg (Shimomura et al., 1989), L. pensylvanicum (synonym: L. dauricum)

(Mimaki et al., 1992), L. pumilum (synonym: L. tenuiflolium) (Mimaki et al., 1989), L. regale E. H. Wilson (Mimaki et al., 1993; Gur'eva et al., 1996; Kintya et al., 1996), L. speciosum var. speciosum (Mimaki and Sashida, 1991), and L. speciosum x L. nobilissimum (Makino) Makino (Nakamura et al., 1994). Steroidal saponins that have been characterized in the Lilium genus are summarized in Table 1.1 and the molecular structures are shown in Figure 1.7.

Table 1.1. Steroidal saponins isolated from Lilium.

species compound reference L. candidum 5, 10, 23, 33, 50, 51, 52, 53, 55 Mimaki et al., 1998 57, 58, 59, 60, 61, 62 Mimaki, et al., 1999 56 Eisenreichova, et al., 2000 L. regale 5, 28, 29, 30, 31 Mimaki et al., 1993 32, 33, 34 ,35, 36 Gur'eva et al., 1996 3, 5, 37, 38, 39, 40 Kintya et al., 1996 L. longiflorum 4, 23, 24, 42, 43, 44, 45, 46, 47 Mimaki et al., 1994 L. pensylvanicum 10, 22, 23, 24, 25, 26, 27 Mimaki et al., 1992 L. pardarinum 11, 12, 13, 14, 15, 16, 17 Shimomura et al., 1989 L. speciosum x nobilissum 4, 5, 10, 23, 41 Nakamura et al., 1994 L. hansonii 18, 19, 20, 21 Ori et al., 1992 L. brownii 3a, 4b, 5b, 6a, 7b Mimaki and Sashida, 1990 a, b 8, 9 Hou and Chen, 1998 L. speciosum 3, 5, 10 Mimaki ans Sashida, 1991 L. martagon 48, 49 Satou et al., 1996 L. henryi 4, 5 Mimaki et al., 1993 L. pumilum 1, 2 Mimaki et al., 1989 L. macklineae 4, 5 Sashida et al., 1991

Figure 1.7. Molecular structures of steroidal saponins from Lilium.

OH OH 18 R

12 17 19 11 13 1 -D-Glcp 16 9 14 15 1 2 -D-Allp 2 10 8

3 5 7 OH 6 RO 4

21 O 25 O 26 OR 18 1 23 24 20 27 R1 R2 22 12 17 3 HMG H 19 11 13 16 O 14 9 15 -D-Glcp 1 4 HMG OH 2 10 8

3 5 5 H 6 7 H O 4 R2O O 2' HO 1' O 1'' H3C O HO OH OH 21 O R3 18 O 20 22 12 6 H 17 19 11 13 16 O 14 OH 9 15 7 -D-Glcp 1 O OH 2 10 8 HO 3 5 7 6 HO 1''' O 4 R3O O OH 2' HO 1' O 1'' H3C O HO OH OH OH O HO O HO OH1'''' 21 26 25 23 18 O 8 24 27 20 12 9 5 -H 19 11 13 17 16 O 9 14 15 1 OH 2 10 8 4' 3 5 7 O 6 HO O 4 HO 2' O 1' O 1'' H3C 21 OCH3 HO O OH 18 23 OH 24 20 27 22 10 12 17 19 11 13 16 O 14 9 15 1 OH 2 10 8

3 5 O 6 7 HO O 4 2' HO 1' O 1'' H3C O HO OH OH

Figure 1.7. continued O R1 R2 R3 R4 O 11 OH CH3 H R2 H OCH3 25 12 H CH3 18 HO 24 27 20 22 13 OH CH3 -D-Glcp 12 17 19 11 13 O 16 CH 9 14 15 H 3 -D-Glcp 1 14 OH 2 10 8 4' 3 5 7 R1 CH2OH H O 4 6 15 H -L-Arap R3O O R O 2' H 4 1' OH CH H -L-Arap O 16 3 17 H CH3 H -L-Arap H3C O HO 21 OH 25 OH O 26 18 23 OH 24 27 20 22 12 17 19 11 13 O 16 18 14 9 15 OH 1 19 5 -H OH 2 10 8 O 4' 3 5 7 HO O 6 HO O O 4 OH 2' HO 1' O OH 1'' H C O O 3 HO O HO HO 1'''' OH OH OH 21 OH 26 25 18 23 24 27 20 22 12 17 19 11 13 16 O OH 9 14 15 1 20 O OH 2 10 8 HO 4' 3 5 7 21 5 -H HO O O 4 6 OH O HO 2' O 1' 1'' H C O 3 21 OCH3 HO OH O 25 18 OH 23 26 20 27 22 24 12 17 R R 19 11 13 1 2 16 O 14 9 15 H 1 22 -L-Arap OH 2 10 8

3 5 7 H -D-Glcp 6 23 R O O 4 2 O 2' R1O 1' O 1'' H3C O HO 21 OH 25 OH O 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 OH 2 10 8 OH 24 4' 3 5 7 6 O O 4 HO O 2' HO O 1' OH O 1'' H3C O HO OH OH

Figure 1.7. continued 21 25 O 2 18 23 6 HO 24 27 20 22 12 17 19 11 13 16 O R3 14 9 15 1 H OH 2 10 8 25

4' 3 5 7 O 6 26 -L-Arap HO O 4 2' R3O 1' H O O 1'' OH H3C O 21 OH HO 26 OH OH 18 22 20

12 17 27 19 11 13 16 14 9 15 1 OH 2 10 8 OH 27 4' 3 5 O 6 7 HO O 4 2' HO 1' H O O 1'' H C O 3 21 HO 25 OH O 26 OH 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 OH OH 2 10 8 28 3 5 7 6 O O 4 HO HO O 2' HO O 1' OH 3' O 1'' H3C O OH HO OH OH O HO O HO OH1'''' 26 25 21 OCH3 18 23 24 27 20 22 12 17 19 11 13 16 O 9 14 15 1 OH OH 2 10 8 3 5 7 29 O O 4 6 HO HO O 2' HO O 1' OH 3' O 1'' H3C O HO OH 21 OH 25 OH O 26 O OH 18 23 24 20 27 22 12 17 O O 19 11 13 16 O 14 9 15 1 OH 2 10 8 R1 R2 4' 3 5 6 7 O 4 R1O O 30 H -D-Glcp 2' R O 3' 1' 2 O 1'' H3C O HO OH OH

Figure 1.7. continued 21 25 O 26 OH 23 18 24 HO 20 27 22 12 17 19 11 13 16 O 9 14 15 1 OH OH 2 10 8 3 5 7 6 O O 4 HO HO O 2' HO 2''' 1''' O 3' 1' O 31 OH O 1'' H3C O HO O HO OH OH HO 1'''' OH 21 25 O 26 18 23 24 20 27 22 12 17 19 11 13 16 O 14 R R 9 15 1 2 1 OH 2 10 8 32 H H 3 5 7 6 O 4 33 -L-Rhap H HO O R O 2' 2 3 ' 1' OH OR1 O HO O HO OH1'''' 21 OH 26 25 18 23 24 27 20 22 12 17 19 11 13 O 16 R1 R2 9 14 15 1 OH 2 10 8 34 H H 3 5 7 O 6 35 -L-Rhap H HO O 4 2' -L-Rhap -D-Glcp R2O 3 ' 1' 36 OR1 OH 27

21 O 26 25 18 23 24 20 22 12 17 R1 R2 19 11 13 16 O 14 9 15 37 H H 1 OH 2 10 8 38 -L-Rhap -D-Glcp 4' 3 5 7 O 6 HO O 4 2' R2O 3' 1' OR1

21 25 OH O 26 O OH 18 23 24 20 27 22 12 17 O O 19 11 13 16 O 14 9 15 1 OH 2 10 8 39 4' 3 5 O 6 7 HO O 4 2' HO 3' 1' OH

Figure 1.7. continued OH O OH 27

21 26 O 25 O O 18 23 24 20 22 12 17 19 11 13 R 16 O 1 9 14 15 1 40 -D-Glcp OH 2 10 8

3 5 7 O 6 HO O 4 2' R O 3' 1' 1 O 1'' H3C O HO 21 OCH3 OH O 25 OH 18 23 26 20 27 22 24 O 12 17 19 11 13 16 O 14 O 9 15 1 O OH 2 10 8 HO 4' 3 5 7 6 O O 4 HO 1''' O 2' OH HO 1' O 41 1'' H3C O HO OH OH OH O HO O HO 1'''' OH 26 25 21 OCH3 18 23 24 27 20 22 12 17 19 11 13 16 O 9 14 15 1 R1 R2 OH 2 10 8 4' 3 5 7 O 6 42 -D-Glcp H R1O O 4 2' H -L-Arap R2O 1' 43 O -D-Xlyp 1'' 44 H H C O 3 21 HO 25 OH O 26 R OH 18 23 24 20 27 22 12 17 19 11 13 O 16 R R 9 14 1 R2 15 1 OH 2 10 8 45 O-HMG -L-Arap 4' 3 5 6 7 -L-Arap O 4 46 OH R2O O 2' H -D-Glcp R O 3' 1' 47 OH 1 O 1'' H3C O HO 21 OH OH 25 O 26 OH 23 18 24 HO 20 27 22 12 17 19 11 13 16 O 9 14 15 OH 1 OH 2 10 8 O 48 HO 3 5 6 7 O 4 HO 2''' 1''' O O 49 5 -H OH HO 3' 1' O OH HO O HO 1'''' OH

Figure 1.7. continued

21 OCH3 O 25 18 R 23 26 1 20 27 22 24 12 17 19 11 13 16 O 14 OR 9 15 2 1 OH 2 10 8 HO O 3 5 7 2' O 6 HO 1' O 4 OH H O 50 O 2' HO 1' O 51 OH -D-Glcp 1'' H3C O HO OH 21 R2 25 OH O 18 23 R 26 27 R 1 20 3 22 24 12 17 19 11 13 16 O 14 9 15 R1 R2 R 1 3 OH 2 10 8 52 H H CH 3 5 3 6 7 HO O 4 O 53 OH OCH3 CH3 2' HO 1' O 1'' H3C O HO 21 25 OH OH O OH 18 23 26 27 20 22 24 12 17 19 11 13 16 O 14 OH 9 15 1 OH 2 10 8 HO O 3 5 7 6 54 HO O 4 1''' O O OH 2' HO 1' O OH 1'' H C O O 3 HO OH HO 3'' O HO 1'''' OH OH O O 26 HO 25 21 OCH HO 1'''' 3 18 23 OH 24 27 20 22 12 17 19 11 13 16 O 9 14 15 1 R1 OH 2 10 8 4' 3 5 7 O 6 R1O O 4 HO 2' 55 H O 1' 1'' H C O 3 OC H HO OH 21 2 5 OH O 25 18 23 26 27 20 22 24 12 17 19 11 13 16 O 14 OH 9 15 OH 1 HO O 2 10 8 3 5 7 O 6 HO 1''' O 4 O OH 2' HO 1' O 56 1'' H3C O HO OH OH

Figure 1.7. continued

21 R4 O 25 18 23 R 26 27 R 1 20 3 OH 22 24 12 R 17 2 19 11 13 O HO O 16 14 R1 R2 R 9 15 R3 4 1 HO 1''' O OH 2 10 8 57 H H CH3 H 3 5 7 6 HO O 4 O 58 H H CH2OH H 2' HO 1' O 59 H OH CH3 H 1'' H3C O H CH OCH HO 60 H 3 3 OH OH OCH 61 OH H CH3 3 OH O HO O HO 1'''' OH 26 25 21 OCH3 18 23 OH 24 27 20 22 12 O 17 HO 19 11 13 16 O HO 1''' O 9 14 15 OH 1 2 10 O 8 HO 3 5 7 62 4 6 2' HO 1' O O 1'' H3C O HO OH OH

1.6 Steroidal Alkaloids

Steroidal alkaloids are a structurally diverse class of natural products that have been isolated and characterized in a wide range of organisms including plants, marine animals, and terrestrial animals (Atta-ur-Rahman and Choudhary, 1998). Structurally, steroidal alkaloids contain a steroid-type backbone and with a nitrogen atom incorporated into the molecule. The biosynthesis of steroidal alkaloids in plants differs from other plant alkaloids due to the fact that the carbon atoms in the molecule are derived from the melavonic acid pathway, whereas the carbon backbones of other alkaloids are derived from amino acids (Atta-ur-Rahman and Choudhary, 1998). Steroidal alkaloids are most commonly found in the plant families of Apocynanceae, Buxaceae, Liliaceae and

Solanaceae. Interestingly, highly cytotoxic steroidal alkaloids have also been isolated and characterized from marine organisms such as the truncate Ritterea tokiokal (Fukuzawa et al., 1994), and amphibians such as Salamandra sp. and Phyllobates sp. (Daly et al.,

2005).

1.6.1 Steroidal Alkaloids in Liliaceae

The occurrence of steroidal alkaloids in the Liliaceae family is well documented

(Li et al., 2006). Fritillaria, a genus in the Liliaceae family, and Veratrum, a genus in the closely related Melanthiaceae family, have been recognized for centuries for their pharmacological activities and have a long history of use in traditional medicine. In fact,

V. viride has been documented to be used by Native Americans for the treatment of

28 catarrah, the treatment of rheumatism, and as an insecticide against lice (Rahman and

Choudhary, 1998). Steroidal alkaloids isolated from the Liliaceae family have been of great interest in pharmacology and have been documented to exhibit various putative biological activities including antihypertensive (Oh et al., 2003), anticholinergic (Gilani et al., 1997), antifungal (Zhou et al., 2003), and anticancer (Jiang et al., 2005). In particular, the genera Veratrum and Fritillaria, have been the subject of extensive chemical characterization and pharmacological investigations. Hundreds of new steroidal alkaloids have been isolated from these plants and over 100 steroidal alkaloids have been isolated between the years of 1980 and 2005 (Li et al., 2006). The steroidal alkaloids isolated from the Liliaceae have been classified into main two groups, isosteroidal alkaloids and steroidal alkaloids, on the basis of connectivity of the carbon skeleton.

1.6.1.1. Classification of Isosteroidal Alkaloids of Liliaceae

Isosteroidal alkaloids, also referred to as Veratrum steroidal alkaloids, are characterized by a C-nor-D-homo-[14(13→12)-abeo] ring system (Li et al., 2006).

Isosteroidal alkaloids are further sub-divided into three groups according to the linkage patterns between rings E and F into cevanine type, veratramine type, and jervine type.

The cevanine type, structurally characterized by the hexacyclic benzo [7,8] fluoreno [2,1- b] quinolizine nucleus, constitutes the largest class. The veratramine type is characterized by the absence of ring E and the presence of an aromatic ring D. Analogues with an unaromatized ring D are also placed in this class. The jervine type contain hexacyclic

29 compounds that have the furan ring E fused onto a piperidine ring system forming an ether bridge between carbon atoms at C17 and C23.

Figure 1.8. Isosteroidal alkaloids of Liliaceae: Representative examples of cevanine type (A), veratramine type (B), and jervine type (C) isosteroidal alkaloids, isolated from F. imperialis, F. ningguoensis, and F. camtschatcensis, respectively (Li et al., 2006).

HO A OH

H N

HO B O

O HO

HO C OH

1.6.1.2. Classification of Steroidal Alkaloids of Liliaceae

The steroidal alkaloids of Liliaceae, also referred to as Solanum steroidal alkaloids, are characterized by a six membered C-ring and a five-membered D-ring. The steroidal alkaloids are further sub-divided to two groups on the basis of the position of the nitrogen atom. Steroidal alkaloids with the nitrogen atom incorporated into an indolizidine ring are referred to as solanidine type. The solanidine type is derived from epiminocholestanes with the amino group incorporated into an indolizidine ring, resulting in a hexacyclic carbon framework. If the nitrogen atom is incorporated into a piperidine ring, they are of verazine type. The verazine type is based on the 22/23,26- epiminocholestane heterocyclic skeleton.

Figure 1.9. Steroidal alkaloids of Liliaceae: Representative examples of solanidine type (A) and verazine type (B) steroidal alkaloids, isolated from F. delavayi and F. ebeiensis var. purpurea, respectively (Li et al., 2006).

HO A HO OH

HO OH B O

1.6.1.3. Steroidal Alkaloids in Lilium

Although extensive work has been done on the characterization of steroidal alkaloids in the Liliacece family, less is known on the steroidal alkaloids in the Lilium genus. In 1996 Noor-e-Ain reported on the identification of two steroidal alkaloids from

L. candidum, commonly known as the Madonna lily (Noor-e-Ain, 1996). Two 22, 26- epiminocholestane-type steroidal alkaloids, named etioline and teinemine, that were previously found in several Solanum and Veratrum species were isolated from L. candidum bulbs (Lin et al., 1986; Atta-ur-Rahman and Choudhary, 1998). In 2001,

Erdoğan et al. also reported on the isolation of etioline from L. candidum bulbs (Erdoğan et al., 2001). Although these compounds were isolated as free form steroidal alkaloids, the isolation procedure was performed over several weeks under acidic conditions and it is unclear whether the conditions caused glycosidic cleavage, resulting in the formation of artifacts during the isolation process. Thus far, the glycosylated forms of these two steroidal alkaloids have not been identified L. candidum. Nevertheless, these compounds are the only free form steroidal alkaloids that have been reported from the Lilium genus.

Figure 1.10. Structures of the steroidal alkaloids (A) etioline and (B) teiemine isolated from the bulbs of L. candidum (Noor-e-Ain, 1996; Erdoğan et al., 2001)

21 25 N 26 18 27 24 20 23 22 12 17 19 11 13 16 14 9 15 1 OH A 2 10 8

3 5 6 7 HO 4 21 H 25 N 26 18 24 27 20 23 22 12 17 19 11 13 16 14 9 15 1 OH 2 10 8 B 3 5 6 7 HO 4

1.6.2. Steroidal Glycoalkaloids

In contrast to the widespread distribution of steroidal alkaloids in plants and animals, steroidal glycoalkaloids, thus far, have only been identified in the Solanaceae and Liliaceae families (Ghisalberti, 2006). Steroidal glycoalkaloids are characterized by a steroidal alkaloid type aglycone glycosidically linked to carbohydrate moieties. The most common aglycones of the steroidal glycoalkaloids can be classified based upon their structural features into six major groups. The first group is referred to as (1) spirosolanes and are characterized by an oxazaspirodecane ring system. The second group, the (2) 22,

26-epiminocholestanes, are characterized by a 22/23, 26-epiminocholestane heterocyclic skeleton. The third group, the (3) solanidanes, are characterized by a fused indolizidine ring system. The fourth group, the (4) epiminocyclohemiketals, are characterized the

33 presence of an epiminocyclohemiketal functionality. The fifth group, the (5) 3- aminospirostanes, is characterized by an amino group at the C3-position. The sixth group, the (6) leptines, are characterized by a fused indolizidine ring system with 23-hydroxy or

23-acetoxy moieties. The most common steroidal glycoalkaloids belong to the solanidane and the spirosolane groups (Ghisalberti, 2006). The carbohydrate moiety of steroidal glycoalkaloids is -glycosidically linked at the C-3 hydroxy position of the steroidal alkaloid backbone. The most common sugars are D-glucose, D-galactose, D-xylose, L- rhamnose, and L-arabinose. Similar to the steroidal saponins, all dextrorotatory form sugars are linked via a -glycosidic linkage and all levorotatory form sugars are linked via an -glycosidic linkage. The composition of the carbohydrate moiety can range from one sugar to multiple sugars, linked in a linear or branched arrangement.

Figure 1.11. Structures of the most common aglycones of the steroidal glycoalkaloids: (A) spirosolanes, (B) 22, 26-epiminocholestanes, (C) solanidanes, (D) epiminocyclohemiketals, (E) 3-aminospirostanes, and (F) leptines

H N N

HO A B HO

N O OH

C D HO H2N

R O

O N

E F R 1 H2N HO OH 2 OAc

1.6.2.1. Dietary Sources of Steroidal Glycoalkaloids

Steroidal glycoalkaloids from the Solanaceae family are found in many agriculturally important foods crops such as tomato, S. lycopersicum, potato, S. tuberosum, eggplant, S. melongena, and pepper, Capsicum annuum. -Solasonine and - solamargine are the two predominant steroidal glycoalkaloids found in the common cultivated eggplant, S. melongena (Blankemeyer et al. 1998). -Solasonine and - solamargine share the same aglycone, solasodine, but differ only in the carbohydrate moiety. The carbohydrate moiety of -solasonine contains a trisaccharide moiety, 3-O- -

L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→3)- -D-galactopyranoside, whereas

-solamargine contains a disaccharide moiety, 3-O- -L-rhamnopyranosyl-(1→2)- -D- galactopyranoside. Solasodine is a common aglycone and has been identified in over 200

Solanum species alone (Dinan et al. 2001). -Solanine and -chaconine are the most prevalent glycoalkaloids found in cultivated potato, S. tuberosum. Similar to -solasonine and -solamargine, -solanine and -chaconine share the same aglycone, solanidine, but differ only in the carbohydrate moiety. The carbohydrate moiety of -solanine contains a trisaccharide, 3-O- -L-rhamnopyranosyl-(1→2)- -D-galactopyranosyl-(1→3)- -D- glucopyranoside, whereas -chaconine contains a disaccharide moiety, 3-O- -L- rhamnopyranosyl-(1→2)- -D-galactopyranoside. -Tomatine and dehyrotomatine are the most abundant steroidal glycoalkaloids in the leaves and unripe fruit of tomato, S. lycopersicum. In the case of -tomatine and dehyrotomatine, they only differ in the degree of saturation between the C5 and C6 carbon of the aglycone and share the same

36 carbohydrate moiety, namely 3-O- -D-glucopyranosyl-(1→2)- -D-xylopyranosyl-

(1→3)- -D-glucopyranosyl-(1→4)- -D-galactopyranoside.

Steroidal glycoalkaloids are toxic to many organisms including bacteria, fungi, and animals. The literature suggests that their toxicological effects are based on the disruption of cellular membranes, anticholinesterase activity, and their effect on the active transport of ions through membranes (Friedman et al., 1992a; Keukens et al., 1995;

Blankemeyer et al., 1992; Blankemeyer et al., 1995). Steroidal glycoalkaloids are amphiphilic in nature due to the lipophilic steroidal aglycone and the hydrophilic carbohydrate moiety. Recent evidence suggests that this structural characteristic contributes to their biological activity. Accordingly, free aglycones are less active as compared to their glycosylated forms (Roddick, 1989). Interestingly, chaconine and

solanine share the same aglycone and only differ in the carbohydrate moiety; however,

chaconine has been shown to be the more teratogenic as compared to solanine

(Blankemeyer et al., 1997; Blankemeyer et al., 1998). Furthermore, spirosolanes have been shown to be less teratogenic as compared to solanidanes. Differences in the carbohydrate moiety, the absolute configuration of the F ring, and saturation between C5 and C6 has been shown to play a role in biological activity (Gaffield and Keeler, 1996).

In addition, different forms of steroidal glycosides that occur in the same plant have been shown to act synergistically, thus the importance of investigating the biological activity of the compounds both individually and in mixtures (Rayburn et al., 1994; Smith et al.,

2001). Due to the structural diversity of steroidal glycoalkaloids, differential biological activities have been observed. For example, some steroidal glycoalkaloids are highly toxic whereas others may potentially have beneficial properties, in particular the

37 spirosolanes. Beneficial biological activities that have been observed include cholesterol and triglyceride lowering, anti-inflammatory activity, anti-viral activity, and anti-cancer activity (Friedman et al. 2000; Kuo et al., 2000; Ikeda et al., 2003; Carter and Lake,

2004).

1.6.2.2. Steroidal Glycoalkaloids in Lilium

Although steroidal alkaloids are well documented in the Solanaceae and Liliaceae families, a very small number of steroidal glycoalkaloids have been isolated from the

Lilium genus. In Lilium, steroidal glycoalkaloids have been identified in L. philippinense

(Espeso et al., 1990), L. mackliniae (Sashida et al., 1991), and L. brownii (Mimaki and

Sashida, 1990a; Mimaki and Sashida, 1990b). In 1990, Espeso et al. reported the isolation of a steroidal glycoalkaloid from the aerial parts of the L. philippinense, commonly known as the Banquet lily. Interestingly, the compound was a veratramine type isosteroidal alkaloid glucoside, containing an unaromatized D ring. Prior to this work, the only steroidal glycoalkaloid previously isolated from the genus was from L. cordatum; however, this plant was taxonomically moved into the closely related genus

Cardiocrinum (Endl.) Lindl. Nevertheless, in 1987 Nakano et al. reported the isolation and structural elucidation of (25R)- and (25S)-22,26-epimino-5α-cholest-22(N)-en-3β,6β- diol O(3)-β-D-glucopyranoside from C. cordatum (Nakano et al., 1987). Also in 1990,

Mimaki et al. reported on the isolation a structural elucidation of two steroidal glycoalkaloids from L. brownii, (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside and (22,R

25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranoside. Both alkaloids were solasodine-based and differed only in the number of sugars in the carbohydrate moiety. In 1991, Sashida et al. reported on the isolation of a solanidine- based steroidal glycoalkaloid, solanidine O- -L-rhamnopyranosyl-(1→2)- -D- glucopyranosyl-(1→4)- -D-glucopyranoside, from the bulbs of L. mackliniae.

Interestingly, this compound was isolated previously from two Fritillaria species, F. thunbergii and F. camtschatcensis (Sashida et al., 1991). In contrast to the steroidal saponins which have been extensively isolated and characterized in the genus, there are a very limited number of reports on the occurrence of steroidal glycoalkaloids in the Lilium genus.

Figure 1.12. Structures of steroidal glycoalkaloids isolated from C. cordatum (1, 2), L. philippinense (3), L. brownii (4, 5), and L. mackliniae (6).

R1 21 25 N 26 18 24 R2 20 23 22 12 17 19 11 13 16 14 9 15 1 OH 2 10 8 R1 R2

3 5 7 O 6 1 CH3 H HO O 4 HO 2 H CH3 OH OH 21 HN 18 26

20 22 25 27 23 24 O 13 17 16 19 11 12 14 15 1 9 3 OH 2 10 8

3 5 7 O 6 HO O 4 HO 21 H 25 OH N 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 R OH 2 10 8 1

4' 3 5 7 6 4 H O 4 R1O O 2' 5 -D-Glcp HO 1' O 1'' H3C O HO OH OH 21 18 20 23 22 26 12 17 19 11 13 16 N 24 14 27 9 15 OH 1 O OH 2 10 8 HO 3 5 7 O 6 HO O O 4 2' 6 OH HO 1' O 1'' H3C O HO OH OH

1.7. Plant organ distribution of steroidal glycosides

Although steroidal glycosides are widespread in plants and have been identified in almost all plant organ types, there are only a few investigations comparing the concentration of these natural products in the different organs of the same plant species.

Even less is known about tissue specific localization and cellular and sub-cellular synthesis and storage of these compounds. This is due, in part, to the complexity of performing quantitative analysis of this diverse class of natural products.

Steroidal glycosides lack a strong chromophore; therefore, quantitative methods using non-specific short wavelength ultraviolet (UV) detection (200 – 210nm) is a challenge due to interference from phytochemicals with strong chromophores that occur in the same plant matrix. Analytical methods using evaporative light scattering detection

(ELSD) have been developed to help overcome this obstacle; however, laborious sample preparation and sensitivity issues persist (Oleszek and Bialy, 2006). Recently, analytical methods utilizing liquid chromatography - mass spectrometry (LC-MS) operating in selected ion monitoring (SIM) mode and liquid chromatography–tandem mass spectrometry (LC-MS/MS) have been developed with increased sensitivity and specificity over other methods (Oleszek and Bialy, 2006). Due to recent advances in analytical chemistry, including better chromatography and analytical instrumentation, more studies on the organ distribution of steroidal glycosides in plants will become increasingly available.

Although reports on the organ distribution of steroidal glycosides in plants are not common, there have been some reports on the concentrations of these compounds within the different plant organs. The organ distribution of steroidal glycosides in plants such as

Solanum nigrum, Solanum incanun (Eltayeb et al., 1997), Asparagus officinalis L. (Wang et al., 2003) and Dioscorea pseudojaponica Yamamoto (Lin et al., 2008) have been reported. Steroidal glycoalkaloids are generally found in all plant organs, with the highest concentrations occurring in the metabolically active parts including in flowers, sprouts, immature fruits, young leaves and shoots. In 1997, Eltayeb et al. reported on the concentrations of steroidal glycoalkaloid, solasodine, in the different organs of S. nigrum and S. incanun throughout maturation using a colorimetric assay (Eltayeb et al., 1997).

Interestingly, all of the plant organs tested contained detectable levels of solasodine with the highest concentrations occurring in small actively growing leaves. In addition, the concentration in the roots was observed to be higher than in the stems. In 2003, Wang et al. developed a LC-MS method and quantified the concentration of the furostanol saponin, protodioscin, in common garden asparagus, A. officinalis (Wang et al., 2003).

The distribution of protodioscin within the shoots was found to vary in concentration, with the highest concentration observed in the lower stem tissue adjacent to the rhizome.

In 2008, Lin and Yang reported on the concentration of several steroidal saponins in the different organs of the yam, D. pseudojaponica, using HPLC-ESLD (Lin et al., 2008).

Saponins were found occur in increasing concentrations in the vine, leaf, rhizophor, tuber flesh and then tuber cortex. Although there are some reports on the distribution of steroidal glycoside in the different plant organs, reports on tissue specific localization are even less common.

Although scarce, there are some reports on the tissue specific localization and cellular localization of steroidal glycosides in plants. In potato, S. tuberosum, the concentration of the steroidal glycoalkaloid, -solanine, has been observed to increase in potato tubers in response to mechanical wounding (McKee, 1955; Ishizaka and Tomiyama, 1972). When potato tubers are wounded, the development of meristematic tissue near the site of the wounded tissue occurs. This area develops a suberized wound periderm that is believed provide a protective structural barrier from fungi and bacteria. It has been shown that the tissues adjacent to the wound periderm accumulate -solanine, suggesting a role in wound healing (McKee, 1955). In addition, in D. pseudojaponica, differential concentrations of steroidal saponins were observed in the inner yam tuber cortex as compared to the tuber flesh (Lin et al., 2008). Furthermore, histological observations of the cellular localization of furostanol saponins in D. caucasia have been reported with furostanol accumulation in specialized idioblasts in the upper and lower leaf epidermis as compared to the leaf mesophyll where no furostanols were detected (Gurielidze et al.,

2004). Although researchers have reported on the occurrences of steroidal glycosides in the different plant organs, plant tissues, and plant cell types, considering the widespread abundance of these compounds in the plant kingdom, investigations this area of plant biology is surprisingly lacking.

1.8. Steroidal glycosides in plant defense

Plants have multiple protection strategies from plant pathogens and herbivory.

The strategies include structural barriers, constitutive chemical defenses and inducible

43 chemical defenses. The first line of plant defense is the morphological structure of the plant‘s surface. Physical structures such as the waxy cuticle and epidermal cell wall serve as a protective barrier from pathogens. Specialized structures such as thorns or spines serve as protection from herbivory. In addition structural features, it has been well established that plant derived secondary metabolites can increase plant resistance to pathogens and herbivory. Plant chemical defenses are broadly classified as constitutive chemical defenses, known as phytoanticipins, or inducible chemical defenses, known as phytoalexins (VanEtten et al, 1994; Müller and Börger, 1940). These definitions are based on how the compounds are regulated rather than by chemical structure.

Accordingly, the definitions are plant species specific and in some case even tissue specific within the same species.

Phytoalexins are inducible secondary metabolites that are produced in response to infection or by chemical or mechanical injury (Müller and Börger, 1940). The synthesis of phytoalexins occur in healthy cells adjacent to wounded or infected cells and are induced by chemical elicitors released from damaged cells. Chemical elicitors trigger the expression of phytoalexin biosynthetic genes resulting in de novo synthesis. Phytoalexins are toxic to many organisms including fungi, bacteria, and animals. Some well known examples of phytoalexins include resveratrol in grape, pisatin in pea, and capsidiol in pepper.

Phytoanticipins are secondary metabolites that are constitutively present in plant cells prior to infection (VanEtten et al, 1994). Some phytoanticipins are present in their biologically active forms and in some cases present as biologically inactive precursors.

Upon cellular disruption by infection or mechanical damage, the biologically inactive

44 precursors are rapidly converted to biologically active forms. Phytoanticipin distribution within plants is often tissue specific and in some plants the compounds are localized in the outer cell layers, creating a defensive barrier to invading pathogens. For example, the plant pathogenic fungus Colletotricum circinans infects white onions but does not infect pigmented onions. In pigmented onions, phenolic compounds that are inhibitory to C. circinans are preferentially accumulated in the bulb skin, thus providing a protective barrier to infection (Link and Walker, 1933).

Steroidal glycosides are a class of compounds that in many plant species are considered phytoanticipins due to the fact that they are constitutively present in healthy plant tissues and are inhibitory or toxic to some organisms (Osbourn, 1996). The role of steroidal glycosides in plant defense, including antifungal and antiherbivory, has been studied extensively (Zullo et al., 1984; Nozzolillo et al., 1997; Adel et al., 2000; Bowyer et al., 1995; Osbourn, 1996; Morrissey and Osbourn, 1999; Osbourn, 1999;

Papadopoulou et al., 1999; Morrissey et al., 2000; Trojanowska et al., 2000; Osbourn et al., 2003; Osbourn, 2003; Hughes et al., 2004; Choi et al., 2005). For example, some steroidal glycosides are toxic to insects such as the European corn borer, Ostrinia nubialis, and army worm, Spodoptera littoralis (Nozzolillo et al., 1997; Adel et al.,

2000). In oats, Avena sativa, biologically inactive steroidal saponins are converted into an antifungal form in response to tissue damage, suggesting a role in the plant-pathogen interaction (Osbourn, 1996; Morrissey et al., 2000; Osbourn, 2003; Hughes et al., 2004).

In addition, the steroidal glycoalkaloids -tomatine and -chaconine play a role in fungal resistance of tomato, Solanum lycopersicum, and potato, Solanum tuberosum,

45 respectively (Morrissey and Osbourn, 1999); however, the exact mechanism of resistance to the pathogens has not been fully elucidated.

In general, the molecular mechanism for the antifungal activity of steroidal glycosides is not well characterized; however, interaction with cellular membranes has been proposed to play a role. Due to the amphipathic nature of the molecules, steroidal glycosides have been shown to disrupt cell membranes both in vitro and in vivo (Steel and Drysdale, 1988; Roddick et al, 2001). Some studies suggest that membrane disruption may be due either to the interaction of the aglycone with membrane bound sterols, resulting in the formation of membrane pores (Armah et al., 1999) or the extraction of membrane bound sterols, causing loss of lipid bilayer integrity and membrane leakage (Keukens et al, 1992; Keukens et al, 1995). Despite the fact that steroidal glycosides have antifungal properties and may play a protective role against potential pathogens, in the case of successful pathogens this is not sufficient and infection occurs. Some mechanisms that fungal pathogens utilize to overcome host defense strategies are avoidance, tolerance, and enzymatic metabolism of plant defense compounds.

1.9 Detoxification of steroidal glycosides

Some successful fungal pathogens have the ability to overcome plant defenses by the metabolism of plant defense compounds. The plant pathogen Botrytis cinerea has been shown to produce enzymes that can metabolize a variety of plant defense compounds from active forms to inactive forms (Staples and Mayer, 1995). For example, B. cinerea

46 produces laccases that metabolize plant defense compounds induced during infection and reduces lignification by the host (Bar-Nun et al., 1988). In addition, fungal produced lactases have been shown to detoxify phytoalexins from a wide variety of plants by cross- linking host produced phenols (Van Etten et al., 1989; Pezet et al., 1992). B. cinerea produced lactases are capable of degrading the grape phytoalexins, pterostilbene and resveratrol, that are up regulated by the plant in response to infection, thus overcoming the plant‘s defense response (Pezet et al., 1992).

The enzymatic detoxification of steroidal glycosides by fungal pathogens is well documented (Arneson et al., 1967; Verhoeff and Liem, 1975; Ford et al., 1977; Bowyer et al., 1995; Morrissey et al., 2000). Plant pathogens such as Gaeumannomyces graminis and Stagonospora avanae have the ability to enzymatically detoxify host plant saponins by sugar cleavage, resulting in loss of antifungal activity of the compounds (Bowyer et al., 1995; Morrissey et al., 2000). In tomato, B. cinerea metabolizes the antifungal steroidal glycoalkaloid, -tomatine, to an inactive form by enzymatic cleavage of the entire carbohydrate moiety, or by the cleavage of the terminal xylose by a -xylosidase enzyme (Verhoeff and Liem, 1975; Quidde and Osbourn, 1998). In addition, other plant pathogenic fungi such as Septoria lycopersici and Fusarium oxysporum f.sp. lycopersici detoxify -tomatine by cleavage of sugar residues through two separate independent metabolic pathways (Arneson et al., 1967; Ford et al., 1977). The ability to efficiently detoxify host plant chemical defenses may play a role in the virulence and host range of some fungal pathogens (Bowyer et al, 1995). Pedras et al. suggested that a better understanding the detoxification pathways utilized by plant pathogenic fungi could potentially lead to new approaches to control plant pathogens (Pedras et al., 2011).

Plants produce both constitutively expressed and inducible plant defense compounds as a means of protection from plant pathogens. Some compounds are stored in an active form and are preferentially localized, creating a protective barrier to potential plant pathogens. In other cases, precursors are stored and rapidly activated upon infection.

Inducible plant defense compounds are produced in healthy cells by signaling molecules that originate in damaged cells. Despite the chemical diversity of plant defense compounds and the differential mechanisms of expression, successful plant pathogens evolved multiple means by which they overcome plant defenses and thus cause infection.

Friedman and McDonald speculated that in response to plant pathogen‘s ability to overcome host plant defenses, some plants may have adapted structural modifications to plant defense compounds, thus increasing their biological activity (Friedman and

McDonald, 1997). For example, in many Solanaceous species ―paired‖ glycoalkaloids occur that share the same aglycone, only differ in the carbohydrate moiety, and express differential biological activity (Friedman and McDonald, 1997; Roddick et al., 2001). A plant defense strategy aimed at inhibiting the pathogens ability of to detoxify plant defense compounds may be an alternative strategy in plant defense. The structural modification of plant defense compounds as an adaptive response to plant pathogens is an interesting hypothesis and needs further exploration.

CHAPTER 2: Isolation and Structural Determination of Steroidal Glycosides from the Bulbs of Easter Lily (Lilium longiflorum Thunb.)

2.1. Abstract

The bulbs of Easter Lily (Lilium longiflorum Thunb.) are used as a food and medicine in several Asian cultures and they are cultivated as an ornamental plant throughout the world. A new steroidal glycoalkaloid and two new furostanol saponins, along with two known steroidal glycosides, were isolated from the bulbs of L. longiflorum. The new steroidal glycoalkaloid was identified as (22R, 25R)-spirosol-5-en-

3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D- glucopyranoside. The new furostanol saponins were identified as (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- arabinopyranosyl-(1→3)- -D-glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-

(1→3)- -D-glucopyranoside. The previously known steroidal glycosides, (22R, 25R)- spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D- glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-

O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside were identified in L. longiflorum for the first time. These new compounds from L. longiflorum and the isolation methodologies employed can be used for studies on the biological role

49 of steroidal glycosides in plant development and plant-pathogen interactions, as well as for studies in food and human health, for which little is known.

2.2. Introduction

The Easter lily (Lilium longiflorum Thunb., family Liliaceae) has beautiful white flowers and a delicate aroma and is appreciated worldwide as an attractive ornamental. In addition to its economic importance and popularity in horticulture, lily bulbs are regularly consumed in Asia, as both food and medicine. The bulbs of several Lilium species, including L. longiflorum, L. brownii, L. pensylvanicum, and L. pumilum, have been used traditionally in China as sedatives, anti-inflammatory and antitussive agents, and a general tonic (Su, 1979; Mimaki et al., 1990; Mimaki et al., 1992, Varrier, 2002). The crude drug ―Bai-he‖, used in traditional Chinese medicine, is prepared from the bulbs of

Lilium sp. and is regularly used for lung ailments in China today. Although the medicinal use of L. longiflorum is well documented, the compounds responsible for the reported properties are not known.

Bulbs of the genus Lilium are a rich source of secondary metabolites, including bitter phenylpropanoid glycosides identified in the bulbs of L. speciosum Thunb.

(Shimomura et al., 1986), antitumor alkaloids identified in L. hansonii Leichtlin ex D. D.

T. Moore (Shimomura et al., 1987), and steroidal glycoalkaloids identified in L. brownii

(Mimaki and Sashida, 1990b). Extensive work has been done on the isolation and characterization of steroidal saponins in Lilium. Steroidal saponins have been reported in

L. brownii (Mimaki and Sashida, 1990a; Mimaki and Sashida, 1990b; Sashida et al.,

1990; Hou and Chen, 1998; Ji et al., 2001), L. candidum (Mimaki et al., 1998; Mimaki et al., 1999; Eisenreichova et al., 2000), L. hansonii (Ori et al., 1992), L. henryi (Mimaki et al, 1993), L. longiflorum (Mimaki et al, 1994), L. mackliniae (Sashida et al., 1991), L. martagon (Satou et al., 1996), L. pardalinum (Shimomura et al, 1989), L. pensylvanicum

(Mimaki et al., 1992), L. pumilum (Mimaki et al, 1989), L. regale (Mimaki et al, 1993,

Gureva et al, 1996, Kintya et al, 1997), L. speciosum var. speciosum (Mimaki et al,

1991), and L. speciosum x L. nobilissimum (Nakamura et al, 1994).

Steroidal saponins have been reported to exhibit a wide range of biological activities including antifungal (Sautour et al., 2005), platelet aggregation inhibition

(Zhang et al., 1999; Huang et al., 2006), antidiabetic (Nakashima et al, 1993), cholesterol lowering (Matsuura et al., 2001), anti-inflammatory (Shao et al., 2207), antiviral (Gosse et al., 2002), and anticancer (Mimaki et al., 1999; Pettit et al., 2005; Acharya et al.,

2009). Although the putative biological activities of steroidal saponins are well documented, the biological role of these compounds within the plant is poorly understood.

The occurrence of steroidal alkaloids in the Liliaceae family is also well documented (Li et al, 2006). Steroidal alkaloids isolated from the Fritillaria genus and the closely related Veratrum genus show various biological activities including antihypertensive (Oh et al., 2003), anticholinergic (Gilani et al., 1997), antifungal (Zhou et al, 2003), and anticancer (Jiang et al., 2005). In Lilium, steroidal alkaloids have been identified in L. candidum L. (Erdoğan et al., 2001) and steroidal glycoalkaloids have been identified in L. philippinense (Espeso and Guevara, 1990), L. mackliniae (Sashida et al.,

1991), and L. brownii (Sashida et al., 1990); however, no steroidal alkaloids or steroidal glycoalkaloids have been previously reported from L. longiflorum.

With regard to the steroidal glycosides of L. longiflorum bulbs, six spirostanol saponins and three furostanol saponins with antitumor activity have been reported

(Mimaki et al, 1994). To set the stage for biological activity studies on the role of steroidal glycosides in plant development and plant-pathogen interactions, as well as for studies in food and human health, this chapter reports the isolation and structural determination of several new steroidal glycosides from the bulbs of L. longiflorum. The structures of the steroidal glycosides were elucidated by a combination of spectroscopic and chemical analysis.

2.3. Materials and Methods

2.3.1. Plant Material.

L. longiflorum, cultivar 7-4, bulbs were provided from the Rutgers University lily breeding program. Bulbs were treated with Captan (Bayer CropScience AG, Monheim am Rhein, Germany) fungicide prior to planting. Bulbs were planted in beds containing

Pro-Mix (Premier Horticulture Inc., Quakertown, PA) soil mix and were grown to mature plants under greenhouse conditions for 9 months prior to harvest. The greenhouse temperatures were set to provide a minimum day temperature of 24 ºC and a minimum

52 night temperature of 18 ºC. Plants were fertilized biweekly with a 100 mg L-1 solution of

NPK 15-15-15 fertilizer (J. R. Peters Inc., Allentown, PA). Each plant produced three to five new bulbs, which were used for extraction. The new bulbs were full-sized, fresh, and mature at harvest. Each plant was harvested by hand, and the bulbs were manually separated, immediately frozen under liquid nitrogen, lyophilized on a VirTis AdVantage laboratory freeze-dryer (SP Industries Inc.,Warminster, PA), and stored at -80 ºC until analysis.

2.3.2. Chemicals.

The following compounds were obtained commercially: (25R)-spirost-5-en-3β-ol,

Sephadex LH-20, N,O-bis(trimethylsiyl)trifluoroacetamide with trimethylchlorosilane

(99:1) silylating reagent, Dragendorff reagent, p-(dimethylamino)benzaldehyde, hydrochloric acid, sodium hydroxide, pyridine-d5 (0.3% v/v TMS), D-(+)-glucose, L-(-)- glucose, D-(+)-rhamnose, L-(-)-rhamnose, D-(+)-arabinose, L-(-)-arabinose, D-(+)- xylose, and L-(-)-xylose (Sigma-Aldrich, St. Louis, MO); and (22R,25R)-spirosol-5-ene-

3β-ol (Glycomix Ltd., Reading, U.K.). All solvents (acetonitrile, chloroform, ethanol, ethyl acetate, formic acid, n-butanol, and n-pentane) were of chromatographic grade

(Thermo Fisher Scientific Inc., Fair Lawn, NJ). Water was deionized (18 MΩ cm) using a

Milli-Q water purification system (Millipore, Bedford, MA).

2.3.3. Isolation and Purification of Steroidal Glycosides 1 – 5 from L. longiflorum.

2.3.3.1. Sequential Solvent Extraction of Lyophilized L. longiflorum Bulbs

Lyophilized lily bulbs (100 g) were frozen in liquid nitrogen, ground into a fine powder with a laboratory mill (IKA Labortechnik, Staufen, Germany), and extracted with n-pentane (3 x 100 mL) on a wrist-action autoshaker (Burrell Scientific, Pittsburgh, PA) at room temperature for 30 min. After centrifugation (5000 rpm for 10 min) (Sorvall RC-

3C Plus, Thermo Fisher Scientific Inc.), the organic layers were discarded and the pellet was freed from residual solvent in a fume hood overnight. The residual defatted material was then extracted with a mixture of ethanol and water (7:3, v/v; 2 x 150 mL) on an autoshaker for 45 min at room temperature (25 ºC). After centrifugation (5000 rpm for 10 min) and vacuum filtration through a Whatman 114 filter paper (Whatman International

Ltd., Maidstone, U.K.), the supernatant was collected and the residue discarded. The supernatant was then evaporated under reduced pressure (30 ºC; 1.0 x 10-3 bar) using a

Laborota 4003 rotary evaporator (Heidolph Brinkman LLC, Elk Grove Village, IL) and lyophilized, yielding a crude bulb extract (13.7 g). The lyophilized crude bulb extract was then dissolved in deionized water (100 mL) and washed with ethyl acetate (5 x 100 mL), and the organic phase was discarded. The aqueous phase was then extracted with n- butanol (5 x 100 mL) and the aqueous phase discarded. The organic phase was then evaporated under reduced pressure (30 ºC; 1.0 x 10-3 bar) and lyophilized, yielding a crude glycoside extract (2.42 g) (Figure 2.1).

Figure 2.1. Isolation scheme for compounds 1 – 5 isolated from the bulbs of L. longiflorum. EtOH, ethanol; EtAC, ethyl acetate; n-BuOH, n-butanol; ACN, acetonitrile.

2.3.3.2. Gel Permeation Chromatography (GPC)

Crude glycoside extract (1.0 g) was dissolved in a solution of ethanol and water

(7:3, v/v; 5.0 mL), filtered with a 0.45 μm PTFE syringe filter (Thermo Fisher Scientific

Inc.), and then applied onto a standard threaded 4.8 cm x 60 cm glass column (Kimble

Chase Life Science and Research Products LLC, Vineland, NJ) packed with Sephadex

LH-20 (Amersham Pharmacia Biotech, Uppsala, Sweden) that was washed and conditioned in the same solvent mixture overnight. Chromatography was performed with

55 isocratic ethanol/water (70:30, v/v) at a flow rate of 3.5 mL min-1. The first 200 mL of effluent was discarded, and 30 fractions (25 mL each) were collected and analyzed by

LC-MS. LC-MS analyses were performed on a HP 1100 series HPLC (Agilent

Technologies Inc., Santa Clara, CA) equipped with an autoinjector, a quaternary pump, a column heater, and a diode array detector and interfaced to a Bruker 6300 series ion-trap mass spectrometer equipped with an electrospray ionization chamber. HP ChemStation and BrukerData Analysis software were used for data acquisition and data analysis.

Reverse phase separations were performed using a Prodigy C18 column (250 mm x 4.6 mm i.d.; 5.0 μm particle size) (Phenomenex, Torrance, CA). The flow rate was set to 1.0 mL min-1, and the column temperature was at 23 ± 2 ºC. The binary mobile phase composition consisted of (A) 0.1% formic acid in deionized water and (B) 0.1% formic acid in acetonitrile. Chromatography was performed using a linear gradient of 15 – 43%

B over 40 min and then to 95% B over 5 min; thereafter, elution with 95% B was performed for 10 min. The re-equilibration time was 10 min. All mass spectra were acquired in positive ion mode over a scan range of m/z 100 – 2000. Ionization parameters included capillary voltage, 3.5 kV; end plate offset, -500 V; nebulizer pressure, 50 psi; drying gas flow, 10 mL min-1; and drying gas temperature, 360 ºC. Trap parameters included ion current control, 30000; maximum accumulation time, 200 ms; trap drive,

61.2; and averages, 12 spectra. On the basis of the LC-MS profile, GPC fractions 6 – 17 were combined, evaporated under reduced pressure (30 ºC; 1.0 x 10-3 bar), and lyophilized, yielding GPC fraction A (180 mg).

2.3.3.3. Semipreparative Reverse-Phase High-Performance Liquid Chromatography (RP-

HPLC)

Fractionation of GPC fraction A was achieved by semipreparative RP-HPLC performed on a Luna C18 column (250 mm x 21.2 mm i.d.; 10 μm particle size)

(Phenomenex) to afford 1 – 5 (Figure 2.2). Chromatography was performed on a

Shimadzu LC-6AD liquid chromatograph (Shimadzu Scientific Instruments Inc.,

Columbia, MD) using a UV-vis detector and a 2 mL injection loop. Mixtures of (A) 0.1% formic acid in deionized water and (B) 0.1% formic acid in acetonitrile were used as the mobile phase. The flow rate was set to 20 mL min-1, the column temperature was 23±2

ºC, and UV detection was recorded at λ=210nm. GPC fraction A was dissolved in a mixture of mobile phase A and mobile phase B (75:25, v/v) and filtered through a 0.45

μm PTFE syringe filter prior to injection. Chromatography was performed using a linear gradient of 5 – 30% B over 45 min and then to 90% B over 10 min; thereafter, elution with 90% B was performed for 10 min. The re-equilibration time was 10 min.

Figure 2.2. RP-HPLC chromatogram (λ = 210 nm) of 1 – 5 isolated from L. longiflorum n-butanol extract fractionated by gel permeation chromatography (GPC).

20 3 4

1 210)

10 5

2 Intensity ( Intensity

32 34 36 38 40 42 time (min)

The target compounds were collected, freed from solvent under reduced pressure (30 ºC;

1.0 x 10-3 bar), and lyophilized. Final purification of 1 and 2 was performed with an isocratic separation using a mixture of 0.1% formic acid in deionized water and 0.1% formic acid in acetonitrile (80:20, v/v). The target compounds were collected, freed from solvent under reduced pressure (30 ºC; 1.0 x 10-3 bar), and lyophilized, yielding 1 (15mg) and 2 (7 mg) as white amorphous powders in high purity (>98%), as determined by LC-

MS (Figure 2.3) and NMR. Final purification of 3 – 5 was performed with an isocratic separation using a mixture of 0.1% formic acid in DI water and 0.1% formic acid in acetonitrile (75:25, v/v). The target compounds were collected, freed from solvent under reduced pressure (30 ºC; 1.0 x 10-3 bar), and lyophilized, yielding 3 (25 mg), 4 (7 mg), and 5 (5 mg) as white amorphous powders in high purity (>98%), as determined by LC-

MS (Figure 2.3) and NMR.

Figure 2.3. (A) Total ion chromatogram (TIC) of crude L. longiflorum n-butanol extract (B – F) Total ion chromatograms (TIC) of 1 – 5 isolated by RP-HPLC.

Intens. 8 A 1 x10 4 2 3 4 2 5 0 B 1 4 2 0 2 3 C 2 1

3 3 D

0 2 E 4

0 F 4 5

0 5 10 15 20 25 30 35 Time [min]

2.3.4. Structural Elucidation

Compounds 1 – 5 (Figure 2.4) were identified by a combination of spectroscopic data (1H NMR, 13C NMR, HMBC, HMQC, MS, IR), chromatographic data, and chemical analysis. Melting points were obtained using a Thomas-Hoover Capillary Melting Point

Apparatus (Arthur H. Thomas Co., Philadelphia, PA) and by differential scanning calorimetry using a Perkin-Elmer Diamond DSC (Perkin-Elmer, Waltham, MA). IR spectra were recorded on a Nexus 670 FT-IR spectrophotometer. Observed rotations were obtained on a Perkin-Elmer model 341LC polarimeter. High -resolution mass spectra

(HRMS) were recorded on a BioTOF II ESI under the following conditions: source temperature, 150 ºC; acceleration voltage, 8500; mass resolution, 10000 fwhm; scan range, m/z 100 – 1000; drying gas, N2. ESI+–MS spectra were recorded on a Bruker 6300 series ion-trap mass spectrometer under the conditions reported above.1D1H NMR and

13CNMR spectra and 2D heteronuclear multiple bond coherence (HMBC) and heteronuclear multiple quantum coherence (HMQC) spectra were acquired on an AMX-

400 spectrometer and an AMX-500 spectrometer (Bruker, Rheinstetten, Germany).

Samples for NMR analysis were dissolved in pyridine-d5, and chemical shifts are given as δ values with reference to tetramethylsilane (TMS).

Figure 2.4. Structures of compounds 1 – 5 isolated from L. longiflorum bulbs.

21 H 25 N 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 OH 2 10 8 R1

4' 3 5 7 6 O 4 1 -D-Glcp R1O O 2' HO 1' 2 6-Ac- -D-Glcp O 1'' H3C O HO OH OH OH O HO O HO OH1'''' 21 OH 26 25 18 23 24 27 20 22 12 17 19 11 13 16 O R2 R3 9 14 15 1 OH 2 10 8 3 -D-Glcp H 4' 3 5 7 -L-Arap O 6 4 H R2O O 4 R O 2' 5 H -D-Xlyp 3 O 1' 1'' H3C O HO OHOH

O OH O OH O O O HO HO O HO HO HO HO HO OH OH OH OH -D-Glcp 6-Ac- -D-Glcp -L-Arap -D-Xlyp

2.3.4.1. Acid Hydrolysis of Compounds 1 – 5.

A solution of each compound (1 mg) in 1NHCl in methanol (0.5 mL) was refluxed at 80 ºC for 2 h. After hydrolysis, the solution was adjusted to pH 7 with NaOH

(4 N) and evaporated to dryness under reduced pressure (30 ºC; 1.0 x 10-3 bar). The residue was dissolved in water (1 mL) and extracted with n-pentane (2 mL). The n- pentane phase was used for aglycone analysis, and the aqueous phase was used for sugar analysis.

2.3.4.2. Aglycone Analysis

The n-pentane phase obtained after hydrolysis was evaporated to dryness under reduced pressure (30 ºC; 1.0 x 10-3 bar), dissolved in a mixture of pyridine and BSTFA with TMCS (99:1) silylating reagent (1:1, v/v, 100 μL), and refluxed in a sealed tube (60

ºC for 1 h). After cooling to room temperature, the solution was analyzed by GC-MS. An

Agilent 6890 series GC system coupled to an Agilent 5973 mass spec detector (Santa

Clara, CA) was used for GC-MS analysis. A capillary column (HP-5, 5% phenyl, 95% dimethyl polysiloxane stationary phase, 30 m x 0.25 mm i.d. x 0.25 mm film thickness) was used for the chromatographic separation. The temperature program was as follows:

70 ºC for 2 min, then increased 8 ºC min-1 to 240 ºC, and held for 10 min. The other parameters used were splitless injector heated at 250 ºC and helium as the carrier gas with a constant flow of 1 mL min-1. MS parameters were as follows: operated in electron

62 impact (EI) ionization mode at 70 eV; scan range, m/z 50 – 550. The transfer line was maintained at 250 ºC. Comparisons were made with retention times and mass spectra of aglycone reference standards prepared according to the same procedure.

2.3.4.3. Sugar Composition Analysis

The aqueous fraction obtained after hydrolysis was evaporated to dryness under reduced pressure (30 ºC; 1.0 x 10-3 bar), dissolved in a mixture of pyridine and BSTFA with TMCS (99:1) silylating reagent (1:1, v/v, 100 μL), and refluxed in a sealed tube (60

ºC for 1 h). After cooling to room temperature, the solution was analyzed by GC-MS.

The temperature program was as follows: 50 ºC for 6 min, then increased at 4 ºC min-1 to

160 ºC, and held for 5 min. The other parameters used were a splitless injector heated to

200 ºC and helium as the carrier gas with a constant flow of 1 mL min-1. MS parameters were the same as reported above. Identifications were made based on retention times and mass spectra of sugar standards prepared according to the same procedure.

2.3.4.4. Determination of Sugar Absolute Configurations

Absolute configuration of sugars was determined by enantioselective GC-FID. A chiral RTBetaDEXsm capillary column (30m x 0.25mm x 0.25 μm; Restek Corp., State

College, PA) was used for chromatographic separation. The temperature program was as

63 follows: 60 ºC for 0 min, then increased at 4 ºC min-1 to 160 ºC, then increased at 15 ºC min-1 to a final temperature 230 ºC, and maintained for 15 min. The other parameters used were as follows: FID detector heated to 230 ºC; split injector with a 10:1 split ratio maintained at 230 ºC; helium as the carrier gas with a constant flow of 1 mL min-1.

Comparisons were made with retention times of optically pure sugar standards prepared following the same procedure.

2.3.4.5. Thin Layer Chromatography (TLC)

Each compound (1 mg) was dissolved in methanol (0.5 mL), spotted on a 20 cm x

20 cm silica gel 60 F254 TLC plate (Merck & Co., Inc., Whitehouse Station, NJ), and developed with chloroform/methanol/water (8:4:1, v/v/v). To detect furostanols, TLC plates were developed with Ehrlich‘s reagent [3.2 g of p-(dimethylamino)benzaldehyde in 60 mL of 95%ethanol and 60 mL of 12 N HCl] and heated to 110 ºC for 5 min. Bright red spots were indicative of a positive reaction. To detect alkaloids, the TLC plates were developed with Dragendorff‘s reagent and heated to 110 ºC for 5 min. Orange spots were indicative of a positive reaction.

2.4. Results and Discussion

2.4.1 Structure Elucidation of Compounds 1 – 5.

Lyophilized lily bulbs were washed with n-pentane and extracted with ethanol and water. After the removal of solvent, the extract was dissolved in deionized water, washed with ethyl acetate and extracted with n-butanol. The organic phase was evaporated under reduced pressure and lyophilized, yielding a crude steroidal glycoside extract. The crude glycoside extract was fractionated by gel permeation chromatography

(Sephadex L-H20) and repeated semi-prep RP-HPLC to yield compounds 1 – 5 (Figure

2.3). Based on 1H NMR, 13C NMR, 2D NMR (HMQC and HMBC), HRESI–TOFMS and chemical analysis, including GC-MS analysis of the sugar and aglycone TMSi derivatives after acid hydrolysis, 1 and 3 were identified as (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, previously isolated from L. brownii (Mimaki and Sashida et al., 1990), and (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -D- glucopyranosyl-(1→4)- -D-glucopyranoside, previously isolated from Dioscorea deltoidea Wall. ex Griseb., Ophiopogon planiscapus Nakai, L. hansonii and Allium nutans L. (Sviridov et al., 1975; Watanabe et al., 1983; Ori et al., 1992; Akhov et al.,

1999 ) (Figure 2.4). This is the first report of these compounds isolated from the bulbs of

L. longiflorum. Compound 2 was characterized as a new acetylated steroidal glycoalkaloid. The structure of compound 2 was determined to be (22R, 25R)-spirosol-5-

65 en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D- glucopyranoside. Compounds 4 and 5 were characterized as new furostanol saponins.

The structure of compound 4 and 5 were determined to be (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- arabinopyranosyl-(1→3)- -D-glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-

(1→3)- -D-glucopyranoside, respectively. This is the first report of these new natural products.

The 1H NMR spectra of steroidal glycoside are complex to analyze due to overlapping proton resonances. For example, the proton resonances of the carbohydrate moiety occur in a narrow spectral width of 3.0 – 4.2 ppm, making assignments difficult.

Despite overlapping signals, there are many diagnostic features in the 1H NMR spectrum.

The methyl peaks of the steroidal backbone are readily distinguishable and informative.

For example, resonances integrating for three protons which are indicative of tertiary methyl groups and secondary methyl groups are diagnostic of the aglycone. Additionally, the 13C NMR spectra have many diagnostic characteristics. For example, olefinic carbon

13 resonances at approximately C 139.7 – 142.1 and 120.9 – 124.3 in the C NMR spectrum are indicative of unsaturation between C-5 and C-6. A quarternary carbon signal at approximately C ~110 ppm suggests a furostane skeleton possessing a hydroxyl group at the C-22 position whereas C ~113 ppm suggests the methoxyl furostane derivative (Agrawal, 1995).

Resonances in the 1H NMR and 13C NMR spectra are useful for the determination of both the number of saccharide residues and anomeric configurations of interglycosidic

66 linkages. Anomeric carbon resonances of the oligosaccharide moiety typically occur in

13 the C 92 – 108 region of the C NMR spectra, thus allowing the number of individual sugar residues to be determined. In the 1H NMR spectra, anomeric proton resonances typically occur as broad singlets or doublets in the range of C 4.2 – 6.4 ppm.

Furthermore, coupling constants for anomeric proton resonances of the C-1 position of - linked sugars have a coupling constant of approximately J = 1.0 – 3.0 Hz and -linked sugars have a coupling constant of approximately J = 6.0 – 8.0 Hz. (Agrawal, 1995).

Although there are many distinguishable and diagnostic features of 1H NMR spectra, overlapping resonances often inhibit complete assignments and do not provide sufficient information for the determination of integlycosidic linkages. With 2-D NMR techniques such as HMBC, often the integlycosidic linkages can be determined; however, the absolute configurations of sugars can not be determined. One approach to determine the absolute configurations of the monosaccharide residues is by acid-catalyzed hydrolysis, derivatization and entantioslective GC-FID analysis. This approach allows for the determination of the absolute configurations of the monosaccharide residues of the carbohydrate moiety.

2.3.4.1. Structure Elucidation of Compound 1

Compound 1 was obtained as a white amorphous powder. The compound was positive to the Dragendorff reaction, indicative of an alkaloid. The IR spectrum showed absorption due to the presence hydroxyl groups at 3400 cm-1. HRESI–TOFMS showed a

+ [M + H] ion at m/z: 884.5028 (calculated for C45H74NO16, 884.5002). Additionally,

++ [M+H+Na] ion was observed at m/z 453.7443 (calculated for C45H74NO16Na,

453.7447) (Figure 2.5). Thus, molecular formula was calculated as C45H73NO16, suggestive of steroidal glycoalkaloid. The aglycone was readily deduced from 1H NMR,

13C NMR (Figure 2.12; Table 2.1), ESI+–MS (Figure 2.33) and chemical analysis. The

1H NMR spectrum showed two singlets at 1.06 and 0.88 which is indicative of tertiary methyl groups of the spirosolane skeleton. Furthermore, two doublet signals at 1.09 and 0.82 were assignable to secondary methyl groups. A quaternary carbon signal at C

13 98.4 and olefinic carbon signals at C 140.8 and 121.8 in the C NMR spectrum were consistent with a 5 spirosolane aglycone. The unsaturation between C-5 and C-6 is further substantiated by a doublet at 5.3 in the 1H NMR spectrum for the H-6 signal.

Figure 2.5. (A) High resolution mass spectrum of compound 1 and (B) expanded view of ++ [M + H + Na] ion at m/z 453.7443 (calculated for C45H74NO16Na, 453.7447). The mass spectrum was acquired on a BioTOF II.

Intens. A

1500 227.2426 453.7443

1250 554.3726

1000 525.3525 583.3928

750

496.3310

129.0494

612.4138 641.4327

500

467.3602 154.1547

250

438.2851 884.5028

670.9643

387.1916

699.9738

369.1752

795.5421

853.5908

911.6248

969.6741

757.5129 1027.7172

0 200 400 600 800 1000 m/z

Intens.

B 453.7443

1250

1000

750 454.2450

500

250 454.7452

0 450 451 452 453 454 455 456 457 458 459 m/z

Upon acid hydrolysis, derivatization and GC-MS analysis, the retention time and mass spectrum of the TMSi derivative of the aglycone was consistent with that of the steroidal alkaloid, (22R, 25R)-spirosol-5-en-3 -ol (solasodine), which was prepared following the same procedure. In 2008, Eanes and Tek reported that during derivitazation of solasodine, two products are formed (Eanes and Tek, 2008). Consistent with their observations, two products were detected by GC-MS for both solasodine and the aglycone of compound 1 (Figure 2.6; Figure 2.7; Figure 2.8). In addition to derivitazation and GC-MS analysis, the retention time and mass spectrum the aglycone of compound 1 was consistent with that of authentic (22R, 25R)-spirosol-5-en-3 -ol analyzed by LC-MS (Figure 2.9).

Figure 2.6. GC-MS chromatogram of the TMSi derivatives of authentic (22R, 25R)- spirosol-5-en-3 -ol standard. Two products are generated during derivatization, labeled A and B.

Abundance B 2200000 2100000 2000000 1900000 1800000 1700000 1600000 1500000 1400000 1300000 1200000 1100000 1000000 900000 800000 700000 600000 A 500000 400000 300000 200000 100000 16.00 17.00 18.00 19.00 20.00 21.00 22.00 23.00 24.00 25.00 26.00 27.00 28.00 29.00 Time-->

Figure 2.7. GCMS mass spectra of TMSi derivatives, A and B, generated from authentic (22R, 25R)-spirosol-5-en-3 -ol standard.

Abundance A 111 75000 70000 65000 60000 55000 50000 45000 40000 35000 30000 25000 73 20000 15000 10000 138 557 162 452 5000 190 347 412 41 214 238 281 384 484 528 0 m/z--> 50 100 150 200 250 300 350 400 450 500 550

Abundance B 125 900000

800000

700000

600000

500000

400000

300000

200000 73 100000 148 452 41 96 181 207 238 308 362 394 484 542 0 m/z--> 50 100 150 200 250 300 350 400 450 500 550

Figure 2.8. GCMS mass spectra of TMSi derivatives, A and B, generated from the aglycone of compound 1.

Abundance 111 30000 28000 26000 24000 22000 20000 18000 16000 14000 12000 10000 8000 73 6000 4000 138 557 162 452 2000 190 412 41 238 281 484 528 0 m/z--> 50 100 150 200 250 300 350 400 450 500 550

Abundance 125 380000

360000

340000

320000

300000

280000

260000

240000

220000 73 452 542 41 96 148 208 238 309 394 484 0 m/z--> 50 100 150 200 250 300 350 400 450 500 550

Figure 2.9. Total ion chromatogram (TIC) of (A) authentic (22R, 25R)-spirosol-5-en-3 - ol standard (Rt = 32.1 min) and (B) the aglycone of compound 1 (Rt = 32.1 min) generated by LC-MS.

Intens. x10 8 A 3

0 5 10 15 20 25 30 35 Time [min]

Intens. B x10 8 5

0 5 10 15 20 25 30 35 Time [min]

The structure of the oligosaccharide moiety was readily deduced from 1H NMR, 13C

NMR, ESI+–MS, and chemical analysis. Three anomeric protons were observed at

4.97, 6.27, and 5.15, which implied the presence of three saccharide residues. Coupling constants of the anomeric proton resonances suggested -interglycosidic linkages.

Additionally, the 13C NMR spectrum contained three anomeric carbon signals observed at

C 105.3, 1 and consistent with the presence of three saccharide residues.

The HMBC experiment showed long-range correlations between the anomeric proton signal at 4.97 [H-1′] and the carbon signal at C 78.1 [C-3], between the anomeric proton signal at 5.15 [H-1′′′] and the carbon signal at C 82.1 [C-4′], and the anomeric proton signal at 6.27 [H-1′′] and the carbon signal at C 77.8 [C-2′] (Figure 2.10).

Figure 2.10. HMBC long-range correlations for the interglycosidic linkages for the carbohydrate moiety of compound 1.

OH 82.1 OH O H H 78.1 HO O HO O O OH H HO 77.8 5.15 (7.6) O H 4.97 (7.2) 1 H3C O H 6.27 HO OH OH

Upon acid hydrolysis, derivatization and GC-MS/entantioslective GC-FID analysis, the sugars of the trisaccharide moiety were identified as D-(+)-glucose and L-(–)-rhamnose in a 2 to 1 ratio.1H NMR spectrum showed a doublet integrating for three protons at 1.78, which is indicative of the methyl group of rhamnose. The ESI+–MS mass spectrum showed the molecular ion 884.7 [M+H]+ and the doublely charged sodium adduct at

453.8 [M+H+Na]++ . Additionally, ion fragments at m/z 738.5 [M–Rha+H]+, 576.4 [M–

Glu–Rha+H]+, and 414.4 [M–2Glu–Rha+H]+ were observed and were consistent with a

trisaccharide moiety containing two D-(+)-glucoses and a L-(–)-rhamnose moiety (Figure

2.11).

Figure 2.11. ESI+-MS mass spectrum of compound 1.

Intens. 6 x10 884.7

1.5

1.0

0.5 576.4 414.4

253.1 453.8 738.5 0.0 200 400 600 800 1000 1200 1400 m/z

Accordingly, the structure of 1 was determined to be (22R, 25R)-spirosol-5-en-3 -yl O-

-L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, previously reported in L. brownii (Mimaki and Sashida, 1990b).

Figure 2.12. (A) 1H NMR spectrum and (B) 13C NMR spectrum of compound 1.

Compound 1, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D-

28 glucopyranosyl-(1→4)- -D-glucopyranoside: amorphous solid; [ D - 76.9 (c 0.02;

-1 MeOH); mp 303 ºC (dec); IR max (film) cm : 3336 (OH), 2930 (CH), 1585, 1450, 1347,

1255, 1029, 984, 897, 811; HRESI–TOFMS, m/z 884.5028 [M + H]+ (calculated for

+ + + C45H74NO16, 884.5002); ESI – MS, m/z 884.7 (100, [M+H] ), 738.5 (3, [M–Rha+H] ),

576.4 (19, [M–Glu–Rha+H]+), 453.8 (1.6, [M+H+Na]++ ), 414.4 (15, [M–2Glu–Rha+H]+

); 1H NMR (400 MHz) 0.82 [d, 3H, J = 5.2 Hz, 27-H], 0.88 [s, 3H, 18-H], 1.06 [s, 3H,

19-H], 1.09 [d, 3H, J = 6.8 Hz, 21-H], 1.78 [d, 3H, J = 6.4 Hz, 6′′-H], 2.75 [m, 2H, 26-

H], 3.88 [m, 1H, 3-H], 3.90 [m, 1H, 2′-H], 4.33 [1H, 6b′′′-H], 4.56-4.40 [1H, 6a′′′-H],

4.48 [m, 2H, 6′-H], 4.61 [dd, 1H, J =9.2, 3.2, 3′′-H], 4.97 [d, 1H, J = 7.2 Hz, 1′-H], 5.15

[d, 1H, J =7.6 Hz, 1′′′-H], 5.30 [d, 1H, J = 4.8 Hz, 6-H], 6.27 [s, 1H, 1′′-H]; 13C NMR

(400 MHz, pyridine-d5) 37.5 [C-1], 30.2 [C-2], 78.1 [C-3], 38.9 [C-4], 140.8 [C-5],

121.8 [C-6], 32.4 [C-7], 31.6 [C-8], 50.3 [C-9], 37.2 [C-10], 21.2 [C-11], 40.1 [C-12],

40.7 [C-13], 56.7 [C-14], 32.6 [C-15], 78.8 [C-16], 63.6 [C-17], 16.5 [C-18], 19.4 [C-19],

41.6 [C-20], 15.7 [C-21], 98.4 [C-22], 34.7 [C-23], 31.1 [C-24], 31.7 [C-25], 48.1 [C-26],

19.9 [C-27], 100.0 [C-1′], 77.8 [C-2′], 76.2 [C-3′], 82.1 [C-4′], 77.3 [C-5′], 62.0 [C-6′],

101.8 [C-1′′], 72.5 [C-2′′], 72.8 [C-3′′], 74.2 [C-4′′], 69.5 [C-5′′], 18.7 [C-6′′], 105.3 [C-

1′′′], 75.0 [C-2′′′], 78.3 [C-3′′′], 71.2 [C-4′′′], 78.5 [C-5′′′], 62.1 [C-6′′′]; 1H NMR and 13C

NMR are consistent with the literature (Mimaki and Sashida, 1990b).

2.3.4.2. Structure Elucidation of Compound 2

Compound 2 was obtained as a white amorphous powder. The compound was positive to the Dragendorff reaction, indicative of an alkaloid (Mimaki and Sashida,

1990). The IR spectrum showed absorption at 1732 cm-1 and 3353 cm-1due were due to the presence of an acetyl group and hydroxyl groups. HRESI–TOFMS showed a [M+H]+

++ ion at m/z: 926.5085 (calculated for C47H76NO17, 926.5108). Additionally, [M+H+Na] ion was observed at m/z 474.7520 (calculated for C47H76NO17Na, 474.7500) (Figure

2.13). Thus, the molecular formula was calculated as C47H75NO17, suggestive of an acetylated steroidal glycoalkaloid. The aglycone was readily deduced from 1H NMR, 13C

NMR (Figure 2.16; Table 2.1), ESI+–MS (Figure 2.34), and chemical analysis. The 1H

NMR spectrum showed two singlets at 1.06 and 0.88 which is indicative of tertiary methyl groups of the spirosolane skeleton. Furthermore, two doublet signals at 1.11 and 0.82 were assignable to secondary methyl groups. A quaternary carbon signal at

13 C 98.4 and olefinic carbon signals at C 140.8 and 121.9 in the C NMR spectrum were consistent with a 5 spirosolane aglycone. The unsaturation between C-5 and C-6 was further substantiated by a doublet at 5.29 in the 1H NMR spectrum for the H-6 signal.

Figure 2.13. (A) High resolution mass spectrum of compound 2 and (B) expanded view ++ of [M + H + Na] ion at m/z 474.7520 (calculated for C47H76NO17Na, 474.7500). The mass spectrum was acquired on a BioTOF II.

Intens.

A 474.7520

5000

4000

3000

2000

1000

129.1164

554.3704

583.3934

227.2858 525.3528

496.3273

612.4120

641.4322

926.5635

393.3024

542.7395

662.3903 438.2871

0 200 400 600 800 1000 m/z

Intens.

B 474.7520

5000

4000

3000 475.2539

2000

1000

475.7533 476.2577

0 472 473 474 475 476 477 478 479 480 m/z

Upon acid hydrolysis, derivatization and GC-MS analysis, the retention time and mass spectrum of the TMSi derivative of the aglycone was consistent with that of (22R, 25R)- spirosol-5-en-3 -ol, which was prepared following the same procedure. The structure of the oligosaccharide moiety was readily deduced from 1H NMR, 13C NMR, HMBC, ESI+–

MS and chemical analysis. Three anomeric protons were observed at 4.98, 6.24 and

5.09, which implied the presence of three saccharide residues. Coupling constants for the anomeric proton resonances suggested -interglycosidic linkages. Additionally, the 13C

NMR spectrum contained three anomeric carbon signals observed at C 105.6, and consistent with the presence of three saccharide residues. The HMBC experiment showed long-range correlations between the anomeric proton signal at

4.98 [H-1′] and the carbon signal at C 78.1 [C-3], between the anomeric proton signal at

5.09 [H-1′′′] and the carbon signal at C 83.3 [C-4′], and the anomeric proton signal at 6.24 [H-1′′] and the carbon signal at C 77.6 [C-2′] (Figure 2.14).

Figure 2.14. HMBC long-range correlations for the interglycosidic linkages for the carbohydrate moiety of compound 2.

O 83.3 OH O H H 78.1 HO O HO O O OH H HO 77.6 5.09 (8) O H 4.98 (7.2) 2 H3C O H 6.24 HO OH OH

1.78, indicative of the methyl group of rhamnose. The ESI+–MS mass spectrum showed the protonated molecular ion 926.6 [M+H]+ and the protonated double charged sodium adduct at 474.8 [M+H+Na]++. Additionally, ion fragments at m/z 780.5 [M–Rha+H]+,

576.4 [M–Glu–Ac– Rha+H]+ and 414.3 [M–2Glu–Ac–Rha+H]+ were observed and were consistent with a trisaccharide moiety containing D-(+)-glucose, an acetylated D-(+)- glucose and a L-(-)-rhamnose moiety (Figure 2.15).

Figure 2.15. ESI+-MS mass spectrum of compound 2.

Intens. 6 x10 926.7

414.3 2 474.8 780.5 576.4 253.2 0 200 400 600 800 1000 1200 m/z

The presence of an acetyl group was shown by the IR absorption at1732 cm-1, H NMR

13 2.06 (s, 3H, CO-CH3) and C NMR C 170.9 and 20.8. The carbon signals corresponding to the saccharide moiety of 2 were similar to those reported for (25R,26R)-

26-methoxyspirost-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside, isolated from L. speciosum x L. nobilissimum (Nakamura et al., 1994). Alkaline hydrolysis of 2 with 1M sodium hydroxide yielded 1. Upon comparison of the 13C NMR spectra of 1 and 2, the C-6′′′ and

C-4′′′ signals were shifted from C 62.1 and 71.2, to C 64.8 and 71.9, respectively. The signal for C-5′′′ was shifted from C 78.5 to 74.9 and all other carbon signals were

1 similar. Upon comparison of the H NMR spectra of 1 and 2, the signals assignable to H2-

6′′′ methylene protons of the terminal glucose were shifted to a lower field as compared to those of 2. Thus, the acetyl moiety was linked to the C-6 hydroxy position of the terminal glucose unit. Accordingly, the structure of 2 was determined to be (22R, 25R)- spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-

(1→4)]- -D-glucopyranoside (Figure 2.4).

Figure 2.16. (A) 1H NMR spectrum and (B) 13C NMR spectrum of compound 2.

Compound 2, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-

28 acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside: amorphous solid; [ D -

-1 32.0º (MeOH; c 0.05); mp 285 ºC (dec); IR max (film) cm : 3353 (OH), 2932 (CH),

1732 (C=O), 1588, 1451, 1370, 1252, 1033, 983, 897, 813; HRESI–TOFMS m/z:

+ 926.5085 [M+H]+ (calculated for C47H76NO17, 926.5108); ESI –MS, m/z 926.6 (100,

[M+H]+), 780.5 (14, [M–Rha+H]+), 576.4 (15, [M–Glu–Ac–Rha+H]+), 474.8 (3,

[M+H+Na]++ ), 414.3 (30, [M–2Glu–Ac–Rha+H]+); 1H NMR (400 MHz) 0.82 [d, 3H,

J = 4.8 Hz, 27-H], 0.88 [s, 3H, 18-H], 1.06 [s, 3H, 19-H], 1.11 [d, 3H, J = 6.4 Hz, 21-H],

1.78 [d, 3H, J = 6 Hz, 6′′-H], 2.06 [s, 3H, CO-CH3], 4.35 [t, 1H, J =9.4, 4′′-H], 4.60 [dd,

1H, J =8.8, 3.2, 3′′-H], 4.64 [d, 1H, J = 8.4 Hz, 6b′′′-H], 4.92 [dd, 1H, J =11.6, 2, 6a′′′-H],

4.98 [d, 1H, J = 7.2 Hz, 1′-H], 5.09 [d, 1H, J =8.1′′′-H], 5.29 [d, 1H, J = 4.4 Hz, 6-H],

13 6.24 [br s, 1H, 1′′-H]; C NMR (500 MHz, pyridine-d5) 37.5 [C-1], 30.2 [C-2], 78.1 [C-

3], 39.0 [C-4], 140.8 [C-5], 121.9 [C-6], 32.4 [C-7], 31.6 [C-8], 50.4 [C-9], 37.2 [C-10],

21.2 [C-11], 40.1 [C-12], 40.7 [C-13], 56.7 [C-14], 32.6 [C-15], 79.0 [C-16], 63.5 [C-17],

16.5 [C-18], 19.4 [C-19], 41.6 [C-20], 15.7 [C-21], 98.4 [C-22], 34.7 [C-23], 31.0 [C-24],

31.7 [C-25], 48.0 [C-26], 19.8 [C-27], 99.9 [C-1‘], 77.6 [C-2‘], 76.1 [C-3‘], 83.3 [C-4‘],

77.4 [C-5‘], 62.0 [C-6‘], 102.0 [C-1‘‘], 72.5 [C-2‘‘], 72.8 [C-3‘‘], 74.2 [C-4‘‘], 69.6 [C-

5‘‘], 18.7 [C-6‘‘], 105.6 [C-1‘‘‘], 75.1 [C-2‘‘‘], 78.2 [C-3‘‘‘], 71.9 [C-4‘‘‘], 74.9 [C-5‘‘‘],

64.8 [C-6‘‘‘], 20.8 [Ac-CH3], 170.9 [AcC=O].

2.3.4.3. Structure Elucidation of Compound 3

Compound 3 was obtained as a white an amorphous powder. The compound was positive to the Ehrlich‘s reaction, indicative of a furostanol saponin (Yoshiaki et al.,

1983). The IR spectrum showed absorption at 3400 cm-1due to the presence of hydroxyl groups. HRESI–TOFMS showed a [M + Na]+ ion at m/z: 1087.5307 (calculated for

- C51H84O23Na, 1087.5296 ) (Figure 2.17). Additionally, [M–H] ion was observed at m/z

1063.6 (Figure 2.18). Thus, the molecular formula was calculated as C51H84O23, consistent with a furostanol saponin. The aglycone was readily deduced from 1H NMR,

13C NMR (Figure 2.22; Table 2.1), ESI+–MS (Figure 2.34), HMBC, and chemical analysis. The 1H NMR spectrum showed two singlets at 1.06 and 0.90 which is indicative of tertiary methyl groups of the furostane skeleton. Furthermore, two doublet signals at 1.35 and 1.0 were assignable to secondary methyl groups. A quarternary carbon signal at C 110.7 suggests a furostane skeleton possessing a hydroxyl group at the C-22 position. (25R)-26-O- -D-glucopyranosyl-22 -methoxy-furost-5-en-3 26-triol

3-O- -L-rhamnopyranosyl-(1→2- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, previously identified in L. longiflorum possessing antitumor activity, contains a methoxy group at the C-22 position with a OCH3 signal at C 47.73 which is missing in 3. Also, the quaternary carbon signal for C-22 had a minor upfield shift to C 110.7 instead of the reported value of C 112.7 for the C-22 methoxy derivative (Mimaki et al., 1994).

Figure 2.17. (A) High resolution mass spectrum of compound 3 and (B) expanded view + of [M + Na] ion at m/z: 1087.5307 (calculated for C51H84O23Na, 1087.5296 ). The mass spectrum was acquired on a BioTOF II.

Intens.

911.6285

555.2711 853.5872

800

969.6663

1087.5307 795.5470

600

1027.7122

393.3144 583.3995

525.3602 612.4196 496.3204

400 641.4386

229.1715

737.5028

309.2261

670.4644

699.4771 1143.7958

200 467.3195

427.2232 371.1137

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Intens.

B 1087.5307

600

400

1088.5343

1085.7554

1088.0263 1086.7567

200 1089.5368

0 1084 1086 1088 1090 1092 1094 m/z

Figure 2.18. LRMS- mass spectrum of compound 3 acquired on BioTOF II.

Intens.

600 212.1

1063.6

145.0 531.3

400

200

1083.6 171.1

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

C-22 hydroxy furostanols are readily converted to C-22 methoxy derivatives upon refluxing in methanol (Watanabe et al., 1983; Wang et al., 2003). It has been suggested that C-22 methoxy furostanol saponins identified in plants may be artifacts formed during the extraction process when methanol is used as a solvent (Oleszek and Bialy, 2006).

The carbon peaks from C-5 to C-27 were similar to those reported for (25R,S)-26-

O- -D-glucopyranosyl-furost-5-en-3 22 -triol 3-O- -D-galactopyranosyl-(1→2)-O-

-D-glucopyranosyl-(1→4)- -D-galactopyranoside, isolated from Tribulus terrestris L., which suggests the presence of an α-hydroxy group at the C-22 position of the furostane

13 skeleton (Wang et al., 1997). Olefinic carbon signals at C 140.8 and 121.9 in the C

NMR spectrum are consistent with a 5 furostane skeleton. The unsaturation between C-5 and C-6 is further substantiated by a doublet at 5.29 in the 1H NMR spectrum for the H-

6 signal. Upon acid hydrolysis, derivatization and GC-MS analysis, the retention time and mass spectrum of the TMSi derivative of the aglycone was consistent with that of

(25R)-spirost-5-en-3 -ol, which was prepared following the same procedure (Figure

2.19). The structure of the oligosaccharide moiety was readily deduced from 1H NMR,

13C NMR, HMBC, ESI–MS+ and chemical analysis. The 1H NMR spectrum contained four anomeric proton signals observed at 6.26, 4.96, and . The coupling constants of the anomeric proton resonances suggested -interglycosidic linkages. The

13 C NMR spectrum contained four anomeric carbon signals observed at C 105.2, 105.0,

and consistent with the presence of four saccharide residues.

Figure 2.19. GCMS spectra of (A) TMSi derivative of the aglycone of compound 3 and (B) TMSi derivative of (25R)-spirost-5-en-3 -ol standard.

Abundance 139 A 600000 550000 500000 450000 400000 350000 282 300000 250000 73 187 200000 150000 243 372 119 100000 93 159 414 213 343 50000 41 471 0 m/z--> 50 100 150 200 250 300 350 400 450

Abundance B 28000 139 26000 24000 22000 20000 18000 282 16000 14000 12000 73 10000 187 8000 6000 119 243 372 95 4000 159 207 414 41 343 2000 471 0 m/z--> 50 100 150 200 250 300 350 400 450

The HMBC experiment showed long-range correlations between the anomeric proton signal at 4.96 [H-1′] and the carbon signal at C 78.2 [C-3], between the anomeric proton signal at 5.14 [H-1′′′] and the carbon signal at C 82.1 [C-4′], and the anomeric proton signal at 6.26 [H-1′′] and the carbon signal at C 77.8 [C-2′] (Figure 2.20).

Figure 2.20. HMBC long-range correlations for the interglycosidic linkages for the carbohydrate moiety of compound 3.

OH 82.1 OH O H H 78.2 HO O HO O O OH H HO 77.8 5.14 (8) O H 4.96 (7.2) 3 H3C O H 6.26 HO OH OH

Upon acid hydrolysis, derivitazation and GC-MS/entantioslective GC-FID analysis, the sugars were identified as D-(+)-glucose and L-(-)-rhamnose in a 3:1 ratio. 1H NMR spectrum showed a doublet integrating for three protons at 1.77, which is indicative of the methyl group of rhamnose. The ESI+–MS mass spectrum showed a base ion peak at

1047.7 [M-18+H]+ and the sodium adduct at 1087.7 [M + H+Na]+. Additionally, ion fragments at m/z 901.5 [M-18-Rha+H]+, 739.4 [ M-18-Glu-Rha+H]+, 577.4 [M-18-

2Glu-Rha+H]+, and 415.3 [M-18-3Glu-Rha+H]+ were observed and are consistent with 3 being bidesmodic with the trisaccharide moiety at the C3 position containing two

90 glucoses and one rhamnose, and a glucose moiety at the C26 position, indicative of a furostanol saponin (Figure 2.21).

Figure 2.21. ESI+–MS mass spectrum of compound 3.

Intens. 6 x10 1047.7

1 1087.7

739.4 253.1 415.3 577.4 901.5 0 200 400 600 800 1000 1200 1400 m/z

Accordingly, the structure of 3 was determined to be (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-

(1→4)- -D-glucopyranoside, previously reported from D. deltoidea, O. planiscapus, L. hansonii, and A. nutans (Sviridov et al., 1975; Watanabe et al., 1983; Ori et al., 1992;

Akhov et al., 1999) (Figure 2.4).

Figure 2.22. (A) 1H NMR spectrum and (B) 13C NMR spectrum of compound 3.

Compound 3, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside: amorphous

28 -1 solid; [ D - 60.0 (c 0.05; MeOH); mp 207 ºC (dec): IR max (film) cm : 3347 (OH),

2893 (CH), 1662, 1377, 1256, 1027, 909, 839, 812; HRESI–TOFMS m/z: 1087.5296

+ + [M+Na] (calculated for C51H84O23Na, 1087.5307); ESI – MS, m/z 1087.7 (20,

[M+Na]+), 1047.7 (100, [M–18+H]+), 901.5 (9, [M–18–Rha+H]+), 739.4 (17, [M–18–

Glu–Rha+H]+), 577.4 (11, [M–18–2Glu–Rha+H]+), 415.3 (9, [M–18–3Glu–Rha+H]+);

1H NMR (500 MHz) 0.91 [s, 3H, 18-H], 1.00 [d, 3H, J = 6.4 Hz, 27-H], 1.07 [s, 3H,

19-H], 1.35 [d, 3H, J = 6.8 Hz, 21-H], 1.76 [d, 3H, J = 6.5 Hz, 6′′-H], 4.56 [dd, 1H, J

=12, 2.5, 6′′′′-H], 4.60 [dd, 1H, J =9.5, 3.5, 3′′-H], 4.83 [d, 1H, J = 8 Hz, 1′′′′-H], 4.96 [d,

1H, J = 6.5 Hz, 1′-H], 5.14 [d, 1H, J =8, 1′′′-H], 5.29 [d, 1H, J = 4 Hz, 6-H], 6.26 [br s,

13 1H, 1′′-H]; C NMR (400 MHz, pyridine-d5) 37.5 [C-1], 30.2 [C-2], 78.2 [C-3], 39.0 [C-

4], 140.8 [C-5], 121.9 [C-6], 32.5 [C-7], 31.7 [C-8], 50.4 [C-9], 37.2 [C-10], 21.2 [C-11],

39.9 [C-12], 40.8 [C-13], 56.6 [C-14], 32.4 [C-15], 81.1 [C-16], 63.9 [C-17], 16.5 [C-18],

19.4 [C-19], 40.7 [C-20], 16.5 [C-21], 110.7 [C-22], 37.1 [C-23], 28.4 [C-24], 34.3 [C-

25], 75.3 [C-26], 17.5 [C-27], 100.0 [C-1′], 77.8 [C-2′], 76.2 [C-3′], 82.1 [C-4′], 77.3 [C-

5′], 62.1 [C-6′], 101.8 [C-1′], 72.5 [C-2′′], 72.8 [C-3′′], 74.2 [C-4′′], 69.5 [C-5′′], 18.7 [C-

6′′], 105.2 [C-1′′′], 75.0 [C-2′′′], 78.5 [C-3′′′], 71.2 [C-4′′′], 78.3 [C-5′′′], 61.9 [C-6′′′],

105.0 [C-1′′′′], 75.2 [C-2′′′′], 78.6 [C-3′′′′], 71.7 [C-4′′′′], 78.2 [C-5′′′′], 62.8 [C-6′′′′]; 1H

NMR and 13C NMR are consistent with the literature (51).

2.3.4.3. Structure Elucidation of Compound 4

Compound 4 was obtained as a white amorphous powder. The compound was positive to the Ehrlich‘s reaction, indicative of a furostanol saponin (Yoshiaki et al.,

1983). The IR spectrum showed absorption at 3367 cm-1due to the presence of hydroxyl groups. HRESI–TOFMS showed a [M+Na]+ ion at m/z: 1057.5211 (calculated for

- C50H82O22Na, 1057.5190) (Figure 2.23). Additionally, [M–H] ion was observed at m/z

1033.6 (Figure 2.24). Thus, the molecular formula was calculated as C50H82O22, consistent with a furostanol saponin. The aglycone was readily deduced from 1H NMR,

13C NMR (Figure 2.28; Table 2.1), HMBC, ESI+–MS (Figure 2.34), and chemical analysis. The 1H NMR spectrum showed two singlets at 1.07 and 0.91 which is indicative of the tertiary methyl groups of the furostane skeleton. Furthermore, two doublet signals at 1.35 and 1.0 were assignable to secondary methyl groups. (25R)-26-

O- -D-glucopyranosyl-22 -methoxy-furost-5-en-3 26-diol 3-O- -L-rhamnopyranosyl-

(1→2)- -L-arabinopyranosyl-(1→3)- -D-glucopyranoside, previously identified in L. longiflorum possessing antitumor activity, contains a methoxy group at the C-22 position with a OCH3 signal at C 47.73 which is missing in 4. Also, the quaternary carbon signal for C-22 had a minor upfield shift to C 110.7 instead of the reported value of C 112.7 for the C-22 methoxy derivative (Mimaki et al., 1994).

Figure 2.23. (A) High resolution mass spectrum of compound 4 and (B) expanded view + of [M + Na] ion at m/z: 1057.5211 (calculated for C50H82O22Na, 1057.5190). The mass spectrum was acquired on a BioTOF II.

Intens. A

1500 393.3182

1250

1000 309.2306 229.1744

750

1057.5211 540.2664

500

583.4028 612.4229

911.6277

641.4428

853.5895

795.5392

496.3441 969.6706

250

699.4792

670.4661

763.6089

467.3260 1027.7128

737.5051

438.3034

1085.7376

1143.7951 1201.8376

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Intens.

B 1057.5211

600

400 1058.5265

200

1058.0141 1059.5273

0 1056 1058 1060 1062 1064 m/z

Figure 2.24. LRMS- mass spectrum of compound 4 acquired on BioTOF II.

Intens.

1500 212.1

1250

1000

750 1033.6

500

171.1 516.2

250 143.1

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Long range coupling was observed between the methyl proton signal at H-21] and the carbon signals at C 40.7 [C-20], 63.9 [C-17] and 110.7 [C-22], supporting that the shift seen in this region is consistent with a C-22 hydroxy compared to the C-22 methoxy derivative with the reported values of C 40.5 [C-20], 64.2 [C-17] and 112.7 [C-

22] (Mimaki et al., 1994) (Figure 2.25). The carbon peaks from C-5 to C-27 were similar to those reported for (25R,S)-26-O- -D-glucopyranosyl-furost-5-en-3 22 -triol 3-O-

-D-galactopyranosyl-(1→2)-O- -D-glucopyranosyl-(1→4)- -D-galactopyranoside, isolated from Tribulus terrestris L., which suggests the presence of an α-hydroxy group at the C-22 position of the furostane skeleton (Wang et al., 1997).

Figure 2.25. Partial HMBC spectrum of compound 4, showing the correlation between H-21 and the carbon signals of C-20, C17 and C-22.

C-20 40

C-17 60 70

100

C-22 110 PPM F1 PPM F2 2.00 1.8 1.60 1.40 1.2 1.0 0.80 0 0 0

1.35 H21

OH 40.7 20 22 110.7 63.9 17 O

Consistent with (25R)-26-O- -D-glucopyranosyl-furost-5-en-3 22 triol 3-O- -D- glucopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, isolated from the rhizomes of Tupistra chinensis Bak., the peak of C-25 at C 34.3 as compared to C

34.42, suggests an R configuration (Wang et al., 1992). Olefinic carbon signals at C

140.7 and 121.9 in the 13C NMR spectrum were consistent with a 5 furostane skeleton.

The unsaturation between C-5 and C-6 is further substantiated by an olefinic proton signal at 5.33 in the 1H NMR spectrum for H-6. Upon acid hydrolysis, derivatization and GC-MS analysis, the retention time and mass spectrum of the TMSi derivative of the aglycone was consistent with that of (25R)-spirost-5-en-3 -ol, which was prepared following the same procedure. The structure of the oligosaccharide moiety was readily deduced from 1H NMR, 13C NMR, HMBC, ESI+–MS and chemical analysis. All four sugars of 4 have similar carbon NMR values to those reported for (25R)-26-O- -D- glucopyranosyl-22 -methoxy-furost-5-en-3 26-diol 3-O- -L-rhamnopyranosyl-(1→2)-

-L-arabinopyranosyl-(1→3)- -D-glucopyranoside (Mimaki et al., 1994). The 1H NMR spectrum contained four anomeric proton signals observed at 6.30, 4.99, 4.92 and

4.84. The coupling constants of the anomeric proton resonances suggested - interglycosidic linkages. The carbon signals were at C 105.6, 105.0, 102.5 and 99.9, which is consistent with the presence of four saccharide residues. The HMBC experiment showed long-range correlations between the anomeric proton signal at 4.99 [H-1′] and the carbon signal at C 77.7 [C-3], between the anomeric proton signal at 4.92 [H-1′′′] and the carbon signal at C 88 [C-3′], and the anomeric proton signal at 6.30 [H-1′′] and the carbon signal at C 78 [C-2′] (Figure 2.26).

Figure 2.26. HMBC long-range correlations for the interglycosidic linkages for the carbohydrate moiety of compound 4.

OH OH H 77.7 O O HO O 78 HO O H OHH 88 O 4.99 (7.2) 4.92 (7.6)

4 H 6.30 H3C O HO OH OH

Upon acid hydrolysis, derivitazation and GC-MS/entantioslective GC-FID analysis, the sugars were identified as D-(+)-glucose, L-(–)-rhamnose and L-(–)-arabinose in a 2:1:1 ratio. 1H NMR spectrum showed a doublet integrating for three protons at 1.76, indicative of the methyl group of rhamnose. The ESI+–MS mass spectrum showed a base ion peak at 1017.7 [M–18+H]+ and the sodium adduct at 1057.7 [M+Na]+ . Additionally, ion fragments at m/z 871.6 [M–18–Rha+H]+, 739.4 [ M–18–Ara–Rha+H]+, 577.2 [M–

18–Ara–Rha–Glu+H]+ and 415.2 [M–18–2Glu–Ara–Rha+H]+ were observed and were consistent with 4 being bidesmodic with the trisaccharide moiety at the C-3 position containing D-(+)-glucose, L-(–)-arabinose and L-(–)-rhamnose, and a D-(+)-glucose moiety at the C-26 position, indicative of a furostanol saponin (Figure 2.27).

Figure 2.27. ESI+–MS mass spectrum of compound 4.

Intens. 1017.7 6 x10

2 1057.7

1 271.1 415.3 577.2 739.4 871.6 0 200 400 600 800 1000 1200 m/z

Accordingly, the structure of 4 was determined to be (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-

(1→3)- -D-glucopyranoside (Figure 2.4).

100

Figure 2.28. (A) 1H NMR spectrum and (B) 13C NMR spectrum of compound 4.

101

Compound 4, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-(1→3)- -D-glucopyranoside:

28 -1 amorphous solid; [ D - 48.6º (MeOH; c 0.07); mp 200 ºC (dec); IR max (film) cm :

3367 (OH), 2899 (CH), 1699, 1377, 1256, 1040, 912, 840, 811, 780; HRESI–TOFMS

+ + m/z: 1057.5211 [M+Na] (calculated for C50H82O22Na, 1057.5190); ESI – MS, m/z

1057.7 (5, [M+Na]+), 1017.7 (100, [M–18+H]+), 871.6 (7, [M–18–Rha+H]+), 739.4 (13,

[M–18–Ara–Rha+H]+), 577.2 (7, [M–18–Ara–Rha–Glu + H]+), 415.2 (4, [M–18–2Glu–

Ara–Rha+H]+); 1H NMR (400 MHz) 0.91 [s, 3H, 18-H], 1.00 [d, 3H, J = 6.4 Hz, 27-

H], 1.07 [s, 3H, 19-H], 1.35 [d, 3H, J = 6.8 Hz, 21-H], 1.76 [d, 3H, J = 6 Hz, 6′′-H], 3.08

[d, 1H, J = 11.6, 5′′′-H], 4.05 [m, 1H, Hz, 2′′′′-H], 3.90-3.85 [m, 1H, 2′-H], 4.84 [d, 1H, J

= 7.6 Hz, 1′′′′-H], 4.91 [m, 1H, 16-H], 4.92 [d, 1H, J = 7.6, 1′′′-H], 4.99 [d, 1H, J = 7.2

Hz, 1′-H], 5.33 [d, 1H, J = 4 Hz, 6-H], 6.30 [br s, 1H, 1′′-H]; for 13C NMR (400 MHz, pyridine-d5) 37.5 [C-1], 30.1 [C-2], 77.7 [C-3], 38.7 [C-4], 140.7 [C-5], 121.9 [C-6], 32.4

[C-7], 31.7 [C-8], 50.3 [C-9], 37.2 [C-10], 21.1 [C-11], 39.9 [C-12], 40.8 [C-13], 56.6 [C-

14], 32.5 [C-15], 81.1 [C-16], 63.9 [C-17], 16.5 [C-18], 19.4 [C-19], 40.7 [C-20], 16.5

[C-21], 110.7 [C-22], 37.2 [C-23], 28.4 [C-24], 34.3 [C-25], 75.2 [C-26], 17.5 [C-27],

99.9 [C-1′], 78.0 [C-2′], 88.0 [C-3′], 69.7 [C-4′], 77.6 [C-5′], 62.4 [C-6′], 102.5 [C-1′′],

72.5 [C-2′′], 72.9 [C-3′′], 74.4 [C-4′′], 69.5 [C-5′′], 18.7 [C-6′′], 105.6 [C-1′′′], 72.3 [C-

2′′′], 74.6 [C-3′′′], 69.7 [C-4′′′], 67.8 [C-5′′′], 105.0 [C-1′′′′], 75.3 [C-2′′′′], 78.6 [C-3′′′′],

71.7 [C-4′′′′], 78.5 [C-5′′′′], 62.8 [C-6′′′′].

102

2.3.4.4. Structure Elucidation of Compound 5

Compound 5 was obtained as a white amorphous powder. The compound was positive to the Ehrlich‘s reaction, indicative of a furostanol saponin. The IR spectrum showed absorption at 3362 cm-1due to the presence of hydroxyl groups. HRESI–TOFMS

+ showed a [M+Na] ion at m/z: 1057.5211 (calculated for C50H82O22Na, 1057.5190)

(Figure 2.29). Additionally, [M–H]- ion was observed at m/z 1033.6 (Figure 2.30). Thus, the molecular formula was calculated as C50H82O22, consistent with a furostanol saponin.

The aglycone was readily deduced from 1H NMR, 13C NMR (Figure 2.33; Table 2.1),

HMBC, ESI+–MS (Figure 2.34), and chemical analysis. The 1H NMR spectrum showed two singlets at 1.08 and 0.91 which is indicative of tertiary methyl groups of the furostane skeleton. Furthermore, two doublets at 1.35 and 1.0 were assignable to secondary methyl groups. The carbon signals for the two tertiary methyl groups were at

C 19.4 and 16.5 and secondary methyl groups at C 17.5 and 16.5, respectively. A quaternary carbon signal at C 110.7 was observed, supporting a furostane skeleton possessing a hydroxyl group at the C-22 position and the C-25 carbon signal at C 34.3 was indicative an R configuration. Similar to 4, long range coupling was observed between the methyl proton signal at H-21 and the carbon signals at C 40.7 C-20,

13 63.9 C-17 and 110.7 C-22. Olefinic carbon signals at C 140.7 and 121.9 in the C NMR spectrum were consistent with a 5 furostane skeleton. The unsaturation between C-5 and

C-6 was further substantiated by a doublet at 5.33 in the 1H NMR spectrum for the H-

6 signal.

103

Figure 2.29. (A) High resolution mass spectrum of compound 5 and (B) expanded view

+ of [M + Na] ion at m/z: 1057.5211 (calculated for C50H82O22Na, 1057.5190). The mass spectrum was acquired on a BioTOF II.

Intens.

A 393.2720

2000

1500 309.1705

1000

1057.5242

229.1005

540.2420

583.3864

911.6323

612.4077

853.5847 969.6699

500 795.5453

641.4385

496.3122

670.4528

1027.7128

467.2894

737.5006

699.4772

763.6047

1085.7482

438.2671

1143.7994 1201.8374

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Intens.

1000 B 1057.5242

800

600 1058.5274

400 1058.0274

200 1059.5229

0 1052 1054 1056 1058 1060 1062 1064 1066 1068 m/z

104

Figure 2.30. LRMS- mass spectrum of compound 5 acquired on BioTOF II.

Intens.

1500 212.1

1250

1000

750 145.0

500 1033.6

250 516.2 171.1

0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Upon acid hydrolysis, derivatization and GC-MS analysis, the retention time and mass spectrum of the TMSi derivative of the aglycone was consistent with that of (25R)- spirost-5-en-3 -ol, which was prepared following the same procedure. The structure of the oligosaccharide moiety was readily deduced from 1H NMR, 13C NMR, HMBC, ESI+–

MS and chemical analysis. 5 showed similar carbon NMR peaks to those of compound 3 and 4, except for the 3′′′ sugar peaks. This identifies 5 to be similar in structure, connectivity and configuration to compound 3 and 4, except for the 3′′′sugar. The 1H

NMR spectrum contained four anomeric proton signals observed at 6.34, 5.01, 4.99 and 4.89. Coupling constants of the anomeric proton resonances suggested - interglycosidic linkages. The 13C NMR spectrum contained four anomeric carbon signals

105

observed at C 105.4, 105.0, 102.4 and 100.0, consistent with the presence of four saccharide residues. The HMBC experiment showed long-range correlations between the anomeric proton signal at 4.99 [H-1′] and the carbon signal at C 77.4 [C-3], between the anomeric proton signal at 5.01 [H-1′′′] and the carbon signal at C 88.2 [C-3′], and the anomeric proton signal at 6.34 [H-1′′] and the carbon signal at C 78 [C-2′]

(Figure 2.31).

Figure 2.31. HMBC long-range correlations for the interglycosidic linkages for the carbohydrate moiety of compound 5.

OH H 77.4 O O HO HO O HO O 78 H OHH 88.2 O 4.99 (7) 4.01 (5.5)

O H 6.34 5 H3C HO OH OH

Upon acid hydrolysis, derivatization and GC-MS analysis, the sugars were identified as

D-(+)-glucose, L-(-)-rhamnose and L-(-)-xylose in a 2:1:1 ratio. 1H NMR spectrum showed a doublet integrating for three protons at 1.77, indicative of the methyl group of rhamnose. The ESI+–MS mass spectrum showed a base ion peak at 1017.7 [M–18+H]+ and the sodium adduct at 1057.7 [M+Na]+. Additionally, ion fragments at m/z 871.6 [M–

18–Rha+H]+, 739.4 [M–18–Xyl–Rha+H] +, 577.2 [M–18–Xyl–Rha–Glu+H]+, and 415.2

[M–18–2Glu–Xyl–Rha+H]+ were observed and were consistent with 5 being bidesmodic with the trisaccharide moiety at the C-3 position containing D-(+)-glucose, L-(-)-xylose

106

and L-(-)-rhamnose, and a D-(+)-glucose moiety at the C-26 position, indicative of a furostanol saponin (Figure 2.32).

Figure 2.32. ESI+–MS mass spectrum of compound 5.

Intens. 1017.7 6 x10 6

2 1057.7

1 271.1 415.3 577.2 739.4 871.6 0 200 400 600 800 1000 1200 m/z

Accordingly, the structure of 5 was determined to be (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-

(1→3)- -D-glucopyranoside (Figure 2.4).

107

Figure 2.33. (A) 1H NMR spectrum and (B) 13C NMR spectrum of compound 5.

108

Compound 5, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -L-xylopyranosyl-(1→3)- -D-glucopyranoside: amorphous

28 -1 solid; [ D - 46.4º (MeOH; c 0.03); mp 200 ºC (dec); IR max (film) cm : 3362 (OH),

2898 (CH), 1636, 1377, 1256, 1035, 912, 838, 811; HRESI–TOFMS m/z: 1057.5242 [M

+ + + Na]+ (calculated for C50H82O22Na, 1057.5190); ESI – MS, m/z 1057.7 (2, [M+Na] ),

1017.7 (100, [M–18+H]+), 871.6 (7, [M–18–Rha+H]+), 739.4 (9, [M–18–Xyl–Rha+H]+),

577.2 (3, [M–18–Xyl–Rha–Glu+H]+), 415.2 (2, [M–18–2Glu–Xyl–Rha+H]+); 1H NMR

(500 MHz) 0.91 [s, 3H, 18-H], 1.00 [d, 3H, J = 6.5 Hz, 27-H], 1.08 [s, 3H, 19-H], 1.35

[d, 3H, J = 7 Hz, 21-H], 1.77 [d, 3H, J = 6.5 Hz, 6′′-H], 4.83 [d, 1H, J = 8 Hz, 1′′′′-H],

4.89-4.88 [m, 1H, 16-H], 4.99 [d, 1H, J = 7 Hz, 1′-H], 5.01 [d, 1H, J = 5.5, 1′′′-H], 5.33

13 [d, 1H, J = 4 Hz, 6-H], 6.34 [br s, 1H, 1′′-H]; C NMR (400 MHz, pyridine-d5) 37.5 [C-

1], 30.1 [C-2], 77.7 [C-3], 38.7 [C-4], 140.7 [C-5], 121.9 [C-6], 32.4 [C-7], 31.7 [C-8],

50.3 [C-9], 37.2 [C-10], 21.1 [C-11], 39.9 [C-12], 40.8 [C-13], 56.6 [C-14], 32.5 [C-15],

81.1 [C-16], 63.9 [C-17], 16.5 [C-18], 19.4 [C-19], 40.7 [C-20], 16.5 [C-21], 110.7 [C-

22], 37.2 [C-23], 28.4 [C-24], 34.3 [C-25], 75.2 [C-26], 17.5 [C-27], 100.0 [C-1′], 78.0

[C-2′], 88.2 [C-3′], 69.7 [C-4′], 77.7 [C-5′], 62.4 [C-6′], 102.4 [C-1′′], 72.5 [C-2′′], 72.9

[C-3′′], 74.4 [C-4′′], 69.5 [C-5′′], 18.7 [C-6′′], 105.5 [C-1′′′], 74.7 [C-2′′′], 78.4 [C-3′′′],

70.7 [C-4′′′], 67.3 [C-5′′′], 105.0 [C-1′′′′], 75.3 [C-2′′′′], 78.6 [C-3′′′′], 71.7 [C-4′′′′], 78.5

[C-5′′′′], 62.8 [C-6′′′′].

109

2.5. Conclusion

Although, saponins are widely distributed secondary metabolites and have been identified in over 100 plant families and in marine organisms such as starfish and sea cucumber (Güçlü-Üstündağ and Mazza, 2006), thus far, steroidal glycoalkaloids are limited to members of the families Solanaceae and Liliaceae (Ghisalberti, 2006). In addition, novel or ―alien‖ glycoalkaloids have been reported from interspecific hybrids of

Solananceous plants (Grassert and Lellbach, 1987). For example, a novel tomatidine glycoalkaloid, not present in either parental species, was identified in the sexual hybrids of S. acaule and Solanum x ajanhuiri (Osman et al., 1986). Interspecific hybridization is widely employed for the development of new Lilium cultivars. Similar to Solananceous plants, interspecific hybridization in the Liliaceae family may result in novel steroidal glycosides. Several novel steroidal saponins including a steroidal saponin with an acetylated glucose, (25R,26R)-26-methoxyspirost-5-en-3 -yl O- -L-rhamnopyranosyl-

(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside, has been isolated from the interspecific hybrid L. speciosum x L. nobilissimum (Nakamura et al., 1994).

Steroidal saponins with acetylation of saccharide residues have been identified; however, the biological significance of acetylation of these compounds is unclear. Some genotypes of Solanum chacoense, a wild potato species that is resistant to Leptinotarsa decemlineata, contain glycoalkaloids that are acetylated at the C-23 position of the steroid aglycone (Sinden et al., 2005), but we are unaware of any steroidal glycoalkaloids that contain naturally occurring acetylated saccharides. Interestingly, the presence or absence of the acetyl moiety of the S. chacoense glycoalkaloids markedly affected

110 resistance to foliar feeding of both adults and larvae of L. decemlineata. Differences in acetylation of the terminal glucose of the trisaccharide moiety of 1 and 2 may also play a biological role in L. longiflorum.

The furostanol saponins 3 – 5, are similar in structure except for the terminal monosaccharide residues and interglycosidic linkages. 3 has a hexose as the terminal sugar linked via the C-4′ carbon of the inner glucose, whereas both 4 and 5 contain a pentose as the terminal sugar linked via the C-3′ carbon of the inner glucose. In fact, differences in oligosaccharide composition and interglycosidic linkages have been shown to affect the biological activity of steroidal saponins possessing the same aglycone moiety

(Yang et al., 2006). Differences in oligosaccharide composition and interglycosidic linkages in 3 – 5 may also play a role in the biology of these compounds in L. longiflorum.

In this chapter, a new acetylated steroidal glycoalkaloid, (22R, 25R)-spirosol-5- en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D- glucopyranoside, and two new furostanol saponins, (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-

(1→3)- -D-glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-

3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-(1→3)- -D- glucopyranoside, from the bulbs of L. longiflorum have been isolated and structures elucidated. Additionally, a known steroidal glycoalkaloid, (22R, 25R)-spirosol-5-en-3 - yl O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, and a known furostanol saponin, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-

111

3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D- glucopyranoside, were isolated for the first time from the bulbs of L. longiflorum. The extraction and purification procedures reported in this chapter may be used for the production of sufficient quantities of pure compounds for biological investigations. These new compounds from L. longiflorum can be used for studies on the biological role of steroidal glycosides in plant development and plant-pathogen interactions, as well as for studies in food and human health, for which little is known.

112

Figure 2.34. ESI+–MS mass spectra of compounds 1 – 5.

Intens + x10. 7 884.7 [M + H] H 1 N 3 O 576.4 414.6 OH OH 2 [M + H + Na]++ O HO O HO O O OH HO O 738.5 1 576.4 414.6 H3C O HO OH OH 271.4 453.8 738.5 7 x10 + 926.6 [M + H] H 1.25 2 N

1.00 O O 576.4 414.3 O OH ++ O 0.75 [M + H + Na] HO O HO O O OH HO O 0.50 780.5 414.3 H3C O HO 0.25 OH 576.4 780.5 OH 271.2 474.8 7 + x10 OH 901.5 1047.7 [M – 18 + H] O 3 HO O HO 1.0 OH OH

O 0.8 577.4 415.3 OH OH O 0.6 HO O HO O O OH HO O 0.4 739.4 H3C O HO 0.2 OH OH 577.4 739.4 901.5 + 271.3 415.3 1087.7 [M + Na]

6 + x10 OH 871.6 1017.7 [M – 18 + H] O 4 HO O HO OH OH 6 O 577.2 415.2 OH OH 4 O O HO O HO O OH O 739.4

2 H3C O HO OH OH 739.4 871.6 577.2 + 271.3 415.2 1057.7 [M + Na] 6 + x10 OH 871.6 1017.7 [M – 18 + H] O 4 5 HO O HO OH OH

3 O 577.2 415.2 OH O 2 HO O HO O HO O OH O 739.4

1 H3C O + HO 1057.7 [M + Na] OH OH 739.4 871.6 271.3 415.2 577.2 0 200 400 600 800 1000 1200 1400 m/z

113

13 Table 2.1. C NMR spectral data of compounds 1 – 5 in pryridine-d5.

compound carbon 1 2 3 4 5 C-1 37.5 37.5 37.5 37.5 37.5 C-2 30.2 30.2 30.2 30.1 30.1 C-3 78.1a 78.1a 78.2a 77.7a 77.4a C-4 38.9 39.0 39.0 38.7 38.7 C-5 140.8 140.8 140.8 140.7 140.7 C-6 121.8 121.9 121.9 121.9 121.9 C-7 32.4 32.4 32.5 32.4 32.4 C-8 31.6b 31.6b 31.7 31.7 31.7 C-9 50.3 50.4 50.4 50.3 50.3 C-10 37.2 37.2 37.2 37.2 37.2 C-11 21.2 21.2 21.2 21.1 21.1 C-12 40.1 40.1 39.9 39.9 39.9 C-13 40.7 40.7 40.8b 40.8b 40.8b C-14 56.7 56.7 56.6 56.6 56.6 C-15 32.6 32.6 32.4 32.5 32.5 C-16 78.8 79.0 81.1 81.1 81.1 C-17 63.6 63.5 63.9 63.9 63.9 C-18 16.5 16.5 16.5 16.5 16.5 C-19 19.4 19.4 19.4 19.4 19.4 C-20 41.6 41.7 40.7b 40.7b 40.7b C-21 15.7 15.7 16.5 16.5 16.5 C-22 98.4 98.4 110.7 110.7 110.7 C-23 34.7 34.7 37.1 37.2 37.2 C-24 31.1 31.0 28.4 28.4 28.4 C-25 31.7b 31.7b 34.3 34.3 34.3 C-26 48.1 48.0 75.3c 75.3c 75.3c C-27 19.9 19.8 17.5 17.5 17.5 C-1' 100.0 99.9 100.0 99.9 100.0 C-2' 77.8 77.6 77.8 78.0 78.0 C-3' 76.2 76.1 76.2 88.0 88.2 C-4' 82.1 83.3 82.1 69.7 69.7 C-5' 77.3 77.4 77.3 77.6a 77.7a C-6' 62.0 62.0 62.1 62.4 62.4 C-1" 101.8 102.0 101.8 102.5 102.4 C-2" 72.5 72.5 72.5 72.5 72.5 C-3" 72.8 72.8 72.8 72.9 72.9 C-4" 74.2 74.2 74.2 74.1 74.1 C-5" 69.5 69.6 69.5 69.5 69.5 C-6" 18.7 18.7 18.7 18.7 18.7 C-1'" 105.3 105.6 105.2 105.6 105.5 C-2'" 75.0 75.1 75.0 72.3 74.7 C-3'" 78.3 78.2 78.5 74.6 78.4 C-4'" 71.2 71.9 71.2 69.7 70.7 C-5'" 78.5 74.9 78.3 67.8 67.3 C-6'" 62.1 64.8 61.9 C-1"" 105.0 105.0 105.0 C-2"" 75.2c 75.2c 75.2c C-3"" 78.6 78.6d 78.6d C-4"" 71.7 71.7 71.7 C-5"" 78.3a 78.5d 78.5d C-6"" 62.8 62.8 62.8 Ac-CH3 20.8 Ac-C=O 170.9 a-d Assignments may be interchanged in each column

114

Chapter 3: Quantitative Analysis of Steroidal Glycosides in Different Organs of

Easter Lily (Lilium longiflorum Thunb.) by LC-MS/MS

3.1. Abstract

The bulbs of the Easter lily (Lilium longiflorum Thunb.) are regularly consumed in Asia as both food and medicine, and the beautiful white flowers are appreciated worldwide as an attractive ornamental. The Easter lily is a rich source of steroidal glycosides, a group of compounds that may be responsible for some of the traditional medicinal uses of lilies. Since the appearance of recent reports on the role steroidal glycosides in animal and human health, there is increasing interest in the concentration of these natural products in plant-derived foods. A LC-MS/MS method performed in multiple reaction monitoring (MRM) mode was used for the quantitative analysis of two steroidal glycoalkaloids and three furostanol saponins, in the different organs of L. longiflorum. The highest concentrations of the total five steroidal glycosides were 12.02

± 0.36, 10.09 ± 0.23, and 9.36 ± 0.27 mg g-1 dry weight in flower buds, lower stems, and leaves, respectively. The highest concentrations of the two steroidal glycoalkaloids were

8.49 ± 0.3, 6.91 ± 0.22, and 5.83 ± 0.15 mg g-1 dry weight in flower buds, leaves, and bulbs, respectively. In contrast, the highest concentrations of the three furostanol saponins were 4.87 ± 0.13, 4.37 ± 0.07, and 3.53 ± 0.06 mg g-1 dry weight in lower stems, fleshy roots, and flower buds, respectively. The steroidal glycoalkaloids were detected in higher concentrations as compared to the furostanol saponins in all of the plant organs except the roots. The ratio of the steroidal glycoalkaloids to furostanol saponins was

115 higher in the plant organs exposed to light and decreased in proportion from the aboveground organs to the underground organs. Additionally, histological staining of bulb scales revealed differential furostanol accumulation in the basal plate, bulb scale epidermal cells, and vascular bundles, with little or no staining in the mesophyll of the bulb scale. An understanding of the distribution of steroidal glycosides in the different organs of L. longiflorum is the first step in developing insight into the role these compounds play in plant biology and chemical ecology and aids in the development of extraction and purification methodologies for food, health, and industrial applications. In this chapter, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D- glucopyranosyl-(1→4)- -D-glucopyranoside, (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside,

(25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, (25R)-26-

O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)-

-L-arabinopyranosyl-(1→3)- -D-glucopyranoside, and (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- xylopyranosyl-(1→3)- -D-glucopyranoside were quantified in the different organs of L. longiflorum for the first time.

116

3.2. Introduction

The Easter lily (Lilium longiflorum Thunb., family Liliaceae), with its showy white flowers and fragrant aroma, is enjoyed worldwide as an attractive ornamental plant.

Easter lilies are most commonly seen as indoor potted plants or floral arrangements around the Easter holidays; however, they are also often planted outdoors as bedding plants in flower gardens. In addition to their esthetic value, lily bulbs and flower buds are regularly consumed as a food in Asia for their distinctive bitter taste and have a long historical use in traditional Chinese medicine. In particular, a preparation of bulbs of various Lilium species, referred to as ―Bai-he‖, is used as a treatment for inflammation and lung ailments (Mimaki and Sashida, 1990; Mimaki et al., 1992). Among many other secondary metabolites, L. longiflorum is a rich source of steroidal glycosides, a structurally diverse class of natural products that includes steroidal saponins and steroidal glycoalkaloids. Steroidal glycosides have been reported to exhibit a wide range of biological activities including antifungal (Sautour et al., 2005; Zhou et al., 2003), platelet aggregation inhibition (Zhang et al., 1999; Huang et al., 2006), anticholinergic (Gilani et al., 1997), antidiabetic (Nakashima et al., 1993), antihypertensive (Oh et al., 2003), cholesterol lowering (Matsuura, 2001), anti-inflammatory (Shao et al., 2007), antiviral

(Gosse et al., 2002), and anticancer (Acharya et al., 2009; Pettit et al., 2005; Mimaki et al., 1999; Jiang et al., 2005). Additionally, steroidal glycosides have a wide variety of commercial uses including as surfactants (Yamanaka et al., 2008), foaming agents (Singh et al., 2003), and vaccine adjuvants (Rajput et al., 2007) and serve as precursors for the industrial production of pharmaceutical steroids (Hansen, 2007). Steroidal saponins have

117 been found in over 100 plant families and in some marine organisms such as starfish and sea cucumber (Güçlü-Üstündağ and Mazza, 2006). They are characterized by a steroid type skeleton glycosidically linked to carbohydrate moieties. Steroidal glycoalkaloids are characterized by a nitrogen-containing steroid type skeleton glycosidically linked to carbohydrate moieties. In contrast to steroidal saponins, the occurrences of steroidal glycoalkaloids are, thus far, limited to members of the plant families Solanaceae and

Liliaceae (Li et al., 2006; Ghisalberti, 2006). Some glycoalkaloids from solanaceaous plants have been shown to play a role in plant defense and are toxic to animals and humans. The potato glycoalkaloids, -solanine and -chaconine, are highly toxic to animals due to their interaction with membrane sterols, disruption of cell membranes, and inhibition of acetylcholinesterase, suggesting a biological role in antiherbivory (Sánchez-

Mata et al., 2010). In Lilium, steroidal glycoalkaloids have been identified in L. philippinense (Espeso and Guevara, 1990), L. mackliniae (Sashida et al., 1991), and L. brownii (Mimaki and Sashida, 1990), and in the previous chapter solasodine-based glycoalkaloids were identified for the first time from L. longiflorum. Interestingly, both the leaves and flowers of L. longiflorum have been reported to be highly nephrotoxic to domesticated cats; however, the toxic compounds have yet to be identified (Rumbeiha et al., 2004; Langston, 2002). Although it has been reported that solasodine-based glycoalkaloids are less toxic then solanidine-based glycoalkaloids (Roddick et al., 2001), the animal and human toxicity of the steroidal glycoalkaloids from L. longiflorum has yet to be investigated. Although the putative biological activities of steroidal glycosides are well documented, the biological role of these compounds in plant metabolism and development is poorly understood. The role of steroidal glycosides in wound response

118 and plant defense, including antifungal and antiherbivory, has been studied extensively

(Zullo et al., 1984; Nozzolillo et al., 1997; Adel et al., 2000; Bowyer et al., 1995;

Osbourn, 1996; Morrissey and Osbourn, 1999; Osbourn, 1999; Papadopoulou et al.,

1999; Morrissey et al., 2000; Trojanowska et al., 2000; Osbourn et al., 2003; Osbourn,

2003; Hughes et al., 2004; Choi et al., 2005). In fact, some steroidal glycosides are toxic to insects such as the European corn borer, Ostrinia nubialis, and army worm,

Spodoptera littoralis (Nozzolillo et al., 1997; Adel et al., 2000). In oats, Avena sativa, biologically inactive steroidal saponins are converted into an antifungal form in response to tissue damage, suggesting a role in the plant-pathogen interaction (Osbourn, 1996;

Morrissey et al., 2000; Osbourn, 2003; Hughes et al., 2004). In addition, the steroidal glycoalkaloids -tomatine and -chaconine play a role in fungal resistance of tomato,

Solanum lycopersicum, and potato, Solanum tuberosum, respectively (Morrissey and

Osbourn, 1999). Although the literature suggests that steroidal glycosides are involved in the plant pathogen interaction and antiherbivory in oats, tomato, and potato, there are no reports on biological role of steroidal glycosides in L. longiflorum.

In the previous chapter, two steroidal glycoalkaloids, (22R, 25R)-spirosol-5-en-

3 -yl O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside

(1), (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside (2), and three furostanol saponins, (25R)-

26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-

119 glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- xylopyranosyl-(1→3)- -D-glucopyranoside (5) were identified for the first time in L. longiflorum. Although steroidal glycosides have attracted scientific attention in recent years for their medicinal and industrial uses, there are only a few studies that have quantified these compounds within different plant organs throughout plant development.

The organ distribution of steroidal glycosides in various plants such as Solanum nigrum,

Solanum incanun (Eltayeb et al., 1997), Asparagus officinalis L. (Wang et al., 2003) and

Dioscorea pseudojaponica (Lin et al., 2008) have been reported, however, there are no studies on the distribution of steroidal glycosides in L. longiflorum.

Steroidal glycosides lack a strong chromophore and occur in complex biological matrices; therefore, nonspecific short-wavelength UV detection is often inadequate.

Analytical methods using evaporative light scattering detection (ELSD) are used to help overcome this obstacle; however, laborious sample preparation and sensitivity issues persist (Oleszek and Bialy, 2006). LC-MS methods operating in selected ion monitoring

(SIM) mode have been developed to increase sensitivity and specificity; however, the separation of structurally similar compounds and shared ions still poses a challenge

(Ghisalberti, 2006; Oleszek and Bialy, 2006). LC-MS/MS in MRM mode overcomes these obstacles, allowing for sensitive quantitative analysis in complex biological matrices with increased specificity over SIM. The purpose of this investigation was to utilize LC-MS/MS in MRM mode to quantify five steroidal glycosides in the different organs of L. longiflorum. Additionally, histological techniques were employed to qualitatively visualize tissue-specific localization of furostanols in bulb scales. An understanding of the distribution and tissue specific localization of steroidal glycosides in

120

L. longiflorum is the first step to develop insight into the biological role these compounds play in plant metabolism, plant development, and chemical ecology. Quantitative analysis of steroidal glycosides in the different organs of L. longiflorum will aid in development studies in animal and human health, toxicology, and optimization of extraction methodologies for potential commercial applications including functional foods, cosmetics, and pharmaceuticals.

3.3. Materials and Methods

3.3.1. Plant material.

Ten L. longiflorum cv. 7-4 plants were grown from tissue-cultured bulbs provided by the Rutgers University lily breeding program. The young bulbs were treated with

Captan (Bayer CropScience AG, Monheim am Rhein, Germany) fungicide prior to planting. Bulbs were planted in raised beds containing Pro-Mix (Premier Horticulture

Inc., Quakertown, PA) soil mix and were grown to mature plants, containing both flower buds and flowers, under greenhouse conditions for 9 months prior to harvest. The greenhouse temperatures were set to provide a minimum day temperature of 24 °C and a minimum night temperature of 18 °C. Plants were fertilized biweekly with a 100mg L min-1 solution of NPK 15-15-15 fertilizer (J. R. Peters Inc., Allentown, PA). Each plant was harvested by hand and manually separated into the following plant organs: bulb scales, fibrous roots, fleshy roots, leaves, lower stems, upper stems, flower buds, and

121 mature flowers. Bulb scales included both inner and outer bulb scales and ranged from

0.8 to 2.0 cm in width and from0.9 to 4.0 cm in length. Fibrous roots were 0.25 – 0.5mm in diameter. Fleshy roots were 2 – 4 mm in diameter. Leaves ranged in size from 6 to 14 cm. Lower stems were defined as the underground portion of the stem, ranged in size from 6 to 10 cm, and were from white to yellow in appearance. Upper stems were defined as the aboveground portion of the stem, ranged in size from 19 to 31 cm, and were green in appearance. Flower buds ranged in size from 3 to 6 cm. Mature flowers ranged in size from 6 to 14 cm. All of the organs from 10 individual plants were pooled together by organ type, immediately frozen under liquid nitrogen, lyophilized on a VirTis AdVantage laboratory freeze-dryer (SP Industries Inc., Warminster, PA), and stored at -80 °C until analyzed (Figure 3.1).

122

Figure 3.1. The different plant organs of L. longiflorum analyzed in this study.

123

3.3.2. Chemicals

The following compounds were obtained commercially: p-

(dimethylamino)benzaldehyde, hydrochloric acid, and pyridine-d5 (0.3%v/v TMS)

(Sigma-Aldrich, St. Louis, MO). All solvents (acetonitrile, chloroform, ethanol, ethyl acetate, formic acid, n-butanol, and n-pentane) were of chromatographic grade (Thermo

Fisher Scientific Inc., Fair Lawn, NJ). Water was deionized (18 MΩ cm) using a Milli-Q water purification system (Millipore, Bedford, MA).

3.3.3. Histology and Microscopy.

Histological detection of furostanols was modified from the method of Gurielidze et al., 2004 (Gurielidze et al., 2004). Bulb scales were carefully cross-sectioned (∼0.5 mm) parallel to the basal plate, soaked for 2 min in a solution of Ehrlich‘s reagent [3.2 g of p-(dimethylamino)-benzaldehyde in 60 mL of 95% ethanol and 60 mL of 12 N HCl], and briefly heated on a microscope slide under an open flame. Transmitted light microscopy was performed with an Axiovert 200 inverted microscope (Carl Zeiss

Microimaging Inc., Thornwood, NY) at magnifications of 10x, 20x, and 40x Axiovision version 3.0 software was used for image acquisition. Furostanol localization was visualized as dark red areas.

124

3.3.4. Purification and Confirmation of Analytical Standards.

Closely following the procedure recently reported in chapter two, the steroidal glycosides 1-5 were isolated as analytical standards from lyophilized L. longiflorum bulbs

(Figure 3.2). Briefly, lyophilized lily bulb powder was washed with n-pentane and extracted with ethanol and deionized water (7:3, v/v). After the removal of solvent, the extract was dissolved in deionized water, washed with ethyl acetate, and extracted with n- butanol. The organic phase was evaporated under reduced pressure and lyophilized, yielding a crude steroidal glycoside extract. The crude glycoside extract was fractionated by gel permeation chromatography and repeated semipreparative RP-HPLC to yield compounds 1 – 5. The standards were obtained as white amorphous powders in high purity, >98%, as determined by LC-MS and NMR.

125

Figure 3.2. Structures of steroidal glycoalkaloids 1 – 2 and furostanol saponins 3 – 5 quantified in the various L. longiflorum organs.

21 H 25 N 26 18 23 24 27 20 22 12 17 19 11 13 16 O 14 9 15 1 OH 2 10 8 R1

4' 3 5 7 6 O 4 1 -D-Glcp R1O O 2' HO 1' 2 6-Ac- -D-Glcp O 1'' H3C O HO OH OH OH O HO O HO OH1'''' 21 OH 26 25 18 23 24 27 20 22 12 17 19 11 13 16 O R2 R3 9 14 15 1 OH 2 10 8 3 -D-Glcp H 4' 3 5 7 4 H -L-Arap O 4 6 R2O O R O 2' 5 H -D-Xlyp 3 O 1' 1'' H3C O HO OHOH

O OH O OH O O O HO HO O HO HO HO HO HO OH OH OH OH

-D-Glcp 6-Ac- -D-Glcp -L-Arap -D-Xlyp

126

3.3.4.1. Nuclear Magnetic Resonance Spectroscopy (NMR).

1D1H NMR and 13C NMR spectra were acquired on an AMX-400 spectrometer and an AMX-500 spectrometer (Bruker, Rheinstetten, Germany). Samples for NMR analysis were dissolved in pyridine-d5 and chemical shifts were calculated as δ values with reference to tetramethylsilane (TMS).

3.3.5. Quantitative Analysis of Steroidal Glycosides in Lilium longiflorum.

3.3.5.1. Sample Preparation

Lyophilized lily organ samples were removed from the freezer and allowed to reach room temperature. The samples were ground to a fine powder with a laboratory mill (IKA Labortechnik, Staufen, Germany) and passed through a sieve (pore size =

270mesh) (W. S. Tyler Inc., Mentor, OH). Each sample (125 mg except 250 mg for fibrous roots and fleshy roots) was weighed separately and transferred into a 50 mL volumetric flask, which was partially filled with ethanol and deionized water (7:3, v/v; 35 mL each). The samples were then extracted on a wrist-action autoshaker (45 min)

(Burrell Scientific, Pittsburgh, PA), sonicated in an ultrasonic water bath (30 min)

(B3500A-DTH ultrasonic bath, VWR International Inc., West Chester, PA), and filled to full volume (50 mL) with ethanol and deionized water (7:3, v/v). The solution was transferred to a centrifuge tube, centrifuged (5000 rpm for 10 min) (Sorvall RC-3C Plus,

127

Thermo Fisher Scientific Inc.), and filtered through a 0.45 μm PTFE syringe filter

(Thermo Fisher Scientific Inc.) prior to LC-MS/MS analysis.

3.3.5.2. Analytical Standard Preparation

Steroidal glycosides 1 – 5, isolated and purified as described in chapter two were used as analytical standards. The analytical standards were accurately weighed into volumetric flasks (10 mL) and partially filled with ethanol and deionized water (7:3, v/v;

7 mL each). Solutions were sonicated (5 min) and filled to full volume (10 mL) with ethanol and deionized water (7:3, v/v). Solutions used for calibration curves were prepared by dilution of the stock solutions. External calibration curves were established over six data points covering a concentration range of 0.086 – 2.75 μg mL-1 for compound 1 and 0.078 – 2.5 μg mL-1 for compounds 2 – 5. Mean areas (n = 3) generated from the standard solutions were plotted against concentration to establish calibration equations. Standard solutions were stored at 4 °C and were allowed to reach room temperature prior to analysis.

3.3.5.3. Liquid Chromatography-Mass Spectrometry (LC-MS/MS).

LC-MS/MS analysis of L. longiflorum extracts was performed using an Agilent

1200 series HPLC system (Agilent Technologies Inc., Santa Clara, CA) equipped with a

FC/ALS Therm autosampler thermostat, a HiP-ALS SL autosampler, a BIN Pump SL

128 binary pump, a TCC SL thermostated column compartment, and a DADSL diode array detector, interfaced to a 6410 triple-quadrupole LC-MS mass selective detector equipped with an API-ESI ionization source. Chromatographic separations were performed on a

Prodigy C18 column (250 x 4.6 mm i.d.; 5.0 μm particle size) (Phenomenex, Torrance,

CA) operated at a flow rate of 1.0 mL min-1, column temperature set to 25 °C, and an injection volume of 10 μL. The binary mobile phase consisted of (A) 0.1% formic acid in deionized water and (B) 0.1% formic acid in acetonitrile. Chromatographic separations were performed using a linear gradient of 15 – 43%B over 40min and then to 95%B over

5 min; thereafter, elution with 95% B was performed for 10 min. The re-equilibration time was 10 min. Mass Hunter Workstation Data Acquisition, Qualitative Analysis, and

Quantitative Analysis software were used for data acquisition and analysis. Quantitative analysis was performed in positive ion mode. Ionization parameters included capillary voltage, 3.5 kV; nebulizer pressure, 35 psi; drying gas flow, 10.0 mL min-1; and drying gas temperature, 350 °C. Full-scan mass data were collected for a mass range of m/z 100

– 1500. MS2 experiments were conducted for precursor and product ion selection for steroidal glycoalkaloids 1 and 2 (Figure 3.3) and furostanol saponins 3 – 5 (Figure 3.4).

129

Figure 3.3. MS2 product ion spectra of steroidal glycoalkaloids 1 – 2.

3 x10 8 884.5 6

4 866.5 2

0 x104 2 926.5

1 908.5

0 850 900 950 Counts vs. Mass-to-Charge (m/z)

130

Figure 3.4. MS2 product ion spectra of furostanol saponins 3 – 5.

3 x10

1047.5 2

885.5 1

0 x103

3 1017.5

2 855.5 1

0 x103

5 1017.5 4 3

2 855.5 1 0 800 900 1000 1100 Counts vs. Mass-to-Charge (m/z)

Flow injection analysis (FIA) experiments were performed to optimize fragmentor voltages and collision energies. The fragmentor voltage was set at 120 V, and collision energies were set to 60, 55, 30, 25, and 25 for compounds 1 – 5, respectively.

By means of the multiple reaction monitoring (MRM) mode, the individual steroidal glycosides were analyzed using the following mass transitions given in parentheses: 1

(m/z 926.5 → 908.5), 2 (m/z 884.5 → 866.5), 3 (m/z 1047.5 → 885.5), 4 (m/z 1017.5 →

855.5), 5 (m/z 1017.5 → 855.5) (Figure 3.5).

131

Figure 3.5. MS/MS chromatograms for the quantitative analysis of compounds 1 – 5 in a L. longiflorum bulb scale using multiple reaction monitoring (MRM) mode.

x10 3 TIC 1

2 2 1 3 4 5 0 x10 3

1 m/z 884.5 → 866.5 4 3 2 1 0 3 x10 2 m/z 926.5 → 908.5

0 x10 2 3 m/z 1047.5 → 885.5 3 2

0 2 x10 4 m/z 1017.5 → 855.5

5 1

0 5 10 15 20 25 30 35 40 Counts vs. Acquisition Time (min)

132

3.3.5.4. Recovery

Recovery rates were calculated using the standard addition method (Skoog et al.,

1997). Lyophilized and finely ground lily organs were separately weighed, transferred into volumetric flasks (50 mL), and spiked with three different concentrations (50, 100, and 200 μg g-1) of purified reference standards dissolved in ethanol and deionized water

(7:3, v/v). The volumetric flasks were then partially filled with ethanol and deionized water (7:3, v/v; 35 mL each). After extraction on a laboratory shaker (45 min) and sonication (30 min), they were filled to full volume (50 mL) with ethanol and deionized water (7:3, v/v). Quantitative analysis was then performed as described above. The recovery rate for each steroidal glycoside in the different plant organs was calculated by comparing the amount of standard in the spiked sample with the content found in the lily organ sample that was not spiked with additional standards (control). Each analysis was performed in triplicate.

3.3.5.5. Thin Layer Chromatography (TLC).

L. longiflorum bulb extract, prepared as described above, was evaporated under reduced pressure (30 °C; 1.0 x 10-3 bar) using a Laborota 4003 rotary evaporator

(Heidolph Brinkman LLC, Elk Grove Village, IL) and lyophilized. Lyophilized bulb extract (1 mg) and compounds 3 – 5 (1 mg) were individually dissolved in methanol (0.5 mL), spotted on a 20 cm x 20 cm silica gel 60 F254 TLC plate (Merck & Co., Inc.,

Whitehouse Station, NJ), and developed with chloroform/methanol/water (8:4:1, v/v/v).

133

To detect furostanols, TLC plates were developed with Ehrlich‘s reagent [3.2 g of p-

(dimethylamino)benzaldehyde in 60 mL of 95% ethanol and 60 mL of 12NHCl] and heated to 110 °C for 5 min. Bright red spots were indicative of a positive reaction.

3.3.5.6. Statistical Analysis.

To examine differences in concentrations of steroidal glycosides in the different plant organs, data were subjected to analysis of variance (ANOVA) and means were separated with Fisher‘s protected LSD (R = 0.05) using SAS version 9.2 for Windows

(SAS Institute Inc., Cary, NC).

3.4. Results and Discussion

3.4.1. Quantification of steroidal glycosides in the different organs of L. longiflorum.

In chapter two, five steroidal glycosides including two steroidal glycoalkaloids and three furostanol saponins were identified for the first time in L. longiflorum. To investigate the natural distribution of these compounds in the different organs of L. longiflorum, compounds 1 – 5 were purified as analytical standards. To quantify compounds 1 – 5 in different organs of L. longiflorum, extracts prepared from bulb scales, fibrous roots, fleshy roots, leaves, lower stems, upper stems, flower buds, and mature flowers were analyzed by LC-MS/MS operating in MRM mode. To assess

134 linearity, calibration curves were constructed over a range of six concentrations. Good linearity was achieved over the concentration ranges of 0.086 – 2.75 μg mL-1 for compound 1 and 0.078 – 2.50 μg mL-1 for compounds 2 – 5 (Figure 3.6; Figure 3.7;

Figure 3.8; Figure 3.9; Figure 3.10). The correlation coefficients for compounds 1 – 5 ranged from R2 = 0.9997 to R2 = 0.9999.

Figure 3.6. Calibration equation for compound 1.

135

Figure 3.7. Calibration equation for compound 2.

136

Figure 3.8. Calibration equation for compound 3.

137

Figure 3.9. Calibration equation for compound 4.

138

Figure 3.10. Calibration equation for compound 5.

To assess the accuracy of the analytical method, recovery rates were calculated for compounds 1 – 5 in each plant organ. Standards were added in defined amounts to each plant organ sample prior to quantitative analysis and compared to a control with no standard addition. Recovery rates were calculated in the bulb scales (98.7 – 100.2%), fibrous roots (95.7 – 102.4%), fleshy roots (98.6 – 102.0%), leaves (95.8 – 103.0%), lower stems (97.7 – 100.1%), upper stems (95.9 – 102.3%), flower buds (95.3 – 101.0%), and mature flowers (96.3 – 101.9%). The recovery rates for compounds 1 – 5 in all

139 organs analyzed were within the range of 95.7 – 103.0%. The precision of the method was tested by multiple injections of the same bulb scale sample (n = 6) and calculating the relative standard deviation (RSD) of compounds 1 – 5. The RSD values for compounds 1 – 5 were 3.24, 2.54, 4.51, 1.63, and 2.58%, respectively. These data clearly demonstrate acceptable recovery rates, RSD, and linearity, suggesting that the LC-

MS/MS method operating in MRM mode is a reliable method for the accurate quantitative determination of compounds 1 – 5 in the different organs of L. longiflorum.

Concentrations of compounds 1 – 5 were determined in bulb scales, fibrous roots, fleshy roots, leaves, lower stems, upper stems, flower buds, and mature flowers of L. longiflorum cv. 7-4. The concentration of compounds 1 – 5 in the different plant organs were significantly different as determined by ANOVA, P < 0.0001 (Table 3.1). The highest concentrations of the total five steroidal glycosides were 12.02 ± 0.36, 10.09 ±

0.23, and 9.36 ± 0.27 mg g-1 dry weight in flower buds, lower stems, and leaves, respectively (Table 3.2). Interestingly, the proportions of the steroidal glycoalkaloids 1 and 2 to furostanol saponins 3 – 5 were variable and decreased from the aboveground plant organs to the underground organs (Figure 3.11). The highest concentrations of the two steroidal glycoalkaloids were 8.49 ± 0.3, 6.91 ± 0.22, and 5.83 ± 0.15 mg g-1 dry weight in flower buds, leaves, and bulbs, respectively (Table 3.2).

140

Table 3.1. ANOVA for concentrations of compounds 1 – 5 in the different organs of L. longiflorum.

Compound Source SS df MS F P-value

1 Between organs 55.0459958 7 7.86371369 813.84 < 0.0001 Error 0.1546 16 0.0096625 Total 55.2005958 23

2 Between organs 30.0797958 7 4.29711369 389.32 < 0.0001 Error 0.1766 16 0.0110375 Total 30.2563958 23

3 Between organs 11.5567167 7 1.65095952 406.81 < 0.0001 Error 0.06493333 16 0.00405833 Total 11.62165 23

4 Between organs 11.18485 7 1.59783571 1681.93 < 0.0001 Error 0.0152 16 0.00095 Total 11.20005 23

5 Between organs 3.81072917 7 0.54438988 706.24 < 0.0001 Error 0.01233333 16 0.00077083 Total 3.8230625 23

141

Table 3.2. Concentrations of compounds 1 – 5 in the different organs of L. longiflorum.

concentration (mg g-1 dw)1,2 compounds bulb scale fibrous root fleshy root leaf lower stem upper stem bud flower

1 3.31c ± 0.12 0.30g ± 0.05 0.55f ± 0.02 4.23b ± 0.01 3.17c ± 0.05 2.87d ± 0.15 4.82a ± 0.19 1.96e ± 0.03 2 2.52bc ± 0.07 0.26f ± 0.04 0.44f ± 0.01 2.68b ± 0.22 2.05d ± 0.05 2.34c ± 0.12 3.67a ± 0.11 0.96e ± 0.05 Total 1-2 5.83c ± 0.15 0.56g ± 0.09 0.99f ± 0.01 6.91b ± 0.22 5.22d ± 0.1 5.21d ± 0.27 8.49a ± 0.30 2.92e ± 0.07

3 1.51c ± 0.08 0.47f ± 0.03 1.55c ± 0.01 2.01b ± 0.07 1.50c ± 0.12 1.15d ± 0.04 2.92a ± 0.06 0.88e ± 0.05 4 0.81c ± 0.04 0.26f ± 0.02 1.76b ± 0.04 0.29f ± 0.01 2.18a ± 0.04 0.67d ± 0.03 0.41e ± 0.01 0.35e ± 0.01 5 0.69c ± 0.05 0.13f ± 0.01 1.06b ± 0.03 0.15f ± 0.01 1.19a ± 0.02 0.37d ± 0.04 0.20e ± 0.01 0.22e ± 0.01 Total 3-5 3.01d ± 0.17 0.86h ± 0.07 4.37b ± 0.07 2.45e ± 0.09 4.87a ± 0.13 2.19f ± 0.05 3.53c ± 0.06 1.45g ± 0.05

Total 1-5 8.84d ± 0.3 1.42h ± 0.14 5.36f ± 0.08 9.36c ± 0.27 10.09b ± 0.23 7.4e ± 0.32 12.02a ± 0.36 4.38g ± 0.12

1 Values with the same letter in each row are not significantly different (p < 0.05). 2 Concentrations are means of triplicates ± SD, expressed on a dry weight basis (dw).

142

Figure 3.11. Proportions of steroidal glycoalkaloids 1 – 2 to furostanol saponins 3 – 5 in the different organs of L. longiflorum. Proportions are based on mg g-1 dry weight basis.

The two steroidal glycoalkaloids, 1 and 2, had a similar pattern of distribution in the various organs of L. longiflorum (Figure 3.12; Figure 3.13). Compound 1 occurred in significantly different concentrations in all of the plant organs analyzed, except for the concentration in the bulb scale compared to the lower stem, which were not significantly different from each other. Compound 1 occurred in the highest concentration of 4.82 ±

0.19 mg g-1 in the flower buds followed by 4.23 ± 0.01 mg g-1 in the leaf tissue.

143

Figure 3.12. Concentrations of steroidal glycoalkaloid 1 in the different organs of L. longiflorum. Bars with the same letter are not significantly different (p < 0.05).

H N

O OH OH O HO O HO O O OH HO O

H3C O HO OH OH

1 a b

c c d

f g

The concentration of 1 was significantly higher in the lower stem as compared to the upper stem, at 3.17 ± 0.05 and 2.87 ± 0.15 mg g-1, respectively. In the underground organs, compound 1 occurred in the highest concentration of 3.31 ± 0.12 mg g-1 in the bulb scale as compared to the lowest concentrations of 0.30 ± 0.05and 0.55 ± 0.02 mg g-1 in the fibrous roots and fleshy roots, respectively. Compound 2, the acetylated derivative

144 of compound 1, was distributed similarly to compound 1; however, it generally occurred in slightly lower concentrations.

Figure 3.13. Concentrations of steroidal glycoalkaloid 2 in the different organs of L. longiflorum. Bars with the same letter are not significantly different (p < 0.05).

H N

O O O OH O HO O HO O O OH HO O

H3C O HO OH OH

b bc c d

e f f

Compound 2 occurred in the highest concentration of 3.67 ± 0.11 mg g-1 in the flower bud followed by 2.68 ± 0.22 mg g-1 in the leaf tissue and 2.52 ± 0.07 mg g-1 in the bulb scale. In contrast to compound 1, compound 2 was significantly higher in the upper stem

145 as compared to the lower stem, at 2.34 ± 0.12 and 2.05 ± 0.05 mg g-1, respectively. The concentration of compound 2 was not significantly different between bulb scales and leaves, 2.52 ± 0.07 and 2.68 ± 0.22 mg g-1, bulbs scales and upper stem, 2.52 ± 0.07 and

2.34 ± 0.12 mg g-1, and between fibrous roots and fleshy roots, 0.26 ± 0.04 and 0.44 ±

0.01 mg g-1. Both steroidal glycoalkaloids occurred in the lowest concentration in the fleshy roots and fibrous roots as compared to the other organs of the plant. Interestingly, compounds 1 and 2 were lower in the flowers as compared to flower buds, which contained the highest concentration of both compounds. Similar to solanaceous plants, glycoalkaloid pairs that differed only in the composition of the carbohydrate moiety were found (Sánchez-Mata et al., 2010; Roddick et al., 2001). Although glycoalkaloids are also present in the edible parts of solanaceous plants, they can be toxic. Solanidine glycoalkaloids found in potato tubers are generally considered to be safe at concentrations of < 0.2 mg g-1 fresh weight (Sinden and Webb, 1972). Lily bulbs contain solasodine glycoalkaloids similar to the glycoalkaloids found in eggplant, and these compounds are less toxic than the solanidine based compounds (Roddick et al., 2001).

The content of steroidal glycoalkaloids, 1 and 2, in lily bulbs and flower buds is > 0.2 mg g-1 fresh weight, but is similar to the content of solasonine and solamargine found in

Solanum macrocarpon, the Gboma eggplant, consumed in parts of Africa, Southeast

Asia, and the Caribbean. The glycoalkaloid levels in the fruits of Gboma eggplant are 5 –

10 times higher than the levels that are considered to be safe for human consumption based on current standards (Sánchez-Mata et al., 2010). Feeding experiments are clearly needed to determine the safe levels of L. longiflorum bulbs and flower buds for human consumption and if the steroidal glycoalkaloids are the toxic compounds in flowers and

146 leaves that are responsible for poisoning in domesticated cats. Nevertheless, solasodine- based glycoalkaloids have been used to treat human skin carcinomas and are of commercial interest as a raw material for the production of pharmaceutical steroids

(Eltayeb et al., 1997).

The distribution of the furostanol saponins 3 – 5 was somewhat different from that of the steroidal glycoalkaloids in the various plant organs of L. longiflorum (Figure

3.14; Figure 3.16; Figure 3.17).

147

Figure 3.14. Concentrations of compound 3 in the different organs of L. longiflorum. Bars with the same letter are not significantly different (p < 0.05).

OH O HO O HO OH OH

O OH O OH HO HO O O OH O HO O

H3C O HO OHOH

b c c c d e f

The highest concentrations of the three furostanol saponins were 4.87 ± 0.13, 4.37

± 0.07, and 3.53 ± 0.06 mg g-1 dry weight in lower stems, fleshy roots, and flower buds, respectively (Table 3.2). Structurally, compounds 3 – 5 are similar except for the interglycosidic linkage and terminal saccharide residues of the C-3 trisaccharide moiety

(Figure 3.15). In compound 3, the terminal sugar is a (+)-D-glucose linked from the C-

1′′′ carbon of the terminal sugar to the C-4′ carbon of the inner glucose. In compound 4,

148 the terminal sugar is (-)-L-arabinose linked from the C-1′′′ carbon to the C-3′ carbon of the inner glucose. Compound 5 has the same interglycosidic linkage as compound 4; however, it contains a (+)-(D)-xylose as the terminal sugar.

Figure 3.15. Differences in saccharide composition and interglycosidic linkages of compounds 3 – 5.

OH OH O 4'' HO O 3 -D-Glcp 4 3 HO O O OH HO -D-Glcp O -L-Rhap 2 H3C O HO OH OH OH OH -L-Arap O O 4 3 4 HO O HO O -D-Glcp 3'' O OH -L-Rhap 2 H3C O HO OHOH OH 5 O O 5 -D-Xlyp 3 HO HO O HO O -D-Glcp OH 3'' O -L-Rhap 2 H3C O HO OH OH

Compound 3 occurred in significantly different concentrations in all of the plant organs except that the bulb scale, lower stem, and fleshy roots were not significantly different. The concentrations in the bulb scale, lower stem, and fleshy roots were 1.51 ±

0.08, 1.50 ± 0.12, and 1.55 ± 0.01 mg g-1, respectively. Compound 3 occurred in the highest concentration of 2.92 ± 0.06 mg g-1 in the flower buds followed by 2.01 ± 0.07 mg g-1 in the leaf tissue. Interestingly, the fibrous roots had the lowest concentration, and

149 similarly to the steroidal glycoalkaloids 1 and 2, the flower buds had a higher concentration than the mature flowers.

Compounds 4 and 5 had a similar pattern of distribution in the various organs of

L. longiflorum; however, compound 5 was slightly lower than compound 4 in all plant organs. The concentrations of both compounds 4 and 5 were not significantly different between fibrous roots and leaves or between flower buds and mature flowers.

150

Figure 3.16. Concentrations of compound 4 in the different organs of L. longiflorum. Bars with the same letter are not significantly different (p < 0.05).

OH O HO O HO OH OH

OH OH O O HO O HO O OH O

H3C O HO OHOH 4

a b

c d f f e e

151

Figure 3.17. Concentrations of compound 5 in the different organs of L. longiflorum. Bars with the same letter are not significantly different (p < 0.05).

OH O HO O HO OH OH

OH O O HO HO O HO O OH O

H3C O HO OHOH

b a c d f f e e

152

3.4.2. Histological visualization of furostanol localization in bulb scale sections of L. longiflorum.

The Ehrlich reagent color reaction was employed to visualize furostanols in bulb scale sections. The furostanols gave a bright red positive reaction, whereas the steroidal glycoalkaloids were not positive for the reaction. Additionally, a crude extract of bulb scales was separated by TLC and only furostanol bands gave a positive reaction, suggesting that staining bulb scales with Ehrlich reagent should not produce positive reactions with non-target compounds. Macroscopically, bulb scale cross sections showed accumulation of furostanols in the outermost layers of the bulb scale. Additionally, a positive reaction was observed surrounding three vascular bundles located in the mesophyll, suggesting furostanol localization is associated with vascular bundles and closely adjacent cells (Figure 3.18).

153

Figure 3.18. Histochemical staining of a bulb scale section. Arrows indicate the epidermis (A) and three vascular bundles (B). Red color indicates the presence of furostanols.

Interestingly, a positive reaction was not observed in the mesophyll, suggesting preferential accumulation and elevated levels of furostanols in the outermost layer of the bulb scale and association with vascular bundles. Microscopically, furostanols were visualized in the highest intensity within the cells of the basal plate, and vascular bundles and preferential accumulation in the intercellular spaces between the mesophyll cells and palisade cell layer were observed (Figure 3.19; Figure 3.20).

154

Figure 3.19. Histochemical analysis of bulb basal plate and bulb scale sections: (A) Bulb basal plate and adjacent bulb scale section. Lettered boxes indicate topography of images (B) and (C); (B) subepidermal intercellular furostanol accumulation along the palisade parenchyma of the basal plate (red); (C) Intercellular furostanol accumulation between spongy tissue cells and palisade parenchyma of the basal plate (red). Furostanol localization is visualized as dark red areas. VB, vascular bundles; BP, basal plate; M, bulb scale mesophyll; EP, epidermal cells.

A C

B BP

20x

B C VB

EP EP

40x 40x

155

Figure 3.20. Histochemical analysis of bulb basal plate and bulb scale sections: (D) Epidermal (red) and mesophyll cells of a bulb scale section; (E) Bulb scale epidermis (red), mesophyll and apical meristem (red); (F) Apical meristem (red) and mesophyll cells; (G) Bulb scale epidermal cells (red) and mesophyll cells. Furostanol localization is visualized as dark red areas. VB, vascular bundles; BP, basal plate; M, bulb scale mesophyll; EP, epidermal cells; AP, apical meristem.

D E

EP EP

AP M

40x 10x

F G

M M

10x 10x

156

These observations were consistent with a histochemical analysis of mature leaf slices of

Achyranthus bidentata that showed triterpenoid saponin accumulation in palisade tissue and phloem cells of the main vein vascular bundles and in phloem cells of the normal and medullary vascular bundles of stem sections (Li and Hu, 2008). Interestingly, in oats, A. sativa, the fluorescent steroidal saponin, avenacin A-1, has been visualized under UV light and found to be localized in the epidermal cell layer of roots (Osbourn, 1999).

Similarly, in a histological study of Dioscorea caucasia, furostanol accumulation was observed in specialized idioblasts in the upper and lower leaf epidermis and was not observed in the leaf mesophyll (Gurielidze et al., 2004). Consistent with the observations in D. caucasia and A. bidentata, there was no positive staining reaction in the mesophyll of the bulb scale of L. longiflorum, suggesting lower levels of furostanols in these tissues.

3.4.3. Quantification of steroidal glycosides within bulb scales of L. longiflorum.

LC-MS/MS operating in MRM mode was employed to quantitatively determine the levels of compounds 1 – 5 within whole bulb scales, the inner portion of bulb scales, and the outermost portion of bulb scales. The outermost layers of intact lyophilized bulb scales were carefully excised with a scalpel. The excised outermost cell layer was approximately 15% of the average mass of the whole intact bulb scale. Quantitative analysis was performed on whole bulb scale, the outermost layer of bulb scale (mostly epidermal and subepidermal cells), and the innermost layer of bulb scale (mostly mesophyll and vascular bundles) (Table 3.3).

157

Table 3.3. Concentrations of compounds 1 – 5 in whole bulb scale, bulb epidermis, and bulb mesophyll.

concentration (mg g-1 dw)1,2 whole bulb bulb bulb compounds scale epidermis mesophyll

3.31b ± 0.12 13.76a ± 0.5 0.06c ± 0.01 1 2 2.52b ± 0.07 11.22a ± 0.3 0.07c ± 0.01 3 1.51 b ± 0.08 7.27a ± 0.4 0.10c ± 0.01 4 0.81b ± 0.04 1.91a ± 0.1 0.20c ± 0.02 5 0.69b ± 0.05 1.90a ± 0.1 0.17c ± 0.02

1 Values with the same letter in each row are not significantly different (p < 0.05). 2 Concentrations are means of triplicates ± SD, expressed on a dry weight basis (dw).

Consistent with what was observed from the histological experiment, the concentrations of furostanols 3 – 5 were higher in the outermost portion of the bulb scale verses the inner bulb scale tissue. These quantitative data confirm the qualitative histological observations made for localization of furostanols in bulb scale sections. Interestingly, the steroidal glycoalkaloids 1 and 2 had the same organ distribution pattern as the furostanols and occurred in elevated levels in the outermost portion of the bulb. The relative portions of compounds 1 – 5 in the outermost layer of bulb scales were 38, 31, 20, 6, and 5%, respectively (Figure 3.21).

158

Figure 3.21. Proportions of compounds 1 – 5 in (A) whole bulb scale, (B) bulb epidermis, and (C) bulb mesophyll. Proportions are based on mg g-1 dry weight basis.

The relative portions of compounds 1 – 5 in the innermost portion of bulb scales were 9, 12, 17, 34, and 28%, respectively. The relative proportions of compounds 1 – 5 in the outermost layer of the bulb scale were similar to that of the whole bulb scale. The relative proportions of compounds 1 – 5 in the innermost section of the bulb scale were most similar to those of the fleshy roots, suggesting that the vascular bundles are most likely contributing to the elevated levels of compound 3 – 5 in this tissue, which is consistent with the histological visualizations (Figure 3.22).

159

Figure 3.22. Proportions of compounds 1 – 5 in different organs of L. longiflorum. Proportions are based on mg g-1 dry weight basis.

In summary, the outermost cell layer of bulb scales that are associated with the bulb epidermis had elevated levels of both steroidal glycoalkaloids and furostanols. The innermost section of bulb scales had lower levels; however, the proportions of furostanols to steroidal glycoalkaloids were different, demonstrating that the cells associated with the vascular bundles have proportions of compounds 1 – 5 similar to those of the fleshy roots as compared to the bulb epidermis. Elevated levels and preferential accumulation of steroidal glycosides in the outermost cell layer of bulb scales and the cells associated with vascular bundles may play a role in wound response and plant-pathogen interaction of L. longiflorum.

160

3.5. Conclusion

In this chapter, two steroidal glycoalkaloids, (22R, 25R)-spirosol-5-en-3 -yl O- -

L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside and (22R,

25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside, and three furostanol saponins, (25R)-26-

O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)-

-D-glucopyranosyl-(1→4)- -D-glucopyranoside, (25R)-26-O-( -D-glucopyranosyl)- furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-

(1→3)- -D-glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-

3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-(1→3)- -D- glucopyranoside, were quantified in the different organs of L. longiflorum for the first time. The highest concentrations of steroidal glycosides were detected in flower buds, lower stems, and leaves. The steroidal glycoalkaloids were detected in higher concentrations as compared to the furostanol saponins in all of the plant organs except for the fibrous and fleshy roots. The proportions of steroidal glycoalkaloids to furostanol saponins were higher in the plant organs exposed to light and decreased from the aboveground organs to the underground organs. The highest concentrations of the steroidal glycoalkaloids were detected in flower buds, leaves, and bulbs. Both steroidal glycoalkaloids had a similar pattern of distribution in the various plant organs; however, the acetylated derivative occurred at lower levels. The furostanol saponins were detected in the highest concentrations in the lower stems, fleshy roots, and flower buds.

Interestingly, the flower buds contained the highest concentrations of compounds 1 – 3,

161 and the fleshy roots contained the highest levels of compounds 4 and 5. In addition, differential accumulation of steroidal glycosides was observed in the basal plate, bulb scale epidermal cells, and vascular bundles.

Steroidal saponins and steroidal glycoalkaloids have been shown to play a role in host defense in several plant species (Osbourn, 1996; Morrissey and Osbourn, 1999;

Osbourn, 1999; Papadopoulou et al, 1999; Osbourn et al., 2003; Osbourn, 2003; Hughes et al., 2004). It has been recognized that pathogen infection is often dependent upon the developmental stage of the plant and is tissue or organ specific (Straathof and Löffler,

1994). It is possible that location specific infection or resistance is related to the level of specific steroidal glycosides present at a developmental stage or tissue specific location.

A correlation between total saponin content and resistance to the plant pathogenic fungus

Fusarium oxysporum f. sp. lilii has been observed in several hybrid Lilium cultivars; however, only the total saponin content was measured and specific compounds were not discriminated in the analysis (Curir et al, 2003). The results in this chapter show that the levels of the two steroidal glycoalkaloids and three furostanol saponins varied in the different organs and are preferentially accumulated in different tissues. Thus, it is possible that the levels of specific steroidal glycosides present at a developmental stage or tissue specific location may play a role in pathogen resistance (e.g. the high levels of compounds 1 – 3 in the developing flower buds). This concept may be investigated by observing whether there is a correlation between resistance or susceptibility in a specific plant tissue and the presence or absence of a specific steroidal glycoside or steroidal glycoside profile exists. If resistance to the pathogen is found to be associated with a specific steroidal glycoside concentration or steroidal glycoside profile, an investigation

162 can be conducted on the exogenous application of isolated steroidal glycosides on other plant species susceptible to the pathogen. This work can also be extended to plant pathogenic fungi that are not pathogenic to L. longiflorum.

Quantitative analysis of steroidal glycosides in the different organs of L. longiflorum is the first step to developing insight into the biological role these compounds play in plant metabolism, plant development, and plant-pathogen interactions. The results of this study will aid in the development of future studies in animal and human health and toxicology and of commercial applications such as functional foods, cosmetics, and pharmaceuticals.

163

Chapter 4: Antifungal Activity and Fungal Metabolism of Steroidal Glycosides of

Easter Lily (Lilium longiflorum) by the Plant Pathogenic Fungus, Botrytis cinerea

4.1. Abstract

Botrytis cinerea Pers. Fr. is a plant pathogenic fungus and the causal organism of blossom blight of Easter lily (Lilium longiflorum Thunb.). Easter lily is a rich source of steroidal glycosides, compounds which may play a role in the plant-pathogen interaction of Easter lily. Five steroidal glycosides, including two steroidal glycoalkaloids and three furostanol saponins, were isolated from L. longiflorum and evaluated for fungal growth inhibition activity against B. cinerea, using an in vitro plate assay. All of the compounds showed fungal growth inhibition activity; however, the natural acetylation of C-6′′′ of the terminal glucose in the steroidal alkaloid, (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside, increased antifungal activity by inhibiting the rate of metabolism of the compound by the

B. cinerea. Acetylation of the glycoalkaloid may be a plant defense response to the evolution of detoxifying mechanisms by the pathogen. The biotransformation of the steroidal glycoalkaloids by B. cinerea led to the isolation and characterization of several fungal metabolites. The fungal metabolites that were generated in a model system were also identified in Easter lily tissues infected with the fungus by LC-MS. In addition, a steroidal glycoalkaloid, (22,R 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-

-D-glucopyranoside, was identified as both a fungal metabolite of the steroidal glycoalkaloids and as a natural product in L. longiflorum for the first time.

164

4.2. Introduction

Botrytis cinerea Pers. Fr. is a necrotrophic plant pathogenic fungus with a broad host range and is the cause of grey mold disease, one of the most important post-harvest diseases of fruits and vegetables worldwide. B. cinerea infects over 200 species of economically important plants and is a causal organism of blossom blight of Easter lily

(Lilium longiflorum Thunb.) and other ornamental lily species (Wehlburg et al., 1975;

Raabe et al, 1981; Williamson et al., 2007). In contrast to Botrytis eliptica, the cause of fire blight of Easter lily, B. cinerea typically does not infect healthy plant tissues and is often found growing on stressed, wounded, or senescing tissues (Van Baarlen et al.,

2004). B. cinerea over winters as sclerotia in crop debris and penetrates young plant tissues where it remains latent until the environmental conditions are conductive to infection (Williamson et al., 2007). The fungus flourishes in cool temperatures and high humidity, often during the end of the growing season in late summer and early fall

(McRae, 1998). In fact, it is recommended to treat Easter lily flower buds with fungicide prior to cold storage as a post-harvest management strategy for the control of blossom blight in potted Easter lily plants and cut flowers prior to shipment (McAvoy, 2010).

Host defense responses to B. cinerea have been investigated in many plants including thale cress, Arabidopsis thaliana, and tomato, Solanum lycopersicum

(Williamson et al., 2007). In response to infection by B. cinerea, activation of host defense pathways including the production of antifungal metabolites and pathogenesis related proteins have been reported (Van Baarlen et al., 2004). In addition to the

165 activation of inducible defense pathways, steroidal glycosides including steroidal saponins and steroidal glycoalkaloids, constitutively present in plant tissues, have been shown to play a role in host defense in several plant species (Osbourn, 1996; Morrissey and Osbourn, 1999; Osbourn, 1999; Papadopoulou et al, 1999; Osbourn et al., 2003;

Osbourn, 2003; Hughes et al., 2004).

Steroidal saponins are widely distributed secondary metabolites and have been found in over 100 plant families (Güçlü-Üstündağ and Mazza., 2006). They are characterized by a steroid type skeleton glycosidically linked to sugar moieties. Steroidal glycoalkaloids are similar in structure to steroidal saponins; however, they have nitrogen present in the steroidal aglycone. In contrast to steroidal saponins, steroidal glycoalkaloids are only found in the Solanaceae and Liliaceae (Li et al., 2006;

Ghisalberti, 2006).

Due to the amphipathic nature of the molecules, steroidal glycosides have been shown to disrupt cell membranes both in vitro and in vivo (Steel and Drysdale, 1988;

Roddick et al., 2001). Some studies suggest that membrane disruption may be due either to the interaction of the aglycone with membrane bound sterols, resulting in the formation of membrane pores (Armah et al., 1999) or the extraction of membrane bound sterols, causing loss of lipid bilayer integrity and membrane leakage (Keukens et al., 1992;

Keukens et al., 1995). The antifungal mechanisms of steroidal glycosides remain unclear and the exact mechanism remains to be elucidated.

Fungal plant pathogens such as Gaeumannomyces graminis and Stagonospora avanae have the ability to enzymatically detoxify host plant saponins (Bowyer et al.,

1995; Morrissey et al., 2000). Interestingly, B. cinerea has been shown to produce

166 enzymes that can metabolize a variety of plant defense compounds from active forms to inactive forms (Staples and Mayer et al., 1995). In tomato, B. cinerea metabolizes the antifungal steroidal glycoalkaloid, -tomatine, to an inactive form by enzymatic cleavage of the entire saccharide moiety, or by the cleavage of the terminal xylose by a - xylosidase enzyme (Verhoeff and Liem, 1975; Quidde and Osbourn, 1998). In addition, other plant pathogenic fungi such as Septoria lycopersici and Fusarium oxysporum f.sp. lycopersici detoxify -tomatine through independent metabolic pathways (Arneson and

Durbin, 1967; Ford et al., 1977). Investigations have been conducted on the interaction of steroidal glycosides and B. cinerea; however, to date there are no studies on the interaction of steroidal glycosides from L. longiflorum and B. cinerea.

Easter lily is a rich source of steroidal glycosides. In the Chapter two and Chapter three, two steroidal glycoalkaloids, (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside (1), (22R,

25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O-acetyl- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside (2), and three furostanol saponins, (25R)-

26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-

(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside (3), (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- arabinopyranosyl-(1→3)- -D-glucopyranoside (4) and (25R)-26-O-( -D- glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L- xylopyranosyl-(1→3)- -D-glucopyranoside (5) were identified and quantified in the various plant organs of L. longiflorum. Steroidal glycosides are known to be inhibitory to fungal growth and may play a role in the plant-pathogen interaction. The goal of this

167 study was to (1): evaluate the biological activity of five steroidal glycosides isolated from the bulbs of L. longiflorum on the growth of the plant pathogenic fungus B. cinerea, using an in vitro assay and (2): to use a model system to generate fungal metabolites of the steroidal glycosides, isolate and characterize the metabolites, and determine if the fungal metabolites were present in plant tissue infected with B. cinerea.

168

4.3. Materials and Methods

4.3.1. Plant material.

L. longiflorum, cultivar 7-4, plants were provided from the Rutgers University lily breeding program. Young bulbs were treated with Captan (Bayer CropScience AG,

Monheim am Rhein, Germany) fungicide prior to planting. Bulbs were planted in raised beds containing Pro-Mix (Premier Horticulture Inc., Quakertown, PA) and were grown to mature plants under greenhouse conditions for 9 months prior to harvest. The greenhouse temperatures were set to provide a minimum day temperature of 24 °C and a minimum night temperature of 18 °C. Plants were fertilized biweekly with a 100 mg L-1 solution of

NPK 15-15-15 fertilizer (J.R. Peters Inc., Allentown, PA). Each plant produced 3 – 5 new bulbs, which were used for extraction. For the purification of steroidal glycosides 1 – 5, each plant was harvested by hand and the bulbs were manually separated, immediately frozen under liquid nitrogen, lyophilized on a VirTis AdVantage laboratory freeze dryer

(SP Industries inc.,Warminster, PA) and stored at -80°C until extraction. For the fungal inoculation studies, small sections of aerial stems and adjacent leaves of healthy growing plants were chosen. Small sections were carefully excised from intact plants approximately 5 cm below the apical meristem.

169

4.3.2. Fungal cultures.

An isolate of Botrytis cinerea was obtained from the Plant Diagnostic Laboratory at Rutgers Cooperative Research and Extension (New Jersey Agricultural Experiment

Station). Cultures were maintained on potato dextrose agar (PDA, 39 g L-1 deionized water) (Thermo Fisher Scientific Inc., Fairlawn, NJ) and incubated in the dark at 25 °C.

4.3.3. Chemicals.

The following compounds were obtained commercially: Sephadex LH-20, hydrochloric acid, sodium hydroxide, Tween-80, chloroform-d (0.03% v/v TMS),

methanol-d4 (0.03% v/v TMS), and pyridine-d5 (0.03% v/v TMS) were purchased from

Sigma-Aldrich (St. Louis, MO); and (22R, 25R)-spirosol-5-ene-3 -ol (Glycomix Ltd,

Reading, UK). All solvents (acetonitrile, chloroform, ethanol, ethyl acetate, formic acid, n-butanol, and n-pentane) were chromatographic grade and purchased from Thermo

Fisher Scientific Inc. (Fairlawn, NJ). Potato dextrose agar and potato dextrose broth was purchased from Thermo Fisher Scientific Inc. (Fairlawn, NJ). Water was deionized (18

MΩ cm) using a Milli-Q-water purification system (Milli-Q, Bedford, MA).

170

4.3.4. Isolation and Purification of Steroidal Glycosides 1 – 5 from Lilium longiflorum.

Closely following the procedure reported in chapter two, the following five steroidal glycosides were isolated from lyophilized L. longiflorum bulbs. The compounds were obtained as white amorphous powders in high purity > 98%, as determined by LC-MS and NMR.

Compound 1, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D- glucopyranosyl-(1→4)- -D-glucopyranoside. 1H NMR and 13C NMR were consistent with the literature (Mimaki and Sashida, 1990).

Compound 2, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O- acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside. 1H NMR and 13C NMR were consistent with the literature (Munafo et al., 2010).

Compound 3, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside. 1H NMR and 13C NMR were consistent with the literature (Ori et al., 1992).

Compound 4, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-(1→3)- -D-glucopyranoside. 1H NMR and 13C NMR were consistent with the literature (Munafo et al., 2010).

Compound 5, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O- -L- rhamnopyranosyl-(1→2)- -L-xylopyranosyl-(1→3)- -D-glucopyranoside. 1H NMR and

13C NMR were consistent with the literature (Munafo et al., 2010).

171

4.3.4.1. Nuclear Magnetic Resonance Spectroscopy (NMR).

1H NMR and 13C NMR spectra were acquired on an AMX-400 spectrometer

(Bruker, Rheinstetten, Germany). For NMR analysis, all compounds were dissolved in pyridine-d5, except for compounds 7 and 10 which were dissolved methanol-d4 and chloroform-d, respectively. Chemical shifts were generated as δ values with reference to tetramethylsilane (TMS).

4.3.4.2. Liquid Chromatography-Mass Spectrometry (LC-MS).

LC-MS analysis was performed using a HP 1100 series HPLC system (Agilent

Technologies Inc., Santa Clara, CA) equipped with an auto injector, quaternary pump, column heater, and diode array detector, interfaced to a Bruker 6300 series ion-trap mass spectrometer equipped with an electrospray ionization chamber. Chromatographic separations were performed using a Prodigy C18 column (250mm x 4.6mm i.d.; 5.0 m particle size) (Phenomenex, Torrance, CA). The flow rate was set to 1.0 mL min-1 and the column temperature was set to 25 °C. The binary mobile phase composition consisted of (A) 0.1% formic acid in deionized water and (B) 0.1% formic acid in acetonitrile.

Separations were performed using a linear gradient of 15 – 43% B over 40 min and then to 95% B over 5 min; thereafter, elution with 95% B was performed for 10 min. The re- equilibration time was 10 min. For instrumentation control and data acquisition, HP

ChemStation and BrukerData Analysis software was used. All mass spectra were

172 acquired in positive ion mode over a scan range of m/z 100 – 2000. Ionization parameters included: capillary voltage, 3.5 kV; end plate offset, -500V; nebulizer pressure, 50 PSI; drying gas flow, 10 ml min-1; and drying gas temperature, 360 °C. Trap parameters included: ion current control, 30000; maximum accumulation time, 200 ms; trap drive,

61.2; and averages, 12 spectra.

4.3.4.3. Partial acid hydrolysis of compound 1.

Compound 1 (1 mg) was refluxed in a reaction vial (1 mL) (Reacti-Vial, Thermo

Fisher Scientific Inc., Fairlawn, NJ) at 80 °C for 2 hours in a solution of 1N HCl in methanol (0.5 mL). After hydrolysis and titration to pH 7 with NaOH (4N), the sample was evaporated to dryness under reduced pressure (30 °C; 1.0 x 10-3 bar) using a

Labarota 4003 rotary evaporator (Heidolph Brinkman LLC, Elk Grove Village, IL). The residue was dissolved in ethanol and water (7:3, v/v; 2 mL), mixed on a vortex mixer (1 min) and filtered through 0.45 m PTFE syringe filter (Thermo Fisher Scientific Inc.,

Fairlawn, NJ) prior to LC-MS analysis (Figure 4.1).

173

Figure 4.1. Total ion chromatogram (TIC) of the partial acid-catalyzed hydrolysis products of compound 1. Compound 9 is (22,R 25R)-spirosol-5-en-3 -yl O- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside, compound 7 is (22,R 25R)-spirosol-5-en- 3 -yl O- -D-glucopyranoside, and compound 10 is (22,R 25R)-spirosol-5-en-3 -ol (solasodine).

x10 7

3 7

9 2

10 1

0 10 15 20 25 30 Time [min]

4.3.5. B. cinerea growth inhibition assay.

Antifungal activity was assessed by an in vitro fungal growth inhibition assay modified from Nicol et al. (Nicol et al., 2001). Fungi were maintained on potato dextrose agar (PDA, 39 g L-1 deionized water) and incubated in the dark at 25 °C. The cultures were continuously maintained by transferring a 5 mm plug of mycelium cut with a cork bore from the periphery of actively growing colonies to freshly prepared media. The fungal growth inhibition of compounds 1 – 5 were evaluated at three concentrations (1,

10, 100 mol) in the final media. Solutions of compound 1 – 5 were prepared in ethanol and water (7:3, v/v), filter sterilized with a 0.22 m sterile syringe filter (Thermo Fisher

174

Scientific Inc., Fairlawn, NJ) and incorporated into autoclaved PDA that was allowed to cool to 50 °C. The media (each plate; 5 mL) was then transferred to polystyrene Petri dishes (50 mm x 12 mm) (VWR International Inc., West Chester, PA), and allowed to solidify. The final concentration of carrier solvent was 1% of the final volume of media for all treatments and control. Plates were inoculated with a 5 mm plug taken from the periphery of an actively growing stock culture. Plates were incubated in the dark at 25 °C and the radial growth of each colony was measured using an ABS Solar Digimatic

Caliper (Mitutoyo America Corporation, Aurora, IL). Treatment and control colonies were measured when the control colonies reached approximately 80% of plate diameter.

The average control colony diameter minus the average treatment colony diameter was used to calculate the relative growth inhibition. (% inhibition = (average control diameter mm – treatment diameter mm)/average control diameter (mm).

4.3.6. In vitro fungal metabolism of compounds 1 and 2.

Solutions of compound 1 and 2 were separately prepared in ethanol and water

(7:3, v/v), filter sterilized with a 0.22 m syringe filter and incorporated into autoclaved

PDA that was allowed to cool to 50 °C. The media (each plate; 5 mL) was then transferred to polystyrene Petri dishes (50 mm x 12 mm) and allowed to solidify. A total of 4 plates were prepared for each compound. The final concentration for each compound was 100 molar. Plates were inoculated in the center with a 5 mm plug taken from the periphery of an actively growing stock culture. Plates were incubated in the dark at 25 °C and harvested at 48 and 72 hours, respectively. At harvest, the total contents of 2 plates

175

(total; 10 mL) were transferred to a centrifuge tube (50 mL) (Thermo Fisher Scientific

Inc., Fairlawn, NJ) containing ethanol and water (7:3, v/v; 35 mL). Each sample was then extracted on a wrist-action autoshaker (15 min) (Burrell Scientific, Pittsburg, PA), sonicated in an ultrasonic water bath (15 min) (B3500A-DTH ultrasonic bath, VWR

International Inc., West Chester, PA), and centrifuged (5000 rpm for 10 min) (Sorvall

RC-3C Plus, Thermo Fisher Scientific Inc.). The supernatant was then filtered through

0.45 m PTFE syringe filter prior to LC-MS analysis

4.3.7. Scale-up fungal metabolism of compound 1.

Potato dextrose broth (PDB, 39 g L-1 deionized water) (each; 100 mL) was prepared and transferred to an Erlenmeyer flask (250 mL) and autoclaved. Once the broth reached room temperature, it was inoculated with B. cinerea and incubated on an orbital platform shaker at 200 rpm for 48 hours at 25 °C. After 48 hours, compound 1 (35 mg) was dissolved in ethanol and water (7:3, v/v; 1 mL), filter sterilized with a 0.22 m sterile syringe filter and introduced to the flask. The reaction was monitored by sampling aliquots (each; 0.5 mL) every 24 hours for 96 hours (Figure 4.2; Figure 4.3; Figure 4.4;

Figure 4.5). Each aliquot was diluted (1:3, v/v) with ethanol and water (7:3, v/v; 1.5 mL), and filtered through 0.45 m PTFE syringe filter prior to LC-MS analysis. At the completion of the reaction, the content of the flask was immediately frozen under liquid nitrogen, lyophilized and stored at -80 °C until extraction.

176

Figure 4.2. Extracted ion chromatograms of m/z 885 (compound 1) from samples taken every 24 hours over the course of 96 hours. This illustrates the decrease over time due to its metabolism by B. cinerea.

H N

O OH OH O HO O HO O O OH HO O

H3C O HO OH OH

Intens. 8 x10

24 hrs

0.8

0.6

0.4 48 hrs

72 hrs 0.2 96 hrs

0.0 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 Time [min]

177

Figure 4.3. Extracted ion chromatograms of m/z 723 (compound 6) of samples taken every 24 hours over the course of 96 hours. This illustrates the increase of the fungal metabolite, compound 6, derived from the metabolism of compound 1 by B. cinerea.

H N

OH O HO O HO O

H3C O HO OH OH

Intens. x10 7 96 hrs 6

72 hrs 5

4 48 hrs

2 24 hrs

18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 Time [min]

178

Figure 4.4. Extracted ion chromatograms of m/z 577 (compound 7) of samples taken every 24 hours over the course of 96 hours. This illustrates the increase of the fungal metabolite, compound 7, derived from the metabolism of compound 1 by B. cinerea.

H N

OH O HO O HO OH

Intens. 8 96 hrs x10 72 hrs 2.5

48 hrs 2.0

1.5

1.0

0.5 24 hrs

0.0 19 20 21 22 23 24 25 Time [min]

179

Figure 4.5. Extracted ion chromatograms of m/z 414.6 (compound 10) of samples taken every 24 hours over the course of 96 hours. This illustrates the increase of the fungal metabolite, compound 10, derived from the metabolism of compound 1 by B. cinerea.

H N

Intens. 7 x10 96 hrs 2.0 72 hrs

1.5 48 hrs

1.0

0.5 24 hrs

0.0 30 31 32 33 34 Time [min]

180

4.3.7.1. Semi-preparative RP-HPLC isolation of the fungal metabolites of compound 1.

The lyophilized reaction mixture, as described above, was ground into a fine powder with a laboratory mill (IKA Labortechnik, Staufen, Germany) and extracted with ethanol and water (7:3, v/v; 2 x 50 mL) on an autoshaker at room temperature for 15 minutes. After centrifugation (5000 rpm for 10 minutes), the supernatant was collected and the residue discarded. The supernatant was then evaporated under reduced pressure, dissolved in a mixture of 0.1% formic acid in deionized water and 0.1% formic acid in acetonitrile (75:25, v/v; 5 mL), and filtered through 0.45 m PTFE syringe filtered prior to purification. Chromatographic separations were achieved by semipreparative RP-

HPLC performed on a Luna C18 column (250 mm x 21.2 mm i.d.; 10 m particle size)

(Phenomenex, Torrance, CA). Chromatography was performed on a Shimadzu LC-6AD liquid chromatograph (Shimadzu Scientific Instruments Inc, Columbia, MD) using a

UV/VIS detector and a 2 mL injection loop. Mixtures of (A) 0.1% formic acid in deionized water and (B) 0.1% formic acid in acetonitrile were used as the mobile phase.

The flow rate was set to 20 mL min-1, the column temperature was 23 ± 2 °C and UV detection was recorded at = 210 nm (Figure 4.6). Chromatography was performed using a linear gradient of 5 – 30% B over 45 min and then to 90% B over 10 min; thereafter, elution with 90% B was performed for 10 min. The re-equilibration time was

10 min.

181

Figure 4.6. RP-HPLC chromatogram (λ = 210 nm) of compounds 6, 7, and 10 isolated from the biotransformation of compound 1 by B. cinerea.

40 7

210)

20 10 6

Intensity ( Intensity 10

30 35 40 45 50 time (min)

The target compounds were collected, freed from solvent under reduced pressure and lyophilized, yielding 6 (2 mg), 7 (5 mg), and 10 (1 mg) as white amorphous powders in high purity > 98%, as determined by LC-MS (Figure 4.7) and NMR.

Compund 6. (22,R 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D- glucopyranoside. 1H NMR and 13C NMR were consistent with the literature (Mimaki and

Sashida, 1990).

Compund 7. (22,R 25R)-spirosol-5-en-3 -yl O- -D-glucopyranoside. 1H NMR and 13C

NMR were consistent with the literature (Kim et al, 1996).

Compound 10 (22,R 25R)-spirosol-5-en-3 -ol. 1H NMR and 13C NMR were consistent with the literature (Bird et al., 1979).

182

Figure 4.7. Total ion chromatograms (TIC) of compounds 6, 7, and 10 isolated by RP-HPLC.

Intens. 8 x10 6 1.5

1.0

0.5

0.0 8 x10 7

8 x10 5 10

10 15 20 25 30 Time [min]

183

4.3.8. Isolation and Purification of Compound 6 from Lilium longiflorum bulbs.

4.3.8.1. Sequential Solvent Extraction of Lyophilized L. longiflorum Bulbs.

Lyophilized lily bulbs (100 g) were frozen in liquid nitrogen, ground into a fine powder with a laboratory mill and extracted with n-pentane (3 x 100 mL) on an autoshaker at room temperature for 30 minutes. After centrifugation (5000 rpm for 10 minutes), the organic layers were discarded and the pellet was freed from residual solvent. The residual material was then extracted with a mixture of ethanol and water

(7:3, v/v; 2 x 150 mL) on an autoshaker for 45 minutes at room temperature. After centrifugation (5000 rpm for 10 minutes) and vacuum filtration through a Whatman 114 filter paper (Whatman International Ltd., Maidstone, UK), the supernatant was collected and the residue discarded. The supernatant was then evaporated under reduced pressure and lyophilized, yielding a crude bulb extract (12.9g). The lyophilized crude bulb extract was then dissolved in deionized water (100 mL), washed with ethyl acetate (5 x 100 mL) and the organic phase was discarded. The aqueous phase was then extracted with n- butanol (5 x 100 mL) and the aqueous phase was discarded. The organic phase was then evaporated under reduced pressure (30 °C; 1.0 x 10-3 bar) and lyophilized, yielding a crude glycoside extract (2.2g).

184

4.3.8.2. Gel Permeation Chromatography (GPC).

Crude glycoside extract (1.0 g) was dissolved in a solution of ethanol and water (7:3, v/v; 5.0 mL), filtered with a 0.45 m PTFE syringe filter and then applied onto a standard threaded 4.8cm x 60cm glass column (Kimble Chase Life Science and Research Products

LLC, Vinland, NJ) packed with Sephadex LH-20 (Amersham Pharmacia Biotech,

Uppsala, Sweden) that was washed and conditioned in the same solvent mixture overnight. Chromatography was performed with isocratic ethanol and water (70:30, v/v) at a flow rate of 3.5 mL min-1. The first 200 mL of effluent was discarded and 25 fractions (25 mL each) were collected and analyzed by LC-MS as described above. Based on the LC-MS analysis, GPC fractions 8 through 10 contained the highest levels of compound 6 and were combined, evaporated under reduced pressure, and lyophilized, yielding GPC fraction A (25mg) (Figure 4.8).

185

Figure 4.8. Isolation scheme for compound 6 purified from the bulbs of L. longiflorum. EtOH, ethanol; EtAC, ethyl acetate; n-BuOH, n-butanol; ACN, acetonitrile.

4.3.8.3 Semipreparative Reverse-Phase High Performance Liquid Chromatography (RP-

HPLC).

Purification of compound 6 from GPC fraction A was achieved by semipreparative RP-HPLC under the same conditions as described for above (Figure

4.9).

186

Figure 4.9. RP-HPLC chromatogram (λ = 210 nm) of compound 6 isolated from L. longiflorum bulb n-butanol extract fractionated by gel permeation chromatography (GPC).

210)

Intensity ( Intensity 6 5

20 25 30 35 40 time (min)

Compound 6 was collected, freed from solvent under reduced pressure, and lyophilized yielding 6 (3 mg) as a white amorphous powder in high purity > 98%, as determined by

LC-MS and NMR. 1H NMR and 13C NMR were consistent with the literature (Mimaki and Sashida, 1990).

187

Figure 4.10. Structures of compounds 1 – 10.

21 H 25 N 26 18 23 24 27 20 22 R R R 12 1 2 3 17 19 11 13 16 O 9 14 15 1 S2 S4 S1 1 2 10 8 2 S2 S4 S3

3 5 7 6 4 R1O

R4 R5 3 S1 H OH 4 H S5 5 H S6 O HO O HO 1'''' OH 21 OH 26 25 18 23 24 27 R1 R2 R3 20 22 12 17 19 11 13 6 S2 S4 H 16 O 14 9 15 7 S2 H H 1 OH 2 10 8 8 S2 H S3 4' 3 5 7 O 4 6 9 S2 H S1 R4O O 2' R O 1' 10 H 5 O 1'' H3C O HO OHOH

O OH OH O O O O HO HO R3O S3 HO S1 HO S2 HO OH OH OR2 -D-Glcp -D-Glcp 6-Ac- -D-Glcp

S4 OH O O H C O HO 3 HO S5 HO S6 HO OH OHOH OH -L-Rhap -L-Arap -D-Xlyp

188

4.3.9. Infection of L. Longiflorum tissue and sample preparation for LC-MS analysis.

Small sections of aerial stems were excised from intact plants approximately 5 cm below the apical meristem, including several small (~3 cm) leaves. The plant tissue was surface sterilized (10 % bleach and 0.01% tween-80, v/v) for 10 minutes, rinsed with sterilized DI water, and transferred aseptically to a Petri dish (90 mm x 15 mm).

Treatment tissues were inoculated with two 5 mm plugs of B. cinerea and incubated at in the dark at 25 C. Control samples were treated under the same conditions without fungal inoculation. Once fully colonized with mycelium (7 days) the samples were frozen under liquid nitrogen and lyophilized. The lyophilized material was ground with a mortar and pestle and passed through a sieve (pore size; 270 mesh) (W.S. Tyler Inc., Mentor, OH).

The fine powder (0.5 g) was transferred to a centrifuge tube (15 mL), extracted with ethanol and water (7:3, v/v; 5 mL) on an autoshaker at room temperature for 10 minutes and sonicated in an ultrasonic water bath (10 min). After centrifugation (5000 rpm for 10 minutes), the supernatant was collected and filtered through 0.45 m PTFE syringe filtered prior LC-MS analysis. LC-MS analysis was performed as described above.

4.3.10. Statistical Analysis.

Data was subjected to analysis of variance (ANOVA) and regression analysis using SAS version 9.2 for Windows (SAS Institute Inc., Cary, NC).

189

4.4. Results and Discussion

4.4.1. Fungal growth inhibition assay.

Analysis of variance was performed to determine if there was a significant effect of treatment (compounds 1 – 5), rate (1, 10, 100 mol), and the interaction between treatment and rate. There was a significant interaction between treatment and rate (P <

0.0001), thus the main effects were ignored and a further investigation of the interaction was performed (Table 4.1). An equation describing the relationship between the response and the rate was generated for each treatment (Figure 4.10).

Table 4.1. ANOVA for treatment (compounds 1 – 5), rate (1, 10, 100 mol), and the interaction between treatment and rate.

Source SS df MS F P-value

Treatment 0.01124694 4 0.00281174 3.88 0.0117 Rate 0.73396072 2 0.36698036 506.92 < 0.0001 Treatment x Rate 0.20059674 8 0.02507459 34.64 < 0.0001

Error 0.02171806 30 0.00072394 Corrected Total 0.96752246 44

190

Figure 4.11. Growth inhibition activity of compounds 1 – 5 on radial mycelia growth of B. cinerea. Dashed lines represent steroidal glycoalkaloids and solid lines represent furostanol saponins.

191

All five compounds were weakly inhibitory to B. cinerea. The furostanol saponins, compounds 3 – 5, all had similar activity ranging from approximately 25 – 30% growth inhibition as compared to control at the highest concentration tested. The steroidal glycoalkaloid, compound 1, had similar inhibitory activity to the furostanol saponins. The steroidal glycoalkaloid, compound 2, had the highest inhibitory activity of 49.2 % at the highest concentration, approximately two times the activity of compound 1 (Figure

4.11). Steroidal glycoalkaloids 1 – 2 are similar in structure and only differ by the presence of an acetyl group linked to the C-6′′′ hydroxy position of the terminal glucose of carbohydrate moiety. The acetylation of the terminal glucose unit resulted in an increased rate of fungal growth inhibition, as compared to compound 1. Similar to the solanaceous glycoalkaloids, -chaconine and -tomatine, compounds 1 and 2 occur together as a pair, share the same aglycone, only differ in the carbohydrate moiety, and exhibit differential biological activity (Roddick et al., 2001). Friedman and MacDonald suggested that glycoalkaloid pairs may occur as a plant defense response to the adaptive ability of the pathogen to detoxify the plant‘s antifungal compounds (Friedman and

McDonald, 1997).

192

Figure 4.12. Growth inhibition activity of (A) compound 1 and (B) compound 2 on the radial mycelia growth of B. cinerea. At the highest concentration, the mycelia growth inhibitory activity was 24.9 and 49.2 % for compounds 1 and 2, respectively.

H N H N O O OH O OH O O HO O OH HO O O O OH HO O HO HO O O O OH HO 1 O 2

H3C O HO OH H3C O OH HO OH OH

A control 1 mol 10 mol 100 mol B control 1 mol 10 mol 100 mol 49.2 % 24.9%

193

4.4.2. Metabolism of compound 1 and 2 by B. cinerea.

Based on the observation that compound 2 had a two-fold increase in fungal growth inhibition as compared to compound 1, an investigation on the ability of B. cinerea to cleave the sugar residues of compounds 1 and 2 was conducted (Figure 4.12).

Figure 4.13. ESI+–MS mass spectra of steroidal glycoalkaloids 1 and 2.

Intens. 7 x10 884.9 H 1 N 4 O 576.4 414.6 OH 3 OH O HO O HO O O OH HO 2 O 738.5

576.4 H3C O 1 414.6 HO OH OH 738.5 0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Intens. 7 x10 926.9 H N 2 2.5 O O 576.4 414.3 2.0 O OH O HO O HO O O 1.5 OH HO O 780.5 1.0 414.3 H3C O HO OH OH 0.5 576.4 780.5 271.3 0.0 200 400 600 800 1000 1200 1400 1600 1800 m/z

After 48 hours of in vitro metabolism of compound 1 by B. cinerea, only trace quantities of compound 1 could be detected by LC-MS (Figure 4.13).

194

Figure 4.14. Metabolism of compound 1 by B. cinerea: (A) Total ion chromatogram (TIC) of fungal media spiked with compound 1 (time = 0) and (B) TIC of metabolites of compound 1 (48 hours).

Intens. x108 8 1 6

Intens. 8 6 x10

1.5

1.0

0.5 10 7 0.0

10.0 12.5 15.0 17.5 20.0 22.5 25.0 27.5 30.0 Time [min]

A metabolite of compound 1, compound 6, was observed with a based peak at m/z 722.8

(Figure 4.14). This ion was consistent with a molecule containing one less glucose molecule then compound 1. Additionally, ion fragments at 576.4 [M– Rha+H]+ and 414.5

[M–Glu–Rha+H]+ were observed and were consistent with a disaccharide moiety containing one glucose and one rhamnose molecule.

195

Figure 4.15. ESI+–MS mass spectra of compund 6.

Intens. 7 x10 6

1.5 722.8

1.0

0.5

414.5 576.4 0.0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Compound 6 was then isolated by semi-preparative RP-HPLC from a scale up fermentation and subjected to further chemical and spectroscopic analysis. Based on

ESI+–MS, and comparison of 1H NMR and 13C NMR with the literature, compound 6, was confirmed to be solasodine 3-O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranoside

(Mimaki and Sashida, 1990), previously isolated from the bulbs of Lilium brownii

(Figure 4.15).

196

Figure 4.16. Molecular structure and fragmentation of compound 6.

H N

O 414.5 OH O HO O HO 576.4 O

H3C O HO OH OH

In addition, another metabolite, compound 7, was observed with a base peak at m/z

576.7 (Figure 4.16).

Figure 4.17. ESI+–MS mass spectra of compund 7.

Intens. 6 x10 576.7 7 2.5

2.0

1.5

1.0

0.5 414.5 0.0 200 400 600 800 1000 1200 1400 1600 1800 m/z

197

The mass spectrum of the metabolite was consistent with a molecule containing one less glucose and one less rhamnose from compound 1 or one less rhamnose from compound

6. Additionally, an ion fragment at 414.5 [M–Glu+H]+ was observed and was consistent with a monosaccharide moiety containing one glucose molecule. Compound 7 was then isolated by semi-preparative RP-HPLC from a scale up fermentation and subjected to further chemical and spectroscopic analysis (Figures 6 and 7). Based on comparison of the retention time of the partial hydrolysis products of compound 1, ESI+–MS, and comparison of 1H NMR and 13C NMR with the literature, compound 7 was confirmed to be solasodine 3-O- -D-glucopyranoside (Bite and Rettegi, 1967; Kim et al., 1996), previously isolated from Solanum umbelliferum (Figure 4.17).

Figure 4.18. Molecular structure and fragmentation of compound 7.

H N

O 414.5 OH O HO O HO OH

A third metabolite, compound 10, was observed with a based peak at m/z 414.5 (Figure

4.18). The mass spectrum was consistent with a molecule containing the loss of two glucose units and one rhamnose from compound 1, one rhamnose and one glucose from

198 compound 6, or one glucose from compound 7. Additionally, an ion fragment at m/z

271.3 was observed.

Figure 4.19. ESI+–MS mass spectra of compound 10.

Intens. 7 414.5 x10 10 5

1 271.3 0 200 400 600 800 1000 1200 1400 1600 1800 m/z

Compound 10 was then isolated by semi-preparative RP-HPLC from a scale up fermentation and subjected to further chemical and spectroscopic analysis (Figures 6 and

7). Based on comparison of the retention time with an authentic standard, ESI+–MS, and comparison of 1H NMR and 13C NMR with the literature, compound 10 was confirmed to be solasodine (Bird et al., 1979), a common aglycone of steroidal glycoalkaloids (Figure

4.19).

199

Figure 4.20. Molecular structure and fragmentation of compound 10.

271.3 H N

HO HO m/z 271.3

Similar to the metabolism of -tomatine by Alternaria solani, sequential cleavage of all of the sugars of the carbohydrate moiety were observed in the model system (Schlösser,

1975). In addition to compounds 6, 7 and 10, several other fungal metabolites with differential degrees of glycosylation and regiospecific mono- and polyhydroxylation of the aglycone were observed.

The fungal metabolism of compound 2 by B. cinerea was markedly different than that of compound 1, as a result of acetylation of the 6′′′ hydroxy position of the terminal glucose unit. After 48 hours of in vitro metabolism of compound 2, most of compound 2 was still present in the media (Figure 4.20), whereas after 48 hours of metabolism of compound 1, only trace amounts were present (Figure 4.13).

200

Figure 4.21. Metabolism of compound 2 by B. cinerea: (A) Total ion chromatogram (TIC) of fungal media spiked with compound 2 (time = 0), (B) TIC of metabolites of compound 2 (48 hours), and (C) TIC of metabolites of compound 2 (72 hours).

Intens. x10 8

2.5

2.0 2

1.5

1.0

0.5

x10 8 2.5

2.0

1.5 2

1.0

0.5 7 8

0.0

x10 8 2.5

2.0

1.5

10 1.0 2

0.5 7 8

0.0 10 15 20 25 30 Time [min]

201

In contrast to the metabolism of compound 1, the metabolite compound 6 was not detected; however, a new metabolite, compound 8, was observed with a based peak at m/z 780.5 (Figure 4.21).

Figure 4.22. ESI+–MS mass spectra of compound 8.

Intens. 6 8 x10 780.5

1.5

1.0

0.5 414.5 576.4 0.0 200 400 600 800 1000 1200 1400 1600 1800 m/z

The mass spectrum of the metabolite was consistent a loss of rhamnose from compound 2. Additionally, ion fragments at 576.4 [M–Glu–Ac+H]+ and 414.5 [M–2Glu–

Ac+H]+ were observed and were consistent with a disaccharide moiety containing glucose and an acetylated glucose moiety. Based on mass spectral analysis of this metabolite, compound 8 is likely to be (22,R 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)-O-[6-O-acetyl- -D-glucopyranoside; however, preparative isolation and full characterization is needed for unequivocal confirmation (Figure 4.22).

202

Figure 4.23. Proposed molecular structure and fragmentation of compound 8.

H N

O O 576.4 414.5 O OH O HO O HO O O OH HO OH

In addition, compound 7 was detected as a metabolite of compound 2 and is consistent with the loss of one rhamnose and an acetylated glucose moiety of compound

2, or the loss of the acetylated glucose moiety of compound 8. Compound 10, was also observed and is consist with the metabolite of compound 1. The mass spectrum was consistent with a molecule containing the loss of one rhamnose, one glucose and an acetylated glucose from compound 2, the loss of one glucose and an acetylated glucose from compound 8, or one glucose from compound 7.

Due to the fact that only a small portion of compound 2 was metabolized after 48 hours of incubation, the metabolism experiment was continued for an additional 24 hours. After 72 hours of metabolism of compound 2 by B. cinerea, compound 2 was still present in the media (Figure 8) as compared to the metabolism of compound 1 which only trace amount were present after 48 hours (Figure 4). The decreased metabolism rate of compound 2 may play a role in the increased fungal growth inhibition of compound 2 as compared to compound 1. Interestingly, after 72 hours, compound 10 increased and

203 small amounts of compound 9 and compound 1 were detected (Figure 8). Compound 9 had a base peak m/z 738.8 (Figure 4.23).

Figure 4.24. ESI+–MS mass spectra of compound 9.

Intens. x10 6 738.8 9 1.25

1.00

0.75

0.50 576.4 0.25 414.5

0.00 200 400 600 800 1000 1200 1400 1600 1800 m/z

This ion was consistent with the de-acetylation of compound 8. Ion fragments at

576.4 [M–Glu+H]+ and 414.5 [M–2Glu+H]+ were observed and were consistent with a disaccharide moiety containing a two glucose moieties. Additionally, Compound 9 had the same retention time and mass spectrum as the product of the partial acid hydrolysis of compound 1 (Figure 4.24).

204

Figure 4.25. (A)Total ion chromatogram (TIC) of the metabolites of compound 2 (72 hours), (B) extracted ion chromatogram (EIC) for compound 9 (m/z 738.8) from the metabolite mixture derived from compound 2, and (C) TIC of the partial acid-catalyzed hydrolysis products of compound 1.

Intens.8 x10 2.5

2.0

1.5

10 1.0 2 9 0.5 7 8

0.0

6 9 x10

1.0

0.8

0.6

0.4

0.2

0.0

x107

3 7

9 2

10 1

0 10 15 20 25 30 Time [min]

205

Accordingly, compound 9 is likely to be (22,R 25R)-spirosol-5-en-3 -yl O- -D- glucopyranosyl-(1→4)]- -D-glucopyranoside; however, preparative isolation and full characterization is needed for unequivocal confirmation (Figure 4.25). The presence of small amounts of compound 1 and compound 9 after 72 hours of incubation suggests acetylase activity; however, due to the presence of a greater abundance of compound 8, the cleavage of the rhamnose moiety at the C-2′ position of the inner glucose is favored over de-acetylation of the acetyl moiety of the C-6′′′ position of the terminal glucose under these conditions.

Figure 4.26. Proposed molecular structure and fragmentation of compound 9

H N

O 576.4 414.5 OH OH O HO O HO O O OH HO OH

Based on these data, the metabolism of compound 1 occurs by the sequential removal of the sugars of the trisaccharide moiety with compounds 6, 7 and 10 as intermediates (pathway A; Figure 4.27). In parallel, hydroxylation of the aglycone occurs. Regiospecific microbial hydroxylation of diosgenin and solasodine is well known and is utilized for production of pharmaceutical steroids (Sato and Hayakawa, 1963a;

Sato and Hayakawa, 1963b). In tomato, B. cinerea has been shown to metabolize the steroidal glycoalkaloid, -tomatine, by both cleavage of the entire carbohydrate moiety

206 and by the cleavage of the terminal xylose (Verhoeff and Liem, 1975; Quidde et al.,

1998). Sequential sugar cleavage in compound 1 is similar to the metabolism of - tomatine by Alternaria solani, where all four sugars of the tetrasaccharide moiety are sequentially cleaved (Schlösser, 1975). In the case of compound 2, acetylation of the terminal glucose moiety inhibits cleavage from the inner glucose and metabolism proceeds through the cleavage the rhamnose at the C-2′ position of the inner glucose. The major metabolic pathway proceeds sequentially with compounds 8, 7, and 10 as intermediates (pathway B; Figure 4.26). Alternatively, evidence of acetylase activity was observed, and two minor metabolic pathways of compound 2 are proposed. One minor metabolic pathway may proceed with de-acetylation of compound 2 with compounds 8, 9, 7, and 10 as intermediates (pathway B1; Figure 4.26) and a second minor metabolic pathway may proceed with compounds 1, 6, 7, and 10 as intermediates

(pathway B2; Figure 4.26).

207

Figure 4.27. Proposed partial metabolic pathways for compounds 1 and 2 (thick arrows): (A) Major metabolic pathway for compound 1. (B) Major metabolic pathway for compound 2. (B1 and B2) Minor metabolic pathways for compound 2. De-Ac, de- acetylation; R-OH, mono/polyhydroxylation of aglycone.

H N H 2 1 N O O O O OH O OH HO O OH De-Ac HO O O O OH HO O HO HO O O O OH HO B O 2 H3C O HO OH H3C O OH HO OH OH A B H N H 8 6 N O O O R-OH O OH O R-OH HO O OH HO O O OH HO O HO O OH HO O

O De-Ac H3C B1 HO OH OH H N H 9 N 7 O O OH OH O OH HO O R-OH HO O O R-OH O OH HO HO O OH HO OH

H N 10

O R-OH

208

4.4.3. In planta identification of compounds 6 – 10 by LC-MS.

Based on the in vitro fungal growth inhibition studies and the characterization of the fungal metabolites of compounds 1 and 2, the objective of this part of the study was to determine if B. cinerea has the ability to metabolize compounds 1 and 2, in planta, into the fungal metabolites that were identified in the model system. In order to investigate this question, plant tissue that was infected with B. cinerea was compared to a control tissue by LC-MS analysis. All fungal metabolites that were characterized in the model system were detected in the infected plant tissue (Figure 4.27; Figure 4.28; Figure 4.29;

Figure 4.30; Figure 4.31). None of the metabolites were detected in the control tissue with the exception of compound 6. Although the infected tissue had elevated levels of compound 6 as compared to the control, interestingly, compound 6 was also present in the control sample that was not infected with B. cinerea (Figure 4.31).

209

Figure 4.28. (A) Extracted ion chromatograms (EIC) for compound 8 (m/z 780.5) of control plant tissue, and (B) plant tissue infected with B. cinerea.

x10 7 A

0 x10 7 B 1.5

1.0 8

0.5

0.0 10 15 20 25 30 Time [min]

210

Figure 4.29. (A) Extracted ion chromatograms (EIC) for compound 9 (m/z 738.8) of control plant tissue, and (B) plant tissue infected with B. cinerea.

7 x10 A

2.0

1.5

1.0

0.5

0.0

8 x10 B 9

1.0

0.5

0.0 10 15 20 25 30 Time [min]

211

Figure 4.30. (A) Extracted ion chromatograms (EIC) for compound 7 (m/z 576.7) of control plant tissue, and (B) plant tissue infected with B. cinerea.

x10 8 A 2.5

2.0

1.5

1.0

0.5

0.0 x10 8 B 7

1.5

1.0

0.5

0.0 10 15 20 25 30 Time [min]

212

Figure 4.31. (A) Extracted ion chromatograms (EIC) for compound 10 (m/z 414.6) of control plant tissue, and (B) plant tissue infected with B. cinerea.

x10 7 A 6

x108 B

4 10 3

0 10 15 20 25 30 Time [min]

213

Figure 4.32. (A) Extracted ion chromatograms (EIC) for compound 6 (m/z 722.8) of control plant tissue, and (B) plant tissue infected with B. cinerea.

x10 8 A

2 6

0 8 x10 B 3 6

0 10 15 20 25 30 Time [min]

214

4.4.4. Isolation and identification of compound 6 from L. Longiflorum bulbs.

In order to confirm the presence of compound 6 as a natural product in L.

Longiflorum, compound 6 was isolated and purified from L. Longiflorum bulbs. Briefly, lyophilized lily bulbs were washed with n-pentane and extracted with ethanol and water.

After the removal of solvent, the extract was dissolved in deionized water, washed with ethyl acetate and extracted with n-butanol yielding a crude steroidal glycoside extract.

The crude glycoside extract was fractionated by gel permeation chromatography and repeated semi-preparative RP-HPLC to yield compound 6 (Figure 4.32). Based on 1H

NMR and 13C NMR, compound 6 was confirmed as (22,R 25R)-spirosol-5-en-3 -yl O-

-L-rhamnopyranosyl-(1→2)- -D-glucopyranoside, previously isolated from L. brownii

(Mimaki and Sashida, 1990). These data confirms that compound 6 is not only a fungal metabolite of compounds 1 and 2, but it is also constitutively present in L. longiflorum.

Figure 4.33. Total ion chromatogram (TIC) of compound 6 isolated by RP-HPLC from L. longiflorum bulbs.

Intens. 8 x10 6

1.5

1.0

0.5

0.0

10 15 20 25 30 Time [min]

215

4.5. Conclusion

In this chapter, two steroidal glycoalkaloids and three furostanol saponins, isolated from L. longiflorum, were evaluated for fungal growth inhibition of the plant pathogenic fungus, B. cinerea. All compounds showed weak fungal growth inhibition activity. In addition, five fungal metabolites of the glycoalkaloids 1 and 2, were characterized from a model system and were observed in living plant tissue infected with the fungus. Furthermore, a structure-function relationship for the fungal growth inhibition for compounds 1 and 2 was established based on the acteylation of the terminal glucose moiety. On the basis of these results, B. cinerea can metabolize compounds 1 and 2 by the sequential removal of the sugars of the trisaccharide moiety. Additionally, these data suggests that a decreased rate of metabolism of compound 2 may play a role it the increased fungal growth inhibition activity. Moreover, compound 6 was determined to be both a fungal metabolite of compounds 1 and 2 and a natural product constitutively present in L. longiflorum. This is the first report of compound 6 from L. longiflorum. This study can be used as a model of characterizing fungal metabolites of plant derived natural products and a means for the generation of new natural products with novel biological activities. In addition, the antifungal activity of the compounds can be pursued further with pathogenic organisms not active in L. longiflorum.

216

Summary and Concluding Remarks

This research project was designed to: (1) Isolate and characterize new steroidal glycosides from the bulbs of L. longiflorum, (2) quantify their contents in all of the organs of L. longiflorum, and (3) perform studies on the antifungal activity and fungal metabolism of the compounds. The results of this study led to the discovery of new steroidal glycosides, a better understanding of their distribution within the different plant organs, and provided insight into structure-function relationships and fungal metabolism of the compounds.

Based on novel isolation methodologies and a combination of extensive spectroscopic and chemical analyses including 1D and 2D NMR, IR, HRESI–TOFMS,

ESI-MS, GC-MS, entantioselective GC-FID, chromatographic data, chemical analysis, and chemical transformations, several novel steroidal glycosides were isolated from L. longiflorum and structures elucidated. L. longiflorum contains two types of steroidal glycosides: steroidal glycoalkaloids and steroidal saponins. A new acetylated steroidal glycoalkaloid and two new furostanol saponins, along with three known steroidal glycosides, were isolated from the bulbs of L. longiflorum. The new steroidal glycoalkaloid was identified as (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-

(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside. The new furostanol saponins were identified as (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-

3 22 -triol 3-O- -L-rhamnopyranosyl-(1→2)- -L-arabinopyranosyl-(1→3)- -D- glucopyranoside and (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol 3-O-

-L-rhamnopyranosyl-(1→2)- -L-xylopyranosyl-(1→3)- -D-glucopyranoside.

Additionally, two known steroidal glycoalkaloids, (22R, 25R)-spirosol-5-en-3 -yl O- -

217

L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside and (22,R

25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranoside, and a known furostanol saponin, (25R)-26-O-( -D-glucopyranosyl)-furost-5-en-3 22 -triol

3-O- -L-rhamnopyranosyl-(1→2)- -D-glucopyranosyl-(1→4)- -D-glucopyranoside, were isolated from L. longiflorum for the first time.

In order to investigate the natural distribution of the newly identified compounds within the plant, a LC-MS/MS method performed in multiple reaction monitoring

(MRM) mode was developed for the simultaneous quantitative analysis of the five new steroidal glycosides in the different organs of L. longiflorum. The highest concentrations of total steroidal glycosides were detected in flower buds, lower stems, and leaves. The steroidal glycoalkaloids were detected in higher concentrations as compared to the furostanol saponins in all of the plant organs except for the fibrous and fleshy roots. The proportions of steroidal glycoalkaloids to furostanol saponins were higher in the plant organs exposed to light and decreased in proportion from the aboveground organs to the underground organs. The highest concentrations of the steroidal glycoalkaloids were detected in flower buds, leaves, and bulbs. Both steroidal glycoalkaloids had a similar pattern of distribution in the various plant organs; however, the acetylated derivative occurred at lower levels. The furostanol saponins were detected in the highest concentrations in the lower stems, fleshy roots, and flower buds. In the bulbs, the steroidal glycosides were not distributed uniformly throughout the bulb scale tissue, but accumulated in the basal plate, bulb scale epidermal cells, and vascular bundles. This work has led to a better understanding of the natural distribution of these compounds within the different plant organs and plant tissues of L. longiflorum.

218

To gain insight into the plant-pathogen interaction, purified steroidal glycosides were evaluated for fungal growth inhibition activity against the plant pathogenic fungus,

Botrytis cinerea, using an in vitro plate assay. All of the compounds showed weak fungal growth inhibition activity; however, the natural acetylation of C-6′′′ of the terminal glucose in the steroidal alkaloid, (22R, 25R)-spirosol-5-en-3 -yl O- -L- rhamnopyranosyl-(1→2)-[6-O-acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside, increased the antifungal activity by inhibiting the rate of metabolism of the compound by the B. cinerea. Acetylation of the glycoalkaloid may be a plant defense response to the evolution of detoxifying mechanisms by the pathogen. A model system was developed to generate fungal metabolites of the steroidal glycoalkaloids which led to the identification of several new fungal metabolites. The fungal metabolites characterized from the model system were subsequently identified by LC-MS to naturally occur in Easter lily tissues infected with the fungus.

The extraction and purification procedures reported in this work may be used for the production of sufficient quantities of pure compounds for biological investigations.

These new compounds from L. longiflorum can be used for studies on the biological role of steroidal glycosides in plant development and plant-pathogen interactions, as well as for studies in food and human health. Furthermore, the quantitative analysis of steroidal glycosides in the different organs of L. longiflorum is the first step to developing insight into the biological role these compounds play in plant metabolism, plant development, and plant-pathogen interactions. An understanding of the distribution of steroidal glycosides in the different organs of L. longiflorum may aid in the development of optimized extraction and purification methodologies for food, health, and industrial

219 applications. Moreover, the results from the fungal metabolism work can be used as a model for characterizing fungal metabolites of plant derived defense compounds, gaining insight into plant-pathogen interactions, and a means for the generation of new natural products with novel biological activities. The structural modification of plant defense compounds may be an adaptive response to plant pathogens and the acetylation of the glycoalkaloid, (22R, 25R)-spirosol-5-en-3 -yl O- -L-rhamnopyranosyl-(1→2)-[6-O- acetyl- -D-glucopyranosyl-(1→4)]- -D-glucopyranoside, may be a plant defense response to the evolution of detoxifying mechanisms by the pathogen. A plant defense strategy aimed at inhibiting the pathogen‘s ability to detoxify plant defense compounds may be an alternative strategy in plant defense. The characterization of detoxification pathways utilized by plant pathogenic fungi helps to provide a fundamental understanding of the plant-pathogen interaction and suggests strategies to chemically modify these anti-fungal compounds to prevent pathogen degradation.

220

Literature Cited

Abe, I.; Rohmer, M.; Prestwich, G. Enzymatic cyclization of squalene and oxidosqualene to sterols and triterpenes. Chem. Rev. 1993, 93, 2189–2206.

Acharya, D.; Mitaine-Offer, A.; Kaushik, N.; Miyamoto, T.; Paululat, T.; Mirjolet, J.; Duchamp, O.; Lacaille-Dubois, M.; Cytotoxic spirostane-type saponins from the roots of Chlorophytum borivilianum. J. Nat. Prod. 2009, 72, 177–181.

Achenbach, H.; Hübner, H.; Brandt, W.; Reiter, M. Cardioactive steroid saponins and other constituents from the aerial parts of Tribulus cistoides. Phytochemistry 1994, 35, 1527–1543.

Adel, M.; Sehnal, F.; Jurzysta, M. Effects of alfalfa saponins on the moth Spodoptera littoralis. J. Chem. Ecol. 2000, 26, 1065–1078.

Agrawal, P. NMR spectroscopy of steroidal sapogenins and steroidal saponins: An update. Mag. Res. Chem. 1995, 33, 923–953.

Ahmad, V.; Khaliq-uz-Zaman, S.; Shameel, S.; Perveen, S.; Ali, Z. Steroidal saponins from Asparagus dumosus. Phytochemistry 1998, 50, 481–484.

Akhov, L.; Musienko, M.; Piacente, S.; Pizza, C.; Oleszek, W. Structure of Steroidal saponins from underground parts of Allium nutans L. J. Agric. Food Chem. 1999, 47, 3193–3196.

Anthony, N.; Balog, J.; Staudinger, F.; Wall, C.; Walker, R.; Huff, W. Effect of urease inhibitor and ceiling fans on ascites in broilers. 1. Environmental variability and incidence of ascites. Poult. Sci. 1994, 73, 801–809.

Armah, C.; Mackie, A.; Roy, C.; Price, K.; Osbourn, A.; Bowyer, P.; Ladha, S. The membrane-permeabilizing effect of avenacin A-1 involves the reorganization of bilayer cholesterol. Biophys. J. 1999, 76, 281–290.

Arneson, P.; Durbin, D. Hydrolysis of tomatine by Septoria lycopersici: a detoxifcation mechanism. Phytopathology 1967, 57, 135–1360.

Atta-ur-Rahman; Choudhary, I. Chemistry and Biology of Steroidal Alkaloids. In The Alkaloids: Chemistry and Biology, Vol 50; Academic Press: San Diego, CA, 1998; pp 61–108.

Bahuguna, S.; Kishor, N.; Sati, O.; Sati, S.; Sakakibara, J.; Kaiya, T. A new spirostanol glycoside from Yucca aloifolia. J. Nat. Prod. 1991, 54, 863–865.

Balog, J.; Anthony, N.; Wall, C.; Walker, R.; Rath, N.; Huff, W. Effect of urease

221

inhibitor and ceiling fans on ascites in broilers. 2. Blood variables, ascites scores and body and organ weights. Poult. Sci. 1994, 73, 810–816.

Bar-Nun, N.; Tal Lev, A.; Harel, E.; Mayer, A. Repression of lactase formation in Botrytis cinerea and its possible relation to phytopathogenicity. Phytochemistry 1988, 27, 2505–2509.

Müller, K.; Börger, H. Experimentelle untersuchungen über die Phytophthora-resistenz der kartoffel. Arb. Biol. Reichsasnstalt. Landw. Forstw. Berlin 1940, 23, 189– 231.

Bird, G.; Collins, D.; Eastwood, F.; Exner, R.; Romanelli, M.; Small, D. The synthesis of 2 2 (25R)-22 N-[15 ,17 - H2]-spirosol-5-en-3 -ol ([15 ,17 - H2] Solasodine) and assignment of the 13C NMR spectra of solasodine and its derivatives. Aust. J. Chem. 1979, 32, 783–796.

Bite, P.; Rettegi, T. Preparation of steroid glycosides. Acta Chim. Acad. Sci. Hung. 1967, 52, 79–82.

Blankemeyer, J.; Stringer. B.; Rayborn. J.; Bantle. J.; Friedman, M. Effect of potato glycoalkaloids, -chaconine and -solanine, on membrane potential of frog embryos. J. Agric. Food Chem. 1992, 40, 2022–2025.

Blankemeyer, J.; Atherton, R.; Friedman, M. Effect of potato glycoalkaloids chaconine and solanine on sodium active transport in frog skin. J. Agric. Food Chem. 1995, 43, 636–639.

Blankemeyer, J.; McWilliams, M.; Rayburn, J.; Weissenebrg, M.; Friedman, M. Developmental toxicology of solamargine and solasonine glycoalkaloids in frog embryos. Food Chem. Toxicol. 1998, 36, 383–389.

Block, E. The chemistry of garlic and onions. Sci. Am. 1985, 252, 114–119.

Braca, A.; Prieto, J.; De Tommasi, N.; Tome`, F.; Morelli, I. Furostanol saponins and quercetin glycosides from the leaves of Helleborus viridis L. Phytochemistry 2004, 65, 2921–2928.

Bowyer, P.; Clarke, B.; Lunness, P.; Daniels, M.; Osbourn, A. Host range of a plant pathogenic fungus determined by a saponin detoxifying enzyme. Science 1995, 267, 371–374.

Capasso, F.; Gaginella, T.; Grandolini, G.; Izzo, A. 2003. Phytotherapy: a quick reference to herbal medicine. Springer, Germany.

Carter, S.; Lake, R. Method using glycoalkaloids for modulating interleukin 6, and therapeutic use. 2004, PCT Int Appl 39 pp.

222

Cheeke, P. Saponins: Surprising benefits of desert plants. In: The Linus Pauling Institute Newsletter. Oregon State University, Corvallis. 1998; 4–5.

Cheeke, P. Actual and potential applications of Yucca schidigera and Quillaja saponaria saponins in human and animal nutrition. J. Anim. Sci. 2000, 77, 1–10.

Choi, D; Bostock, R; Avdiushko, S; Hildebrand, D. Lipid-derived signals that discriminate wound- and pathogen-responsive isoprenoid pathways in plants: methyl jasmonate and the fungal elicitor arachidonic acid induce different 3- hydroxy-3-methylglutaryl-coenzyme A reductase genes and antimicrobial isoprenoids in Solanum tuberosum L. Plant Mol. Biol. 2005, 57, 225–239.

Corea, G.; Fattorusso, E.; Lanzotti, V. Saponins and flavonoids of Allium triquetrum. J. Nat. Prod. 2003, 66, 1405–1411.

Cui, C.; Xu, C.; Gu, Q.; Chu, S.; Ji, H.; Jing, G. A new furostanol saponin from the water-extract of Dioscorea nipponica Mak., the raw material of the traditional Chinese herbal medicine Wei Ao Xin. Chin. Chem. Lett. 2004, 15, 1191–1194.

Curir, P.; Dolci, M.; Pasini, C.; Capponi, A.; Galeotti, F.; Grassotti, A. Contenuto in saponine nel Lilium quale indice della resistenza a Fusarium oxysporum f. sp. lilii. Italus Hortus 2003, 10, 291–293.

Daly, J.; Spande, T.; Garraffo, H. Alkaloids from Amphibian Skin: A Tabulation of Over Eight-Hundred Compounds. J. Nat. Prod. 2005, 68, 1556–1575.

Debella, A.; Haslinger, E.; Kunert, O.; Michl, G.; Abebe, D. Steroidal saponins from Asparagus africanus. Phytochemistry 1999, 51, 1069–1075.

Dinan, L.; Harmatha, J.; Lafont, R. Chromatographic procedures for the isolation of plant steroids. J. Chromat. A 2001, 953, 105–123.

Dragalin, I.; Kintia, P. Steroidal saponins of Yucca filamentosa. Phytochemistry 1975, 14, 1817–1820.

Dong, M.; Feng, X.; Wang, B.; Ikejima, T.; Wu, L. Steroidal saponins from Dioscorea panthaica and their cytotoxic activity. Pharmazie 2004, 59, 294–296

Eanes, R.; Tek, N. Solid-phase microextraction (SPME) followed by on-fiber derivatization of solasodine and solanidine aglycones of steroidal glycoalkaloids. J. Liq. Chrom. Relat. Tech. 2008, 31, 1132–1146.

Eisenreichova, E.; Haladova, M.; Mucaji, P.; Budesinsky, M.; Ubik, K. A new steroidal saponin from the bulbs of Lilium candidum L. Pharmazie. 2000, 55, 549–550.

223

Eltayeb, E; Al-Ansari, A.; Roddick, J. Changes in the steroidal alkaloid solasodine during development of Solanum nigrum and Solanum incanum. Phytochemistry 1997, 46, 489–494.

Erdoğan, I.; Sener, B.; Atta-ur-Rahman, A. Etioline, a steroidal alkaloid from Lilium candidum L. Biochem. Sys. Ecol. 2001, 29, 535–536.

Espejo, O.; Campos, L.; Jung, H.; Giral, F. Spirostanic diosgenin precursors from Dioscorea composita tubers yams. Phytochemistry 1982, 21, 413–416.

Espeso, E.; Guevara, B. Studies of the alkaloids of Lilium philippinense Baker. Acta Manilana 1990, 38, 83–87.

Fattorusso, E.; Lanzotti, V.; Magno, S.; Taglialatela-Scafati, O.; Di Rosa, M.; Ianaro, A. Cytotoxic saponins from bulbs of Allium porrum L. J. Agric. Food Chem. 2000, 46, 4904–4908.

Fattorusso, E.; Iorizzi, M.; Lanzotti, V.; Taglialatela-Scafati, O. Chemical composition of shallot (Allium ascalonicum Hort.). J. Agric. Food Chem. 2002, 50, 5686–5690.

Ferro, E.; Alvarenga, N.; Ibarrola, D.; Hellio´n-Ibarrola, M.; Ravelo, A. A new steroidal saponin from Solanum sisymbriifolium roots. Fitoterapia 2005, 76, 577–579.

Ford, J.; McCance, D.; Drysdale, R. The detoxifcation of -tomatine by Fusarium oxysporum f. sp. lycopersici. Phytochemistry 1977, 16, 545–546.

Francis, J.; Rumbeiha, W.; Nair, M. Constituents in Easter lily flowers with medicinal activity. Life Sci. 2004, 76, 671–683.

Friedman, M.; McDonald, G. Potato glycoalkaloids: chemistry, analysis, safety and plant physiology. Crit. Rev. Plant Sci. 1997, 16, 55–132.

Friedman, M.; Fitch, T.; Yokoyama, W. Lowering of plasma LDL cholesterol in hamsters by the tomato glycoalkaloid tomatine. Food Chem. Toxicol. 2000, 38, 549–553.

Fukuzawa, S.; Matsunaga, S.; Fusetani, N. Ritterazine A, a highly cytotoxic dimeric steroidal alkaloid, from the tunicate Ritterea tokiokal. J. Org. Chem. 1994, 59, 6164–6166.

Gaffield, W.; Keeler, R. Induction of terata in hamsters by solanidane alkaloids derived from Solanum tuberosum. Chem. Res. Toxicol. 1996, 9, 426–433.

Ghisalberti, E. Steroidal glycoalkaloids: Isolation, structure, analysis and biosynthesis. Nat. Prod. Commun. 2006, 10, 859–884.

224

Gilani, H.; Shaheen, F.; Christopoulos, A.; Mitchelson, F. Interaction of ebeinone, an alkaloid from Fritillaria imperialis, at two muscarinic acetylcholine receptor subtypes. Life Sci. 1997, 60, 535–544.

Gosse, B.; Gnabre, J.; Bates, R.; Dicus, C.; Nakkiew, P.; Huang, R. Antiviral saponins from Tieghemella heckelii. J. Nat. Prod. 2002, 65, 1942–1944.

Grassert, V.; Lellbach, H. Investigations of glycoalkaloid content of potato hybrids resistant to potato cyst eelworms Globodera rostochiensis and Globodera pallida. Biochem. Physiol. Pflanzen. 1987, 182, 473-479.

(GRIN) Germplasm Resources Information Network (2011). Published on the Internet http://www.ars-grin.gov

Güçlü-Üstündağ, Ö.; Mazza, G. Saponins: Properties, Applications and Processing. Crit. Rev. Food Sci. Nutr. 2006, 47, 231–258.

Guo, S.; Kenne, L.; Lundgren, L.; Rönnber, B.; Sundquist, B. Triterpenoid saponins from Quillaja Saponaria. Phytochemistry 1998, 48, 175–180.

Gureva, A.; Kintya, P.; Mashchenko, N.; Shashkov, A. Glycosides of the bulbs of Lilium regale. Chem. Nat. Compd. 1996, 32, 1573–8388.

Gurielidze, K.; Gogoberidze, M.; Dadeshidze, I.; Vardosanidze, M.; Djaoshvili, M.; Lomkatsi, N. Tissue and subcellular localization of oligofurostanosides and their specific degrading -glucosidase in Dioscorea caucasica Lipsky. Phytochemistry 2004, 65, 555–559.

Hansen, J. Steroids: Partial synthesis in medicinal chemistry. Nat. Prod. Rep. 2007, 24, 1342–1349.

Haraguchi, M.; Zaccharias, D.; Young, M.; Chu, E. Steroidal prosapogenins from Dioscorea olfersiana. Phytochemistry 1994, 36, 1005–1008.

Haralampidis, K., Trojanowska, M., Osbourn, A. Biosynthesis of triterpenoid saponins in plants. Adv. Biochem. Eng. 2002, 75, 31–49.

Harmatha, J.; Mauchamp, B.; Arnault, C.; Sla´ma, K. Identification of a spirostane-type saponin in the flowers of leek with inhibitory effects on growth of leek-moth larvae. Biochem. Syst. Ecol. 1987, 15, 113–116.

Hayes, P.; Jahidin, A.; Lehmann, R.; Penman, K.; Kitching, W.; De Voss, J. Steroidal saponins from the roots of Asparagus racemosus. Phytochemistry 2008, 69, 796– 804.

225

Hiramatsu, M.; Ii, K.; Okubo, H.; Huang, K.; Huang, C. Biogeography and origin of Lilium longiflorum and L. formosanum (Liliaceae) endemic to the Ryukyu Archipelago and Taiwan as determined by allozyme diversity. Am. J. Bot. 2001, 88, 1230–1239.

Hou, X.; Chen, F. Studies on chemical constituents of Lilium brownii. Yao-Xue-Xue-Bao. 1998, 33, 923–926.

Hu, K.; Yao, X. The cytotoxicity of methyl protoneogracillin (NSC-698793) and gracillin (NSC-698787), two steroidal saponins from the rhizomes of Dioscorea collettii var. hypoglauca, against human cancer cells in vitro. Phytother. Res. 2003, 17, 620–626.

Huang, H.; Tsai, W.; Morris-Natschke, S.; Tokuda, H.; Lee, K.; Wu, Y.; Kuo, Y. Sapinmusaponins F-J, bioactive tirucallane-type saponins from the galls of Sapindus mukorossi. J. Nat. Prod. 2006, 69, 763–767.

Huang, X.; Kong, L. Steroidal saponins from roots of Asparagus officinalis. Steroids 2006, 71, 171–176.

Hughes, H.; Morrissey, J.; Osbourn, A. Characterisation of the saponin hydrolysing enzyme avenacoside-α-L-rhamnosidase from the fungal pathogen of cereals, Stagonospora avenae. Eur. J. Plant Pathol. 2004, 110, 421–427.

Ikeda, T.; Tsumagari, H.; Honbu, T.; Nohara, T. Cytotoxic activity of steroidal glycosides from Solanum plants. Biol. Pharm. Bull. 2003, 26, 1198–1201.

Itoh, T.; Tamura, T.; Mitsuhashi, T.; Matsumoto, T. Sterols of Liliaceae. Phytochemistry 1977, 16, 140–141.

Ishizaka, N.; Tomiyama, K. Effect of wounding or infection by Phytophthora infestans on the contents of terpenoids in potato tubers. Plant Cell. Physiol. 1972, 13, 1053–1063.

Ji, H.; Ding, X.; Tao, G. Screening of steroidal saponins from the bulbs of Lilium brownii var. colchesteri by combination of high performance liquid chromatography- electrospray ionization mass spectrometry and electron impact mass spectrometry. Se-Pu. 2001, 19, 403–406.

Jiang, Y.; Li, H.; Li, P.; Cai, Z.; Ye, W. Steroidal alkaloids from the bulbs of Fritillaria puqiensis. J. Nat. Prod. 2005, 68, 264–267.

Keukens, E.; de Vrije, T.; Fabrie, C.; Demel, R.; Jongen, W.; de Kruijff, B. Dual specificity of sterol-mediated glycoalkaloid induced membrane disruption. Biochim. Biophys. Acta 1992, 1110, 127–136.

226

Keukens, E.; de Vrije, T.; van den Boom, C.; de Waard, P.; Plasman, H.; Thiel, F.; Chupin, V.; Jongen, W.; de Kruijff, B.. Molecular basis of glycoalkaloid induced membrane disruption. Biochim. Biophys. Acta 1995, 1240, 216–228.

Kim, G.; Kim, H.; Seong, J.; Oh, S.; Lee, C.; Bang, J.; Seong, N.; Song, K. Cytotoxic steroidal saponins from the rhizomes of Asparagus oligoclonos. J. Nat. Prod. 2005, 68, 766–768.

Kim, Y.; Che, Q.; Gunatilaka, A.; Kingston, D. Bioactive steroidal alkaloids from Solanum umbelliferum. J. Nat. Prod. 1996, 59, 283–285.

Kimura, M. Quaternary paleogeography of the Ryukyu Arc. J. Geograph. 1996, 105, 259–285.

Kintya, P.; Gur'eva, A.; Mashchenko, N.; Shashkov, A. Structures of some lilioglycosides from the bulbs of Lilium Regale. Chem. Nat. Compd. 1997, 33, 658–662.

Kitajima, J.; Noda, N.; Ida, Y.; Komori, T.; Kawasaki, T. Chem. Pharm. Bull. 1982, 30, 3922–3931.

Kizaki, K. and Oshiro, I. Paleogeography of the Ryukyu Islands. Mar. Sci. Mon. 1977, 9, 543–549.

Kuljanabhagavad, T.; Thongphasuk, P.; Chamulitrat, W.; Wink, M. Triterpene saponins from Chenopodium quinoa Willd. Phytochemistry 2008, 69, 1919–1926.

Kuo, K.; Hsu, S.; Li, Y.; Lin, W.; Liu, L.; Chang, L.; Lin, C.; Lin, C.; Sheu, H. Anticancer activity evaluation of the Solanum glycoalkaloid solamargine. Triggering apoptosis in human hepatoma cells. Biochem. Pharmacol. 2000, 60, 1865–1873.

Lacaille-Dubois, M. A propos des Dioscorea (le ‗‗yam‘‘). La Phytothe´rapie Europe´enne 2002, 7, 9–16.

Langston, C. Acute renal failure caused by lily ingestion in six cats. J. Am. Vet. Med. Assoc. 2002, 220, 49–52.

Lanzotti, V. Bioactive saponins from Allium and Aster plants. Phytochemistry Reviews 2005, 4, 95–110.

Lanzotti, V. The analysis of onion and garlic. J. Chromat. A 2006, 21, 3–22.

Li, B.; Zhou, Z. Chemistry of the new drug Di-ao-xin-xue-kang treating angina pectoris. New Drug Clin. Remed. 1994, 13, 75–76.

227

Li, H.; Jiang, Y.; Li, P. Chemistry, bioactivity and geographical diversity of steroidal alkaloids from the Liliaceae family. Nat. Prod. Rep. 2006, 23, 735–752.

Li, J.; Hu, Z. Accumulation and dynamic trends of triterpenoid saponin in vegetative organs of Achyranthus bidentata. J. Integr. Plant Biol. 2008, 51, 122–129.

Li, T.; Zhou, R.; Zhao, S.; Hu, L. Clinical studies on Zu-shi-ma injection. New Trad. Chin. Med. Clin. Pharmacol. 2000, 11, 266–268.

Lin, C.; Chung, M.; Lin, S. Steroidal alkaloids fro Solanum capsicastrum. Phytochemistry 1986, 26, 305–307.

Lin, J.; Yang, D. Determination of steroidal saponins in different organs of yam (Dioscorea pseudojaponica Yamamoto). Food Chem. 2008, 108, 1068–1074.

Lin, R.; Hanquet, B.; Lacaille-Dubois, M. Aferoside A, a steroidal saponin from Costus afer. Phytochemistry 1996, 43, 665–668.

Link, K.; Walker, J. The isolation of catechol from pigmented onion scales and its significance in relation to disease resistance in onions. J. Biol. Chem. 1933, 100, 379–383.

Liu, H.; Wang, S.; Cai, B.; Qu, G.; Yang, X.; Kobayashi, H.; Yao, X. New furostanol glycosides from the rhizomes of Dioscorea futschauensis R. Kunth. J. Asian Nat. Prod. Res. 2003, 5, 241–247

Mader, T.; Brumm, M. Effect of feeding sarsaponin in cattle and swine diets. J. Anim. Sci. 1987, 65, 9–15.

Marker, R.; Tsukamoto, T.; Turner, D. Sterols. C. Diosgenin. J. Am. Chem. Soc. 1940, 62, 2525–2532

Matsuura, H. Saponins in garlic as modifiers of the risk of cardiovascular disease. J. Nutr. 2001, 131. 1000–1005.

McAvoy, R. The 2010 Easter lily schedule. University of Connecticut: Storrs, CT, 2010; 4.

McKee, R. Host-parasite relationships in the dry-rot disease of potatoes. Ann. appl. Biol. 1955, 43, 147.

McRae, E. Diseases. In Lilies: a guide for growers and collectors. Timber Press: Portland, OR, 1998; 84–85.

Mimaki, Y.; Sashida, Y. Steroidal saponins from the bulbs of Lilium brownii. Phytochemistry 1990a, 29, 2267–2271.

228

Mimaki, Y.; Sashida, Y. Steroidal saponins and alkaloids from the bulbs of Lilium brownii var. colchesteri. Chem. Pharm. Bull. (Tokyo) 1990b, 31, 3055–3059.

Mimaki, Y.; Sashida, Y. Steroidal and phenolic constituents of Lilium speciosum. Phytochemistry 1991, 30, 937–940.

Mimaki,Y.; Ishibashi, N.; Ori, K.; Sashida, Y. Steroidal glycosides from the bulbs of Lilium dauricum. Phytochemistry 1992, 31, 1753–1758.

Mimaki,Y.; Kuroda, M.; Fukasawa, T.; Sashida, Y. Steroidal glycosides from the bulbs of Allium jesdianum. J. Nat. Prod. 1999, 62, 194–197.

Mimaki, Y.; Kuroda, M.; Sashida, Y. Steroidal saponins from the bulbs of Allium aflatunense. Nat. Med. 1999a, 53, 88–93.

Mimaki, Y.; Kuroda, M.; Sashida, Y. Steroidal saponins from the bulbs of Allium ampeloprasum. Nat. Med. 1999b, 53, 134–137.

Mimaki, Y.; Nakanura, O.; Sashida, Y.; Nishino, A.; Nishino, H. Steroidal saponins from the bulbs of Lilium longiflorum and their antitumor-promoter activity. Phytochemistry 1994, 37, 227–232.

Mimaki, I.; Sashida, Y.; Shimomura, H. Lipid and steroidal constituents of Lilium auratum var. platyphyllum and Lilium tenuifolium. Phytochemistry 1989, 38, 3453–3458.

Mimaki, Y.; Sashida, Y.; Nakamura, O.; Nikaido, T.; Ohmoto, T. Steroidal saponins from the bulbs of Lilium regale and Lilium henryi. Phytochemistry 1993, 33, 675–682.

Mimaki, Y.; Satou, T.; Kuroda, M.; Sashida, Y.; Hatakeyama, Y. New steroidal constituents from the bulbs of Lilium candidum. Chem. Pharm. Bull. (Tokyo) 1998, 46, 1829–1832.

Mimaki, Y.; Satou, T.; Kuroda, M.; Sashida, Y.; Hatakeyama, Y. Steroidal saponins from the bulbs of Lilium candidum. Phytochemistry 1999, 51, 567–573.

Mimaki, Y.; Yokosuka, A.; Kuroda, M.; Sashida, Y. Cytotoxic activities and structure- cytotoxic relationships of steroidal saponins. Biol. Pharm. Bull. 2002, 24, 1286– 1289.

Morita, T.; Ushiroguchi, T.; Hayashi, N.; Matsuura, H.; Itakura, Y.; Toru, F. Steroidal saponins from elephant garlic, bulbs of Allium ampeloprasum L. Chem. Pharm. Bull. 1988, 36, 3480–3486.

229

Morrissey, J.; Osbourn, A. Fungal resistance to plant antibiotics as a mechanism of pathogenesis. Microbiol. Mol. Biol. Rev. 1999, 63, 708–724.

Morrissey, J.; Wubben, J.; Osbourn, A. Stagonospora avenae secretes multiple enzymes that hydrolyze oat leaf saponins. M. P. M. I. 2000, 13, 1041–1052.

Munafo, J.; Ramanthan, A.; Jimenez, L.; Gianfagna, T. Isolation and structural determination of steroidal glycosides from the bulbs of Easter Lily (Lilium longiflorum Thunb.). J. Agric. Food Chem. 2010, 58, 8806–8813.

Munafo, J.; Gianfagna, T. Quantitative Analysis of Steroidal Glycosides in Different Organs of Easter Lily (Lilium longiflorum Thunb.) by LC-MS/MS. J. Agric. Food Chem. 2010, 59, 995–1004.

Murakami, T.; Kishi, A.; Matsuda, H.; Yoshikawa, M. Medicinal Foodstuffs. XVII.1) fenugreek seed. (3): Structures of new furostanol-type steroid saponins, trigoneosides Xa, Xb, XIb, XIIa, XIIb, and XIIIa, from the Seeds of Egyptian Trigonella foenum-graecum L. Chem. Pharm. Bull. 2000, 48, 994—1000.

Nakamura, O.; Mimaki, Y.; Nishino, H.; Sashida, Y. Steroidal saponins from the bulbs of Lilium speciosum x L. nobilissimum 'Star Gazer' and their antitumour-promoter activity. Phytochemistry 1994, 36, 463–467.

Nakano, K.; Niahizawa, K.; Murakami, K.; Takaishi, Y.; Tomimatsu. Steroidal alkaloid glycosides from Lilium cordatum. Phytochemistry 1987, 26, 301–303.

Nakano, K.; Midzuta, Y.; Hara, Y.; Murakami, K.; Takaishi, Y.; Tomimatsu, T. 12-keto steroidal glycosides from the caudex of Yucca gloriosa. Phytochemistry 1991, 30, 633–636.

Nakashima, N.; Kimura, I.; Kimura, M.; Matsuura, H. Isolation of pseudoprototimosaponin AIII from rhizomes of Anemarrhena asphodeloides and its hypoglycemic activity in streptozotocin-induced diabetic mice. J. Nat. Prod. 1993, 56, 345–350.

Nicol, R.; Traquair, J.; Bernards, M. Ginsenosides as host resistance factors in American ginseng (Panax quinquefolius) Can. J. Bot. 2002, 80, 557–562.

Nørbæk, R.; Kondo, T. Anthocyanins from flowers of Lilium (Liliaceae). Phytochemistry 1999, 50, 1181–1184.

Noor-e-Ain, F. Studies on the steroidal alkaloids from Buxus longifolia and other related plants. 1996. PhD thesis, University of Karachi, Karachi.

Nozzolillo, C.; Arnason, J.; Campos, F.; Donskov, N.; Jurzysta, M. Alfalfa leaf saponins and insect resistance. J. Chem. Ecol. 1997, 23, 995–1002.

230

Oh, H.; Kang, D.; Lee, S.; Lee, Y.; Lee, H. Angiotensin converting enzyme (ACE) inhibitory alkaloids from Fritillaria ussuriensis. Planta Med. 2003, 69, 564–565.

Oketch-Rabah, H.; Dossaji, S. Antiprotozoal compounds from Asparagus africanus. J. Nat. Prod. 1997, 60, 1017–1022.

Oleszek, W.; Bialy, Z.; Chromatographic determination of plant saponins—An update (2002–2005). J. Chromat. A 2006, 1112, 78–91.

Oleszek, W.; Sitek, M.; Stochmal, A.; Piacente, S.; Pizza, C.; Cheeke, P. Steroidal saponins of Yucca schidigera Roezl. J. Agric. Food Chem. 2001, 49, 4392–4396.

Ori, K.; Mimaki, Y.; Mito, K.; Sashida, Y.; Nikaido, T.; Ohmoto, T.; Masuko, A. Jatropham derivatives and steroidal saponins from the bulbs of Lilium Hansonii. Phytochemistry 1992, 38, 2767–75.

Osbourn, A. Preformed antimicrobial compounds and plant defense against fungal attack. Plant Cell, 1996, 8, 1821–1831.

Osbourn, A. Antimicrobial phytoprotectants and fungal pathogens: A commentary. Fungal Genet. Biol. 1999, 26, 163–168.

Osbourn, A. Saponins in cereals. Phytochemistry 2003, 62, 1–4.

Osbourn, A.; Qi, X.; Townsend, B.; Qin, B. Dissecting plant secondary metabolism: Constitutive chemical defenses in cereals. New Phytologist 2003, 59, 101–108.

Osman, S.; Johns, T.; Price, K. Sisunine, a glycoalkaloid found in hybrids between Solanum acaule and Solanum x ajanhuiri. Phytochemistry 1986, 25, 967-968.

Osorio, J.; Martinez, O.; Navarro, Y.; Mimaki, Y.; Sakagami, H.; Sashida, Y. Diospolysaponin A, a new polyoxygenated spirostanol saponin from the tubers of Dioscorea polygonoides. Heterocycles 2003, 60, 1709– 1715

Papadopoulou, K.; Melton, R.; Leggett, M.; Daniels, M.; Osbourn, A. Compromised disease resistance in saponin-deficient plants. PNAS 1999, 96, 12923–12928.

Park, J. Isolation and characterization of N-feruloyltyramine as the p-selectin expression suppressor from garlic (Allium sativum). J. Agric. Food Chem. 2009, 57, 8868– 8872.

Pedras, M.; Hossain, S.; Snitynsky, R. Detoxification of cruciferous phytoalexins in Botrytis cinerea: Spontaneous dimerization of a camalexin metabolite. Phytochemistry 2011, 72, 199–206.

231

Peng, J.; Yao, X.; Kobayashi, H.; Ma, C. Macrostemonosides, novel furostanol glycosides from Allium macrostemon. Planta Med. 1995, 61, 58–61.

Pettit, G.; Zhang, Q.; Pinilla, V.; Hoffmann, H.; Knight, J.; Doubek, D.; Chapuis, J.; Pettit, R.; Schmidt, J. Antineoplastic agents. 534. Isolation and structure of sansevistatins 1 and 2 from the African Sansevieria ehrenbergii. J. Nat. Prod. 2005, 68, 729–733.

Pezet, R.; Pont, V.; Hoang-Van, K. Enzymatic detoxification of stilbenes by Botrytis cinerea and inhibition by grape berries proanthocyanidins. In Recent Advances in Botryris Research (Verhoeff, K., Malathrakis, N.E. and Williamson. B.. Eds.), 1992, pp. 87–92. Pudoc Scientific Publishers, Wageningen.

Preston, I. 1933. Hybridization of Lilies. Royal Horticultural Society Lily Yearbook 2:173–193.

Rajput, Z.; Hu, S.; Xiao, C.; Arijo, A. Adjuvant effects of saponins on animal immune responses. J. Zhejiang Univ Sci B. 2007, 8, 153–161.

Quidde, T.; Osbourn, A.; Tudzynski, P. Detoxification of -tomatine by Botrytis cinerea. Physiol. Mol. Plant Pathol. 1998, 52, 151–165.

Raabe, R.; Conners, I.; Martinez, A. Host Genera and their Pathogens. In Checklist of plant diseases in Hawaii. Hawaii Inst. Trop. Agric. Human Resources, College of Tropical Agriculture and Human Resources, University of Hawaii: Honolulu, HI, 1981; No. 22, 120.

Rajput, Z.; Hu, S.; Xiao, C.; Arijo, A. Adjuvant effects of saponins on animal immune responses. J. Zhejiang Univ. Sci. B. 2007, 8, 153–161.

Rayburn, J.; Bantle, J.; Friedman, M. Role of carbohydrate side chains of potato glycoalkaloids in developmental toxicity. J Agric. Food Chem. 1994, 42, 1511– 1515.

Roddick, J.; Weissenberg, M.; Leonard, A. Membrane disruption and enzyme inhibition by naturally-occurring and modified chacotriose-containing Solanum steroidal glycoalkaloids. Phytochemistry 2001, 56, 603–610.

Roddick, J. The acetylcholinesterase inhibitory activity of steroidal glycoalkaloids and their aglycones. Phytochemistry 1989, 28, 26314.

Rumbeiha, W; Francis, J.; Fitzgerald, S.; Nair, M.; Holan, K.; Bugyei, K.; Simmons, H. A comprehensive study of Easter lily poisoning in cats. J. Vet. Diagn. Invest. 2004, 16, 527–541.

232

Sánchez-Mata, M.; Yokoyama, W.; Hong, Y.; Prohens, J. -Solasonine and - solamargine contents of Gboma (Solanum macrocarpon L.) and Scarlet (Solanum aethiopicum L.) eggplants. J. Agric. Food Chem. 2010, 58, 5502–5508.

Sang, S.; Mao, S.; Lao, A.; Chen, Z.; Ho, C. New steroid saponins from the seeds of Allium tuberosum L. Food Chem. 2003, 83, 499–506

Sashida, Y.; Ori, K.; Mimaki, Y. Studies on the chemical constituents of the bulbs of Lilium mackliniae. Chem. Pharm. Bull. 1991, 39, 2362–2368.

Sati, O.; Pant, G. Spirostanol glycosides from Asparagus plumosus. Phytochemistry 1985, 24, 123–126.

Satou, T.; Mimaki, Y.; Kuroda, M.; Sashida, Y.; Hatakeyama, Y. A pyrroline glucoside ester and steroidal saponins from Lilium martagon. Phytochemistry 1996, 41, 1225–1230.

Sautour, M.; Miyamoto, T.; Lacaille-Dubois, M. Steroidal saponins from Asparagus acutifolius. Phytochemistry 2007, 68, 2554–2562.

Sautour, M.; Miyamoto, T.; Lacaille-Dubois, M. Steroidal saponins from Smilax medica and their antifungal activity. J. Nat. Prod. 2005, 68, 1489–1493.

Sautour, M.; Miyamoto, T.; Lacaille-Dubois, M. Steroidal saponins and flavan-3-ol glycosides from Dioscorea villosa. Biochem. Syst. Ecol. 2006, 34, 60–63.

Sautour, M.; Mitaine-Offer, A.; Lacaille-Dubois, M. The Dioscorea genus: a review of bioactive steroid saponins. J. Nat. Med. 2007, 61, 91–101.

Sautour, M.; Mitaine-Offer, A.; Miyamoto, T.; Dongmo, A.; Lacaille-Dubois, M. Antifungal steroid saponins from Dioscorea cayenensis. Planta Med. 2004, 70, 90–92.

Sashida, Y.; Ori, K.; Mimaki, Y. Studies on the chemical constituents of the bulbs of Lilium mackliniae. Chem. Pharm. Bull. 1991, 39, 2362–2368.

Sato, Y.; Hayakawa, S. The microbiological transformation of diosgenin. J. Org. Chem. 1963a, 28, 2742–2743.

Sato, Y.; Hayakawa, S. The microbiological hydroxylation of solasodine. J. Org. Chem. 1963b, 28, 2739–2742.

Satou, T.; Mimaki, Y.; Kuroda, M.; Sashida, Y.; Hatakeyama, Y. A pyrroline glucoside ester and steroidal saponins from Lilium martagon. Phytochemistry 1996, 41, 1225–1230.

233

Sautour, M.; Miyamoto, T.; Lacaille-Dubois, M. Steroidal saponins from Smilax medica and their antifungal activity. J. Nat. Prod. 2005, 68, 1489–1493.

Schlösser, E. Role of saponins in antifungal resistance IV. Tomatin-dependent development of species of Alternaria on tomato fruits. Acta Phytopathologica 1975, 10, 77–87.

Shao, B.; Guo, H.; Cui, Y.; Ye, M.; Han, J.; Guo, D. Steroidal saponins from Smilax china and their anti-inflammatory activities. Phytochemistry 2007, 68, 623–630.

Sharma, S.; Thakur, N. Oligofurostanosides and oligospirostanosides from roots of Asparagus filicinus. Phytochemistry 1996, 41, 599–603.

Sharma, S.; Chand, R.; Sati, O.; Steroidal saponins of Asparagus adscendens. Phytochemistry 1982, 21, 2075–2078.

Shen, P.; Wang, S.; Liu, X.; Yang, C.; Cai, B.; Yao, X. A new steroidal saponin from Dioscorea deltoidea Wall var. orbiculata. Chin. Chem. Lett. 2002, 13, 851–854

Shimomura, H.; Sashida, Y.; Mimaki, Y. Bitter phenylpropanoid glycosides from Lilium speciosum var. rubrum. Phytochemistry 1986, 25, 2897–2899.

Shimomura, H.; Sashida, Y.; Mimaki, Y.; Minegishi, Y. Jatropham glucoside from the bulbs of Lilium Hansonii. Phytochemistry 1987. 26, 582–583.

Shimomura, H.; Sashida, Y.; Mimaki, Y. Steroidal saponins, pardarinoside A–G from the bulbs of Lilium pardarinum. Phytochemistry 1989, 28, 3163–3170.

Shimoyamada, M.; Suzuki, M.; Sonta, H.; Maruyama, M.; Okubo, K. Antifungal activity of the saponin fraction obtained from Asparagus officinalis L. and its active principle. Agric. Biol. Chem. 1990, 54, 2553–2557.

Shimoyamada, M.; Suzuki, M.; Maruyama, M.; Watanabe, K. An antifungal saponin from white asparagus (Asparagus officinalis L.) bottoms. J. Sci. Food Agric. 1996, 72, 430–434.

Shoyama, Y.; Hatano, K.; Nishioka, I.; Yamagishi, T. Phenolic glycosides from Lilium longiflorum. Phytochemistry 1987, 26, 2965–2968.

Sinden, S.; Webb, R. Effect of variety and location on the glycoalkaloid content of potatoes. Am. Potato J. 1972, 49, 334–338.

Sinden, S.; Sanford, L.; Cantelo, W.; Deahl, K. Bioassays of segregating plants. A strategy for studying chemical defenses. J. Chem. Eco. 2005, 14, 1573–1561.

234

Singh, B.; Bhat, T.; Singh, B. Potential therapeutic applications of some antinutritional plant secondary metabolites. J. Agric. Food Chem. 2003, 51, 5579–5597.

Skoog, D.; Holler, F.; Nieman, T. Introduction. In Principles of instrumental analysis, 5th edition; Brooks/Cole Publishing: Belmont, CA, 1997; 15–18.

Smith, D.; Roddick, J.; Jones, J. Synergism between the potato glycoalkaloids chaconine and solanine in inhibition of snail feeding. Phytochemistry 2001, 57, 229–234.

Staples, R.; Mayer, A. Putative virulence factors of Botrytis cinerea acting as a wound pathogen. Fems. Microbiol. Lett. 1995, 134, 1–7.

Steel, C.; Drysdale, R. Electrolyte leakage from plant and fungal tissues and disruption of liposome membranes by α-tomatine. Phytochemistry 1988, 27, 1025–1030

Stewart, R. and Bamford, R. Morphology of somatic chromosomes in Lilium. Am. J. Bot. 1947, 34, 9–26.

Stohs, S.; Obrist, J. Sapogenins of Yucca glauca seed pods. Phytochemistry 1975, 14, 239–24.

Straathof, T. and Löffler, H. Resistance to Fusarium oxysporum at different developmental stages of Asiatic hybrid lilies. J. Amer. Soc. Hort. Sci. 1994, 119, 1068–1072.

Sun, W.; Tu, G.; Zhang, Y. A new steroidal saponin from Dioscorea zingiberensis Wright. Nat. Prod. Res. 2003, 17, 287–292.

Sviridov, A.; Paseshnichenko, V.; Kadentsev, V.; Chizhov, O. The structure of saponins from rhizomes of Diioscorea deltoidea wall. Russ. Chem. Bull. 1975, 24, 80–84.

Tai, C.; Uemoto, C.; Shoyama, Y.; Nishioka, I. Biologically active phenolics from Lilium longiflorum. Phytochemistry 1981, 20, 2565–2568.

Takeda, K.; Harborne, J.; Self, R. Identification of malonated anthocyanins in the Liliaceae and Labiatae Phytochemistry 1986, 25, 2191–2192.

Teponno, R.; Tapondjou, A.; Djoukeng, J.; Abou-Mansour, E.; Tabacci, R.; Tane, P.; Lontsi, D.; Park, H. Isolation and NMR assignment of a pennogenin glycoside from Dioscorea bulbifera L. var sativa. Nat. Prod. Sci. 2006, 12, 62–66

Tewtrakul, S.; Itharat, A. Nitric oxide inhibitory substances from the rhizomes of Dioscorea membranacea. J. Ethnopharmacol. 2007, 109, 412–416.

235

Tewtrakul, S.; Itharat, A. Anti-allergic substances from the rhizomes of Dioscorea membranacea. Bioorg. Med. Chem. 2006, 14, 8707–8711

Trojanowska, M.; Osbourn, A.; Daniels, M.; Threlfall, D. Biosynthesis of avenacins and phytosterols in roots of Avena sativa cv. Image. Phytochemistry 2000, 54, 153– 164.

Tsukida, K. and Ikeuchi, K. Epoxycarotenoids. VIII. Pollen carotenoids of Lilium longiflorum and of its cultivated hybrid. Bitamin. 1965, 32, 222–226.

Van Etten, H.; Matthews, D.; Matthews, P. Phytoalexin detoxification: importance for pathogenicity and practical implications. Annu. Rev. Phytopathol. 1989, 27, l43– 164.

VanEtten, H.; Mansfield, J.; Bailey, J.; Farmer, E. Two classes of plant antibiotics: Phytoalexins versus ―phytoanticipins.‖ Plant Cell 1994, 6, 1191–1192.

Van Baarlen, P.; Staats, M.; Van Kan, J. Induction of programmed cell death in lily by the fungal pathogen Botrytis elliptica. Mol. Plant. Path. 2004, 5, 559–574.

Verhoeff, K.; Liem J. Toxicity of tomatine to Botrytis cinerea, in relation to latency. Phytopathologische Zeitschrift 1975, 82, 333–338.

Wang, Y.; Ohtani, K.; Kasai, R.; Yamasaki, K. Steroidal saponins from fruits of Tribulus terrestris. Phytochemistry 1997, 45, 811–817.

Wang, M.; Tadmor, Y.; Wu, Q.; Chin, C.; Garrison, S.; Simon, J. Quantification of protodioscin and rutin in asparagus shoots by LC/MS and HPLC methods. J. Agric. Food Chem. 2003, 51, 6132–6136.

Watanabe, Y.; Sanada, S.; Ida, Y.; Shoji, J. Comparative studies on the constituents of ophiopogonis tuber and its congeners. II. Studies on the constituents of the subterranean part of Ophiopogon planiscapus NAKAI. Chem. Pharm. Bull. 1983, 31, 486–3495.

Wehlburg, C.; Alfieri, S.; Langdon, K.; Kimbrough, J. Liliacae. In Index of Plant Diseases in Florida. Bull. 11.; Florida Dept. of Agriculture and Consumer Services, Division of Plant Industry: Gainesville, FL, 1975; 179.

Wij, M.; Rajaraman, K.; Rangaswami, S. New glycosides of diosgenin and prazergenin- A from the rhizomes of Dioscoria prazeri. Ind. J. Chem. 1977, 15B, 451–454.

Williamson, B.; Tudzynski, B.; Tudzynski, P.; Van Kan, J. Botrytis cinerea: the cause of grey mould disease. Mol. Plant. Path. 2007, 8, 561–580.

Wilson, E. The Lilies of Eastern Asia. Dulau and Co., London. 1925.

236

Wu, Z.; Raven P. Liliaceae. In Flora of China Vol. 24 (Flagellariaceae through Marantaceae). Science Press, Beijing, and Missouri Botanical Garden Press, St. Louis, eds. 2000; 135.

Wu, J.; Morizane, C.; Iida, A.; Ueda, S.; Zhou, Z.; Xu, M.; Zhang, M.; Li, R.; Fujita, T. Structures of Three New Diterpenoids, Fritillebic Acid and Fritillebins A and B, from Bulbs of Fritillaria ebeiensis G. D. Yu et G. Q. JI. Chem. Pharm. Bull. 1995, 43, 1448–1453.

Yamanaka, T.; Vincken, J.; De Waard, P.; Sanders, M.; Takada, N.; Gruppen, H. Isolation, characterization, and surfactant properties of the major triterpenoid glycosides from unripe tomato fruits. J. Agric. Food Chem. 2008, 56, 11432– 11440.

Yang, C.; Huang, S.; Yang, X.; Jia, Z.; Nor-lignans and steroidal saponins from Asparagus gobicus. Planta Med. 2004, 70, 446–451.

Yang, C.; Zhang, Y.; Jacob, M.; Khan, S.; Zhang, Y.; Li, X. Antifungal activity of C-27 steroidal saponins. Antimicrob. Agents Chemother. 2006, 50, 1710–1714.

Yang, D.; Lu, T.; Hwang, L. Isolation and identification of steroidal saponins in Taiwanese Yam cultivar (Dioscorea pseudojaponica Yamamoto). J. Agric. Food Chem. 2003, 51, 6438–6444.

Yang, S.; Ma, Y.; Liu, X. Steroidal constituents from Dioscorea parviflora. Yao Xue Xue Bao. 2005, 40, 145–149.

Yin, J.; Kouda, K.; Tezuka, Y.; Tran, Q.; Miyahara, T.; Chen, Y.; Kadota, S. Steroidal glycosides from the rhizomes of Dioscorea spongiosa. J. Nat. Prod. 2003, 66, 646–650.

Ying, S. Liliaceae. In Flora of Taiwan Vol. 5 second edition. Taipei 2000; 52.

Yoshiaki, W.; Shuichi, S.; Yoshiteru, I.; Junzo, S. Comparative studies on the constituents of ophiopogonis tuber and its congeners. II. Studies on the constituents of the subterranean part of Ophiopogon planiscapus Nakai. Chem. Pharm. Bull. 1983, 31, 486–3495.

Zhang, J.; Meng, Z.; Zhang, M.; Ma, D.; Xu, S.; Kodama, H. Effect of six steroidal saponins isolated from Anemarrhenae rhizoma on platelet aggregation and hemolysis in human blood. Clin. Chim. Acta 1999, 289, 79–88.

Zhang, Y.; Zhang, Y.; Jacob, M.; Li, X.; Yang, C. Steroidal saponins from the stem of Yucca elephantipes. Phytochemistry 2008, 69, 264–270

237

Zhang, H.; Sydara, K.; Tan, G.; Ma, C.; Southavong, B.; Soejarto, D.; Pezzuto, J.; Fong, H. Bioactive constituents from Asparagus cochinchinensis. J. Nat. Prod. 2004, 67, 194–200.

Zhou, C.; Liu, J.; Ye, W.; Liu, C.; Tan, R. Neoverataline A and B, two antifungal alkaloids with a novel carbon skeleton from Veratrum taliense. Tetrahedron Lett. 2003, 59, 5743–5747.

Zhou, C.; Tanaka, J.; Cheng, C.; Higa, T.; Tan, R. Steroidal alkaloids and stilbenoids from Veratrum taliense. Planta Med. 1999, 65, 480–482.

Zhou, L.; Chen, T.; Bastow, K.; Shibano, M.; Lee, K.; Chen, D. Filiasparosides A–D, cytotoxic steroidal saponins from the roots of Asparagus filicinus. J. Nat. Prod. 2007, 70, 1263–1267.

Zou, K.; Wang, J.; Du, M.; Li, Q.; Tu, M. Pair of diastereoisomeric steroidal saponins from cytotoxic extracts of Tupistra chinensis rhizomes. Chem. Pharm. Bull. 2006, 54, 1440–1442.

Zullo, M.; Teixeira, J.; Spoladore, D.; Lourencao, A.; Aranha, C. Content of glycoalkaloids in Solanum americanum infested by Aphis fabae solanella. Bragantia 1984 ,43, 255–259.

238

CURRICULUM VITA

John Peter Munafo, Jr

Education

2011 Ph. D. Plant Biology, Department of Plant Biology and Pathology, Rutgers - The State University of New Jersey 1998 B. S. Biology, The Richard Stockton College of New Jersey

Professional Experience

2007–Present Flavor Research Scientist, Mars Chocolate North America LLC 2006–2007 Flavor Chemistry Technician, Masterfoods USA 2005–2006 Natural Products Chemistry Technician, Masterfoods USA (Contractor) 1999–2005 Manager Member, Medicinal Natural Products LLC

Publications

Munafo, J.; Gianfagna, T. Quantitative Analysis of Steroidal Glycosides in Different Organs of Easter Lily (Lilium longiflorum Thunb.) by LC-MS/MS. J. Agric. Food Chem. 2010, 59, 995–1004.

Munafo, J; Gianfagna, T. Antifungal Activity and Fungal Metabolism of Steroidal Glycosides of Easter Lily (Lilium longiflorum) by the Plant Pathogenic Fungus, Botrytis cinerea. J. Agric. Food Chem. - In Press