Isolation, Identification, and Biological Evaluation of Potential Flavor Modulatory

Flavonoids from Eriodictyon californicum

Dissertation

Presented in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy in the Graduate School of The Ohio State University

By

Joshua Nehemiah Fletcher

Graduate Program in Pharmacy

The Ohio State University

2011

Dissertation Committee:

A. Douglas Kinghorn, Advisor

Esperanza J. Carcache de Blanco

Steven. J. Schwartz

Werner Tjarks

Copyright by

Joshua Nehemiah Fletcher

2011

Abstract

The avoidance of bitter-tasting substances and a preference for sweet-tasting substances have deep roots in the evolutionary history of mankind. With the advent of agriculture, and more recent advancements in food production the necessity of these preferences has diminished somewhat. In fact, the evolutionary biases of humanity can be considered responsible for certain heath problems in the modern world. It is against this backdrop that the present study of Eriodictyon californicum Decne.

(Hydrophyllaceae) for bitterness-masking and sweetness-enhancing activity, in collaboration with Givaudan Flavors Corporation, Cincinnati, OH, was commenced.

In this dissertation study, ten flavonoids were isolated from E. californicum, with two of these having been publisheed previously presented without defined stereochemistry and another with incomplete spectroscopic and physical data to confirm its structure. Additionally, this is the first time that the natural absolute configuration of flavanones from E. californicum has been resolved. Four of the isolated compounds and an additional purchased compound were evaluated for sweetness-enhancing activity in a chimeric cell-based assay. All ten compounds isolated were tested for bitterness-masking activity in a cell-based assay employing the bitterness receptor hTAS2R31. The use of cell-based assays for taste research is a very recent development in the field, with access

ii to these techniques generally exclusive to the industrial sector due to the patenting of the taste receptors.

It was determined that hesperetin (44) and sakuranetin (72) from E. californicum and the purchased compound, 3,2'-dihydroxy-4,4',6'-trimethoxychalcone (105), are potential sweetness-enhancing compounds, due to their in vitro activity in the sweetness- enhancing assay employed. The results of the in vitro bitterness-masking assay showed the flavone, jaceosidin (66), and the flavanones sakuranetin (72) and 6- methoxysakuranetin (98) to be potential bitterness-masking components of the leaves of

E. californicum. When taken together, these results strongly indicate that flavonoids with different A-ring functionalities may have a role in potential in taste modulation applications.

iii

Dedication

This dissertation is dedicated to my family, and all those who aided me along the way.

iv

Acknowledgments

I would like to extend a heartfelt thanks to my advisor Dr. A. Douglas Kinghorn, for giving me the opportunity to conduct research in his laboratory. Without his support and influence this work would not have been possible.

I would like to thank my dissertation defense committee Drs. Esperanza J.

Carcache de Blanco, Steven. J. Schwartz, and Werner Tjarks for sacrificing their valuable time in order to review this dissertation.

I wish to thank the members of the Kinghorn laboratory that I have had the honor of studying with during my tenure as student in the College of Pharmacy. I wish to extend special thanks to Drs. Marcy J. Balunas, William P. Jones, and Young-Won Chin, who all served as mentors in my growth as a researcher. I would also like to especially thank Mr. Mark Bahar for being a constant friend.

I would like to thank our collaborators at the Givaudan Flavors Corporation,

Cincinnati, OH for funding this dissertation project and for conducting the sweetness- enhancing and bitterness-masking cell-based assays used in this work. I am especially grateful to Drs. Zhonghua Jia and Jay P. Slack for their mentorship and efforts in support of this project.

v

I am grateful to Dr. Karl A. Werbovetz for initially suggesting that I investigate

Eriodictyon californicum, and Dr. John M. Cassady for donating the plant material used in this research. Dr. Richard Spjut, World Botanical Associates, Laurel, MD, is acknowledged for collecting this plant material.

I would like to thank Mr. J. Fowble for his service in maintaining many of the analytical instruments used at the College of Pharmacy, and the Central Campus

Instrument Center for the use of its mass spectrometry facilities.

I am extremely grateful for the financial support that I was given as a graduate student through the Chemistry and Biology Interface Training Program (NIH) (2003-

2005), the Raymond W. Doskotch Fellowship (2009-2010), the Jack L. Beal Award

(2010), and as a Teaching Assistant (2005-2008, 2009-2010) and a Research Assistant

(2008-2009, 2011). I am especially thankful to the Givaudan Flavors Corporation for funding me as a Research Assistant (2011).

I would like to sincerely thank all who contributed to my education; from my grade school teachers at Steubenville City Schools, to each professor that I have learned from at The Ohio State University.

I would like to thank my friends and family for their loving support throughout my life. I especially thank my mother, Linda (a key collaborator on many successful science fair projects), my father, Herbert, my siblings, Christopher and Nicholas, and my wife, Gwyndolyn.

Lastly, I would like to thank the Creator, for the wonderful world in which I have the pleasure of conducting research.

vi

Vita

1999...... Steubenville High School 2003...... B.S. Arts & Sciences (Chemistry), The Ohio State University 2003-2011...... Graduate Student, College of Pharmacy, The Ohio State University

Awards and Honors

Jack L. Beal. Award, The Ohio State University, College of Pharmacy, 2010.

Raymond W. Doskotch Graduate Student Fellowship, The Ohio State University, College of Pharmacy, 2009-2010.

Chemistry-Biology Interface Training Program Fellowship, National Institutes of Health,

2003-2005.

Meyers Math and Physical Sciences Scholarship, The Ohio State University, College of

Arts and Physical Sciences 2001-2003.

Trustees Scholarship, The Ohio State University, 1999-2003.

Office of Minority Affairs Excellence Scholarship, The Ohio State University, 1999-

2003.

Ella Stafford Scholarship, Steubenville High School, 1999.

vii

Poster Presentations

Fletcher J. N., Kinghorn, A. D., Slack, J. P., Odley, A., Jia Z. Evaluation of Potential Bitterness-Masking Flavonoids from Eriodictyon californicum. Presented at the 51st Annual American Society of Pharmacognosy Meeting, St. Petersburg, FL, July 2010. Fletcher J. N., Kinghorn A. D., Slack J. P., Jia Z. Evaluation of Flavonoids from Eriodictyon californicum as Potential Taste Modifiers in a Cell-Based Assay. Presented at the 50th Annual American Society of Pharmacognosy Meeting, Honolulu, HI, June 2009. Fletcher J. N.; Chin, Y.-W.; Bahar M.; Werbovetz, K. A.; Kinghorn, A. D. Antileishmanial Activity of Some Naturally Occurring Phenolics. Presented at the 40th Annual Central Regional Meeting of the American Chemical Society, Columbus, OH, June 2008. Fletcher, J. N.; Doskotch, R. W.; Werbovetz, K. A.; Kinghorn, A. D. Anti-leishmanial Bioactivity-Guided Fractionation of Ficus elastica. Presented at the 48th Annual American Society of Pharmacognosy Meeting, Portland, ME, July 2007. Fletcher J. N.; Bahar M.; Werbovetz K. A.; Kinghorn, A. D. Antileishmanial Activity of Some Naturally Occurring Flavonoids. Presented at the 39th Annual Mid- Atlantic Graduate Student Symposium in Medicinal Chemistry, Columbus, OH, June 2006.

viii

Publications

Bahar M.; Deng Y.; Fletcher, J. N.; Kinghorn A. D. Plant-Derived Natural Products in

Drug Discovery and Development: an Overview. In Selected Topics in the

Chemistry of Natural Products, Ikan, R., Ed.; World Scientific Publications:

Singapore, 2008, pp 11-48.

Kinghorn A. D.; Pan L.; Fletcher J. N.; Chai, H. The relevance of higher plants in lead

compound discovery programs. J. Nat. Prod. 2011, 74, 1539–1555.

Fletcher, J. N.; Kinghorn, A. D.; Slack, J. P.; McCluskey, T. S.; Odley, A.; Jia Z. In vitro

evaluation of flavonoids from Eriodictyon californicum for antagonist activity

against the bitterness receptor hTAS2R31. Published on the Web Nov. 8, 2011

in J. Agric. Food Chem. DOI: 10.1021/jf204359q

Fields of Study

Major Field: Pharmacy

ix

Table of Contents

Abstract ...... ii

Dedication ...... iv

Acknowledgments ...... v

Vita ...... vii

Table of Contents ...... x

List of Tables ...... xv

List of Figures ...... xvi

List of Abbreviations ...... xxi

Chapter 1: Overview of the Field of Taste Perception and Its Relevance to Health...... 1

A. Overview of the Biological Role and Mechanism of Taste Perception...... 1

1. Evolutionary Basis for Taste ...... 1

2. Current Debates in the Field of Taste ...... 2

3. Recent Scientific Advances in the Field of Taste Perception...... 5

3.1 Elucidation of the bitter and sweet signal transduction pathways. . . . 5

3.2 Discovery of sweet taste and umami taste receptors...... 9

3.3 Discovery of bitter receptors...... 11

B. Negative Health Effects Associated with Taste Preferences ...... 12

x

1. Bitterness and "functional foods"...... 13

1.1. Examples of “functional foods” containing bitter components – soy. . .

...... 17

1.2. Examples of "functional foods" containing bitter components – citrus

juices...... 18

1.3. Examples of "functional foods" containing bitter components –

cruciferous vegetables ...... 19

1.4. Examples of "functional foods" containing bitter components – cocoa

...... 21

2. The Search for Healthy Sugar Substitutes ...... 22

2.1. High-potency sweeteners ...... 23

2.2. Commercially useful natural high-potency sweeteners...... 25

2.3. Methods to discover new high-potency sweeteners...... 30

2.4. Bulk sweeteners...... 33

3. Efforts to Modulate Taste Perception ...... 35

3.1. Sweetness enhancers ...... 35

3.2. Bitterness maskers ...... 38

4. The Use of Cell-based Approaches in the Discovery of Flavor Modulators and

Sweetener Discovery...... 42

Chapter 2: Review of Eriodictyon californicum and the Genus Eriodictyon...... 44

A. Background on Eriodictyon...... 44

1. The family Hydrophyllaceae...... 44

xi

2. Background on the genus Eriodictyon and species studied phytochemically. . . .

...... 46

2.1 Eriodictyon angustifolium Nutt...... 47

2.2 Eriodictyon sessilifolium Greene...... 49

2.3 Eriodictyon tomentosum Benth...... 51

2.4 Eriodictyon trichocalyx var. trichocalyx A. Heller...... 52

2.5 Eriodictyon californicum Decne...... 53

2.5.1 Background on Eriodictyon californicum...... 53

2.5.2 Early phytochemical investigations of Eriodictyon californicum. .54

2.5.3 Recent studies on Eriodictyon californicum...... 60

Chapter 3: Isolation, Identification, and in vitro Evaluation of Flavor-modulating

Flavonoids from Eriodictyon californicum...... 63

A. Statement of Problem...... 63

B. Experimental...... 64

1. General Experimental Procedures...... 64

2. Plant Material...... 65

3. Extraction of the leaves of Eriodictyon californicum...... 65

4. Chromatography of the chloroform-soluble extract of the leaves of Eriodictyon

californicum...... 66

5. Characterization of isolated compounds...... 69

5.1 Notes on the characterization of isolated compounds...... 69

xii

5.2 Characterization of hesperetin (44)...... 70

5.3 Characterization of homoeriodictyol (51)...... 71

5.4 Characterization of jaceosidin (66)...... 71

5.5 Characterization of naringenin (69)...... 72

5.6 Characterization of sakuranetin (72)...... 73

5.7 Characterization of 4′-isobutyrylhomoeriodictyol (95)...... 73

5.8 Characterization of 6-methoxyhomoeriodictyol (97)...... 74

5.9 Characterization of 6-methoxysakuranetin (98)...... 75

5.10 Characterization of pinocembrin (99)...... 76

5.11 Characterization of 6-methoxyhesperetin (104)...... 76

C. Discussion...... 77

1. Identification of hesperetin (44)...... 77

2. Identification of homoeriodictyol (51)...... 82

3. Identification of jaceosidin (66)...... 85

4. Identification of naringenin (69)...... 89

5. Identification of sakuranetin (72)...... 92

6. Identification of 4′-isobutyrylhomoeriodictyol (95)...... 95

7. Identification of 6-methoxyhomoeriodictyol (97) ...... 100

8. Identification of 6-methoxysakuranetin (98)...... 103

9. Identification of pinocembrin (99)...... 106

10. Structure elucidation of 6-methoxyhesperetin (104) ...... 110

xiii

11. Discussion of the potential sweetness-enhancing activity of compounds

evaluated in vitro...... 115

12. Discussion of the bitterness-masking activity of compounds evaluated in vitro.

...... 117

D. Conclusions...... 119

References ...... 123

Appendix A...... 149

1. Sweetness-enhancing bioassay ...... 149

2. Bitterness-masking bioassay ...... 151

xiv

List of Tables

Table 1.1. Purported activities of selected “functional foods.” ...... 15

Table 1.2. Internationally used natural and semi-synthetic high-potency sweeteners. . . .26

Table 1.3. Comparison of some bulk sweeteners relative to sucrose...... 35

Table 3.1 Evaluation of compounds from E. californicum in an in vitro chimeric sweetness-enhancing assay...... 116

Table 3.2. Evaluation of compounds from E. californicum in an in vitro cell-based bitterness-masking assay...... 118

xv

List of Figures

Figure 1.1. A reproduction of a taste map appearing in the recent scientific literature by

Hoon et al., 1999...... 5

Figure 1.2. Proposed signal transduction pathways for sweet and bitter tastants...... 8

Figure 1.3. Structures of bitter-tasting compounds from selected “functional foods.”. . 16

Figure 1.4. Production of the bitter compound limonin in orange juice...... 18

Figure 1.5. Conversion of glucoraphanin to sulforaphane...... 20

Figure 1.6. Structures of cyclamate and current high-potency sweeteners used in the

United States...... 25

Figure 1.7. Structures of some commercially used natural high potency sweeteners. . . 27

Figure 1.8. Diagram of the sweetness receptor and binding sites...... 31

Figure 1.9. Structure of the naturally occurring sweetness-inhibitor lactisole...... 32

Figure 1.10. Structures of selected bulk sweeteners...... 34

Figure 1.11. Structures of selected sweetness-enhancing compounds...... 37

Figure 1.12. Structures of selected of bitterness-masking compounds...... 39

Figure 2.1. Compounds isolated from Eriodictyon angustifolium ...... 48

Figure 2.2. Compounds identified from Eriodictyon sessilifolium ...... 50

Figure 2.3. Compounds isolated from Eriodictyon tomentosum ...... 51

xvi

Figure 2.4. Compounds isolated from Eriodictyon trichocalyx var. trichocalyx ...... 52

Figure 2.5. Compounds isolated from Eriodictyon californicum ...... 55

Figure 2.6. Degradation leading to the identification of flavonoids as phloroglucinol derivatives ...... 57

Figure 2.7. The interconversion between flavanones and chalcones discovered by Perkin and Hummel ...... 58

Figure 2.8. Proposed mechanism for the base-catalyzed racemization of a flavanone. . 59

Figure 3.1. Fractionation tree employed for Eriodictyon californicum leaves...... 66

Figure 3.2. Compounds isolated from Eriodictyon californicum in the present investigation...... 69

Figure 3.3. Numbering system employed for flavonoids isolated from Eriodictyon californicum...... 70

1 Figure 3.4. H NMR spectrum of hesperetin (44) in acetone-d6 ...... 79

13 Figure 3.5. C NMR spectrum of hesperetin (44) in acetone-d6 ...... 79

Figure 3.6. HMBC correlation NMR spectrum of hesperetin (44) ...... 80

Figure 3.7. HSQC correlation NMR spectrum of hesperetin (44) ...... 80

Figure 3.8. NOESY correlation NMR spectrum of hesperetin (44) ...... 81

Figure 3.9. CD spectrum of hesperetin (44) ...... 81

1 Figure 3.10. H NMR spectrum of homoeriodictyol (51) in acetone-d6 ...... 83

13 Figure 3.11. C NMR spectrum of homoeriodictyol (51) in acetone-d6 ...... 83

Figure 3.12. HSQC correlation spectrum of homoeriodictyol (51) ...... 84

Figure 3.13. HMBC correlation spectrum of homoeriodictyol (51) ...... 84

xvii

Figure 3.14. CD spectrum of homoeriodictyol (51) ...... 85

1 Figure 3.15. H NMR spectrum of jaceosidin (66) in DMSO-d6 ...... 86

13 Figure 3.16. C NMR spectrum of jaceosidin (66) in DMSO-d6 ...... 87

Figure 3.17. HMBC correlation NMR spectrum of jaceosidin (66) ...... 87

Figure 3.18. HSQC NMR correlation spectrum of jaceosidin (66) ...... 88

Figure 3.19. NOESY NMR correlation spectrum of jaceosidin (66) ...... 88

1 Figure 3.20. H NMR spectrum of naringenin (69) in acetone-d6 ...... 90

13 Figure 3.21. C NMR spectrum of naringenin (69) in acetone-d6 ...... 90

Figure 3.22. HSQC correlation NMR spectrum of naringenin (69) ...... 91

Figure 3.23. HMBC correlation NMR spectrum of naringenin (69) ...... 91

Figure 3.24. CD spectrum of naringenin (69) ...... 92

1 Figure 3.25. H NMR spectrum of sakuranetin (72) in acetone-d6 ...... 93

13 Figure 3.26. C NMR spectrum of sakuranetin (72) in CDCl3 ...... 94

Figure 3.27. HMBC spectrum of sakuranetin (72) ...... 94

Figure 3.28. HSQC spectrum of sakuranetin (72) ...... 95

Figure 3.29. CD spectrum of sakuranetin (72) ...... 95

1 Figure 3.30. H NMR spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6 . . 97

13 Figure 3.31. C NMR spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6 . .98

Figure 3.32. HSQC correlation spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6 ...... 98

Figure 3.33. NOESY correlation spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6 ...... 99

xviii

Figure 3.34. HMBC correlation spectrum of isobutyrylhomoeriodictyol (95) in acetone- d6 ...... 99

Figure 3.35. CD spectrum of 4′-isobutyrylhomoeriodictyol (95) ...... 100

1 Figure 3.36. H NMR spectrum of 6-methoxyhomoeriodictyol (97) in acetone-d6 . . . 101

13 Figure 3.37. C NMR spectrum of 6-methoxyhomoeriodictyol (97) in acetone-d6 . . . 102

Figure 3.38. HMBC correlation NMR spectrum of 6-methoxyhomoeriodictyol (97) . .102

Figure 3.39. HSQC correlation NMR spectrum of 6-methoxyhomoeriodictyol (97) . . 103

Figure 3.40. CD spectrum of 6-methoxyhomoeriodictyol (97) ...... 103

1 Figure 3.41. H NMR spectrum of 6-methoxysakuranetin (98) in acetone-d6 ...... 104

13 Figure 3.42. C NMR spectrum of 6-methoxysakuranetin (98) in acetone-d6 ...... 105

Figure 3.43. HMBC correlation NMR spectrum of 6-methoxysakuranetin (98) . . . . . 105

Figure 3.44. HSQC correlation NMR spectrum of 6-methoxysakuranetin (98) ...... 106

Figure 3.45. CD spectrum of 6-methoxysakuranetin (98) ...... 106

1 Figure 3.46. H NMR spectrum of pinocembrin (99) in acetone-d6 ...... 108

13 Figure 3.47. C NMR spectrum of pinocembrin (99) in acetone-d6 ...... 108

Figure 3.48. HMBC NMR spectrum of pinocembrin (99) ...... 109

Figure 3.49. HSQC NMR spectrum of pinocembrin (99) ...... 109

Figure 3.50. CD spectrum of pinocembrin (99) ...... 110

1 Figure 3.51. H NMR spectrum of 6-methoxyhesperetin (104) in acetone-d6 ...... 112

13 Figure 3.52. C NMR spectrum of 6-methoxyhesperetin (104) in acetone-d6 ...... 112

Figure 3.53. HSQC NMR spectrum of 6-methoxyhesperetin (104) ...... 113

Figure 3.54. HMBC NMR spectrum of 6-methoxyhesperetin (104) ...... 113

xix

Figure 3.55. 1H-1H NOESY NMR spectrum of 6-methoxyhesperetin (104) ...... 114

Figure 3.56. B-ring 1H NMR signals of 6-methoxyhesperetin (104) and hesperetin (44). .

...... 114

Figure 3.57. Selected correlations from HMBC and NOESY for 6 methoxyhesperetin

(104) ...... 114

Figure 3.58. CD spectrum of 6-methoxyhesperetin (104) ...... 115

Figure 3.59. Structure of a chalcone (105) evaluated in the sweetness-enhancing assay. . .

...... 115

xx

List of Abbreviations Used

AC: adenylate cyclase

Acetone-d6: deuterated acetone

25 [α] D: specific optical rotation aq: aqueous cAMP: cyclic adenosine monophosphate

CaSR: calcium sensing receptor

CDCl3: deuterated chloroform cGMP: cyclic guanosine monophosphate

CHCl3: chloroform d: doublet

DAG: diacyl glycerol dd: doublet of doublets

δC: carbon-13 chemical shift

δH: hydrogen-1 chemical shift

DMSO-d6: deuterated dimethylsulfoxide

DNA: deoxyribonucleic acid

EAF: Everything Added to Food xxi

FAO: Food and Agriculture Organization of the United Nations

FDA: Food and Drug Administration

FT-IR: Fourier-transformed infrared

G-protein: guanine nucleotide binding protein

Gal: galactopyranose

GC-MS: tandem gas chromatography and mass spectrometry

GI: gastrointestinal

Glc: glucopyranose

GlcA: glucuronic acid

GMP: guanosine monophosphate

GPCR: G-protein coupled receptor

GRAS: Generally Regarded as Safe

HEK-293: human embryonic kidney 293 cells

HMBC: heteronuclear multiple bond correlation spectroscopy

HPLC: high-performance liquid chromatography

HREISIMS: high-resolution electrospray ionization mass spectroscopy

HSQC: heteronuclear single-quantum coherence spectroscopy

Hz: hertz

IC50: sample concentration that inhibits response to 50% of the untreated control

IP3: inositol 1,4,5 triphosphate

IP3R3: inositol 1,4,5 triphosphate receptor type 3

IR: infrared light or infrared spectroscopy

xxii

J: coupling constant m/z: mass to charge ratio

Me: methyl mp: melting point

MeOH: methanol

MGGR: monoglucuronide of glycyrrhizin min: minutes mol: moles

MSG: monosodium glutamate

NMR: nuclear magnetic resonance

NOE: nuclear Overhauser effect

NOESY: nuclear Overhauser effect spectroscopy

ν(cm-1): infrared absorption frequency in reciprocal centimeters

PDE: phosphodiesterase

PKA: protein kinase A

PLCβ2: phospholipase C-β2

Rf: retention factor rha: rhamnopyranose s: singlet

TAS1R: taste receptor type 1

TAS2R: taste receptor type 2

TLC: thin-layer chromatography

xxiii tR: retention time

TRPM5: transient receptor potential cation channel subfamily M member 5

UV: ultraviolet light

WHO: World Health Organization

xxiv

Chapter 1: OVERVIEW ON THE FIELD OF TASTE PERCEPTION AND ITS

RELEVANCE TO HEALTH

A. Overview of the Biological Role and Mechanism of Taste Perception

1. Evolutionary basis for taste

While taste in humans today is generally reflected by hedonistic experiences, the

biological foundation of this sense is built upon basic survival needs. Taste may be

regarded as the gateway sense allowing for a final direct probe of the chemical

constituents of a given food prior to consumption (Lindemann, 2001). There are five

recognized basic taste types, namely, bitter, salty, sour, sweet, and savory (umami). Each

of these tastes plays a primary biological function, which is readily evident upon analysis

of their role. A bitter taste is generally considered to be unpleasant in the majority of

foods, and this has a sound rationale, as there are very many toxic compounds in Nature

that elicit a bitter taste response (Lindemann, 2001). It has been hypothesized that

bitterness may also serve as a marker for medicinal efficacy, as evidenced by

chimpanzees seeking bitter herbs when ill, as well as an increased tolerance to bitter

plants in both sick humans and chimpanzees (Hart, 2005). The salty taste is crucial in the

detection of the vital mineral sodium. Salty foods are especially sought out when sodium

1 levels are low, and sensitivity to salty substances increases with the demand for salt in the diet. Highly sour foods are generally perceived to be unpleasant, which helps to eliminate unripe fruits and fermented products from the diet. However, the sour taste is accepted at lower levels, which allows for necessary sour compounds in the diet such as

L-ascorbic acid (vitamin C) to be ingested. Sweet taste is generally seen as pleasant and is used to encourage the selection of foods rich in calories. Sweetness perception is also used in regulating consumption, with sensitivity to sweet stimuli decreased when in a fed- state, thus discouraging further eating. The umami taste detects L-amino acids (L- glutamate most strongly) and is generally considered to be a reward stimulus encouraging the consumption of protein-rich foods (Lindemann, 2001).

2. Current Debates in the Field of Taste

Evidence for additional taste sensations, such as metallic or fatty tastes, outside the accepted five basic tastes is debatable. Metal salts, although a necessary part of the diet, are toxic at higher levels, and thus there is an evolutionary onus for the detection of metals (Schiffman, 2000). The metallic taste has been found to be both dependent and independent on the sense of smell, depending on the stimulus, thereby adding further debate to the validity of a metallic taste as a fundamental taste (Lawless et al., 2005). (A gustatory referral by retro-nasal smell is considered a result of the sense of smell instead of taste). The proposed fat taste is speculative, as the fat content of foods is also detected by senses other than taste. Adding further confusion to the evidence for a “fat taste” is the possibility of fat-sensitive and -insensitive individuals. It does appear that fatty foods

2 are less appealing with increased amounts of fat in the GI tract. This confirms at the very least the ability of the tongue to detect fat, and shows the potential usefulness of a fat taste in managing caloric intake (Mattes, 2005). These two additional fundamental tastes most likely will not be accepted for some time if they follow previous attitudes toward umami, which although first reported in the scientific literature over 100 years ago is still considered a “new” taste (Kurihara, 2009). Indeed, the umami taste was described as

“toothyness” and “meaty” in the classic tome The Physiology of Taste by the French gastronome Jean Antheleme Brillat-Savarin, which was published in 1825 (Iordachescu,

2008). Regardless of the validity of metallic and fatty tastes as being fundamental, it does appear that taste is more complex than the “only four (or possibly five) basic taste qualities (sweet, sour, salty, and bitter, and possibly 'umami')” sentiment that is frequently conveyed in the scientific literature (Schiffman, 2000).

Perhaps the most challenged theory in the field of taste, and rightly so, is the proposal for a taste map of the tongue. The first such tongue map appeared in the literature in 1901, in the Ph.D. thesis of D. P. Hänig, whose work detailed threshold levels for taste on different areas of the tongue, in order to ascribe distinct physiological mechanisms to the basic tastes (Bartoshuk, 1993; Lindemann, 1999). Also included in this dissertation was a figure with maximum sensitivity (lowest threshold) and minimum sensitivity (highest threshold) of sour, sweet, and bitter tastants plotted against regions of the tongue; however, these values were normalized against each other losing the concentration data in the process (Bartoshuk, 1993). This unit-less figure is likely the source of the modern tongue map, since the minima could be interpreted as null taste

3

(Bartoshuk, 1993). The original data were further obscured in the volume Sensation and

Perception in the History of Experimental Psychology by Edwin Boring, where a graphical representation using Hänig’s data was plotted as the lowest threshold over threshold vs. tongue location (Bartoshuk, 1993). This plot was published with the vague ordinate “sensitivity” and does not illustrate clearly that even the greatest difference in threshold levels for any taste is by a five times concentration factor, which, while seemingly a significant difference, corresponds to less than an order of magnitude change in the limit of detection (Bartoshuk, 1993).

Although the dismissal of the tongue map has been widespread among those with practical experience in the field of taste, this “scientific myth” remains often published

(Bartoshuk, 1993; Lindemann, 1999). A more scientifically accurate taste map, as in the thesis of D. P. Hänig, has been presented as a density plot in which the rim of the tongue is shown to be sensitive to all of the basic tastes, although with areas of increased sensitivity (Lindemann, 1999). This taste map, however, excludes umami, which was not accepted as a basic taste at the time the figure was prepared. The promulgation of a taste map with distinct regions of sensitivity can still be found in even the most technically advanced work in the field of taste, such as the discovery of the sweet and umami receptors (Hoon et al., 1999). Ironically, the taste map used in this report did not include a projected region on the tongue for the perception of the umami taste. An example of a taste map is presented in Figure 1.1.

4

Bitter Salty Sour Sweet

Figure 1.1. A reproduction of a taste map appearing in the recent scientific literature by

Hoon et al., 1999.

3. Recent Scientific Advances in the Field of Taste Perception

The field of taste research may seem to be somewhat primitive and having been rather slow to advance as a result of the combined scientific debates concerning fundamental principles of taste, such as what are the extent of basic tastes, and misinformation such as the taste map of the tongue. However, many modern advances have been made in the field in recent years, as scientists have begun to unravel the biological mechanisms of taste. These advances include the elucidation of the signal transduction pathway for bitter and sweet taste detection (discussed in the next section and shown in Figure 1.2), and the identification and characterization of the bitter, umami, and sweet receptors (discussed in sections 3.2 and 3.3).

3.1 Elucidation of the bitter and sweet signal transduction pathways

It had been posited for some time that the perception of the bitter and sweet tastes is mediated by guanine nucleotide binding proteins (G-proteins), as opposed to the ion

5 channels involved in the salty and sour tastes (McKinnon et al., 1992). G-proteins are commonly employed by physiological systems for chemical perception (Spielman, 1998).

In 1992, Robert Margolskee and associates at the Roche Institute of Molecular Biology,

Nutley, New Jersey identified the α subunit of a G-protein called α-gustducin as being selectively expressed in taste buds, with this protein showing striking homology to the rod and cone transducins known to play a role in sight (McKinnon et al., 1992). This discovery was made by probing DNA from taste cells with primers of conserved regions of G-proteins, in order to find those that were selectively expressed in taste cells

(McKinnon et al., 1992; Spielman, 1998). In this research, eight G-proteins were found, of which seven were known from other cell types, with two of these determined to be transducins from the retina (rod and cone transducin), and the last a novel protein named

α-gustducin (McKinnon et al., 1992; Spielman, 1998).

α-Gustducin was found to be very similar to α-transducin, with 80% sequence identity and 90% similarity on an amino acid basis between rat α-gustducin and rat rod α- transducin (McKinnon et al., 1992; Spielman, 1998). α-Transducin is known to function by photoreceptors detecting the presence of a photon, and, in turn, activating a phosphodiesterase, which produces guanosine monophosphate (GMP) from cyclic guanosine monophosphate (cGMP) (Spielman, 1998). The cell is then hyperpolarized by the closure of a cGMP-gated channel, which then allows for the release of an as yet unknown neurotransmitter (Spielman, 1998). A similar mechanism was believed to occur in bitter taste signal transduction as bitter tastants were found to lower the amount of

6 cyclic adenosine monophosphate (cAMP) and cGMP in mouse taste cells (Yan et al.,

2001; Roper, 2007).

A second messaging pathway involving inositol 1,4,5-trisphosphate (IP3) was found to be involved in bitter-taste signal transduction (Yan et al., 2001). IP3 serves as a signal to release the intracellular stores of calcium ions (Ca2+) into the cytoplasm (Roper,

2007). This influx of Ca2+ opens a monovalent cation channel, TRPM5, which results in an influx of cations and membrane depolarization, leading to the release of the neurotransmitter (Roper, 2007). It has been hypothesized that these two signal transduction mechanisms (cAMP and IP3) may be linked in a “yin-yang” mechanism of cross inhibition (Lindemann, 1996). However, this link was proposed in order to explain why deletion of gustducin stops detection of bitter compounds that are known to activate the IP3 pathway (Lindemann, 1996). This hypothesis, although potentially true, is not needed due to α-gustducin having been found to form complexes with β-and γ-gustducin in taste receptor cells, with the βγ-gustducin dimer playing a role in the IP3 pathway

(Huang et al., 1999; Yan et al., 2001). The role of β- and γ-gustducin in taste signal transduction was then verified by demonstrating that antibodies to phospholipase C-β2

(PLCβ2), a known binder of the βγ-gusducin dimer, inhibits IP3 formation (Yan et al.,

2001). This, in turn, links the binding of the βγ-gusducin dimer with PLCβ2 to the release of the neurotransmitter in taste cells (Yan et al., 2001).

The mode of sweet taste signal transduction has been found to be rather similar to that of the bitter taste, with a few key differences (Roper, 2007). Perhaps the greatest known difference at the present time is that sweet tastants have been determined to

7 increase intracellular cAMP levels, which is the opposite effect of that seen in bitter-taste transduction (Roper, 2007). Interestingly, this pathway is used for sucrose signal transduction, but not at a detectable level for saccharin or a guanidinium-containing sweetener, SC-45647 (Bernhardt et al., 1996). Saccharin, however, generates a much greater IP3 response than sucrose (Bernhardt et al., 1996). The exact role of raised levels of cAMP in sweet-taste signal transduction is still a bit unclear, with both the opening of calcium channels and the blocking of potassium channels being proposed as mechanisms by which the cell is depolarized and the neurotransmitter released (Bernhardt et al., 1996;

Roper, 2007). A compilation of these data is given in graphical form in Figure 1.2.

bitter

sweet

Figure 1.2. Proposed signal transduction pathways for sweet and bitter tastants. (Dashed lines indicate predicted steps in the pathway. From Huang et al., 1999; Yan et al., 2001, and Roper 2007). 8

3.2 Discovery of sweet taste and umami taste receptors

As has been demonstrated, the seminal discovery of α-gustducin has allowed for the elucidation of the signal transduction pathway for sweet and bitter taste perception.

However, this discovery facilitated even greater insight into the biological basis of taste, namely, the identification of the sweet, umami, and bitterness receptors, as a result of a collaboration between the laboratories of Nicholas Ryba at the National Institutes of

Health, Bethesda, Maryland and Charles Zuker at the University of California at San

Diego (Hoon et al., 1999; Adler et al., 2000; Chandrashekar et al., 2000).

The initial work leading to the discovery of the sweet receptor was performed by screening rat taste cells for G-protein coupled receptors (GPCRs) (Hoon et al., 1999).

Two novel GPCRs were identified as being selectively expressed in taste sensitive cells, namely, Tas1R1 and Tas1R2 (initially known as TR1 and TR2) (Hoon et al., 1999).

These were found to have 40% amino acid sequence homology with one another, and about 30% homology to the calcium receptor, CaSR, which shares the feature of having a large extracellular binding domain with the Tas1Rs (Hoon et al., 1999). Their similarity to the then known umami receptor mGLuR, suggested a possible role in the detection of umami (Hoon et al., 1999). However, reliance on a taste map much like that displayed in

Figure 1.1 led the authors to implicate erroneously TasR1 in the detection of sweetness and Tas1R2 in the detection of bitterness (Hoon et al., 1999). In fact, the reliance upon a taste map in this study was heavily criticized in the scientific literature shortly after the publication of these results (Lindemann, 1999).

9

It was soon observed that neither Tas1R1 nor Tas1R2 corresponds to the Sac locus in mice, which through research on sweetness-insensitive mice is known to play a role in the detection of sweeteners, particularly saccharin (Roper, 2007). This led to several more or less simultaneous reports of a new member of the Tas1R family, Tas1R3, a GPCR that occurs in the Sac locus (Kitagawa et al., 2001; Max et al., 2001;

Montmayeur et al., 2001; Nelson et al., 2001). It was noted by Nelson and associates that while Tas1R1 and Tas1R2 expression do not generally overlap, Tas1R3 overlaps with both the other Tas1Rs (Nelson et al., 2001). This observation permitted the inference to be made that Tas1R1 and Tas1R2 form heterodimers with Tas1R3 (Nelson et al., 2001).

Tas1R2::Tas1R3-expressing cells were found to be sensitive to sweet stimuli (Nelson et al., 2001). Later work on Tas1R1::Tas1R3 showed its role in the detection of umami substances (Nelson et al., 2002). The formation of homodimers is also a possibility as

Tas1R3 knockout mice have been shown to display a preference for both sweet- and umami-tasting substances, although this preference was observed only when knockout mice were given a higher concentration of tastant than necessary to elicit the same response in wild type mice for most compounds (Montmayeur et al., 2001; Damak et al.,

2003). Since the sweet and umami receptors are known to form dimers, the most likely means of producing these responses is the formation of homodimers in the absence of

Tas1R3. The existence of single receptor may seem odd when one considers the structural diversity of the known sweeteners. However, simple sugars are most probably the ligands targeted by this receptor, with the other known sweeteners being coincidental activators (Nelson et al., 2001).

10

3.3 Discovery of bitter receptors

The bitterness receptors were identified initially by dual lines of investigation

(Adler et al., 2000; Chandrashekar et al., 2000). The first method used was to select prospective bitterness receptors, which were then expressed in HEK-293 cells and subsequently incubated with bitter (and sweet) tastants in order to observe if the cells were stimulated selectively by bitter compounds (Adler et al., 2000). This method targeted G-protein coupled receptors (GPCRs) from bitter-sensitive cells determined to express α-gustducin, which had been shown to be instrumental in the perception of bitter tastants (Ming et al., 1999; Chandrashekar et al., 2000). The second method used involved analyzing the location of mutations in bitter-insensitive mice (Chandrashekar et al., 2000).

There are at least 25 functioning bitterness receptor genes (TAS2Rs) encoded in the human genome, with an additional eight identified pseudogenes (McKinnon et al.,

1992; Go et al., 2005). A combination of the switch by humans from a nearly exclusively vegetarian diet to a diet that includes meat, along with advances in food preparation techniques, may have resulted in relaxed constraints upon the conservation of bitterness receptor genes; this hypothesis is consistent with the number of hTAS2R pseudogenes that exist (Wang et al., 2004). In fact, hTAS2R38 has been identified as a gene undergoing loss of function, as a result of research on tasters and non-tasters for the thioamides (Mennella et al., 2005).

Human bitter-taste receptors (hTAS2Rs) consist of seven transmembrane domains, and generally have a low affinity for their ligands when compared to other

11

GPCRs. Some hTAS2Rs respond to a wide range of bitter compounds (up to one third of the known bitter compounds), but can also discriminate between structurally similar bitter compounds (Brockhoff et al., 2010). It does appear that the C terminus of bitter receptors plays a greater role in taste perception, as the N terminal two-thirds of hTAS2R31 and hTAS2R46 can be interchanged with the proteins maintaining the same bitterness- detecting profiles of the native proteins that share the same C terminal sequence

(Brockhoff et al., 2010). However, this does not account for all binding pockets, since the very bitter-tasting substance denatonium appears to interact with the central portion of hTA2R46, although the interaction is weak (Brockhoff et al., 2010).

B. Negative Health Effects Associated with Taste Preferences

While seemingly unrelated subjects, the fields of taste and health do share several instances of overlap. For example, the previously mentioned nutritional aspects of taste, and the special role of bitterness in implying both toxicity and potential medicinal activity are primarily health roles. The suspected dual role of bitter taste detection may confer to humankind an evolutionary predisposition in agreement with the oft quoted notion of Paracelsus that “solely the dose determines that a thing is not a poison”

(Borzelleca, 2000). Perhaps in today's world the most prominent evidence that any chemical can be toxic given a sufficient dose is exemplified by the over-consumption of calorigenic sugars (Johnson et al., 2009).

12

Mankind's preference for sweet foods coupled with a stable food supply and other modern conveniences renders the developed world as an “obesogenic environment”

(Krebs, 2009). Obesity has been associated with a great number of health risks: cancer, coronary heart disease, dyslipidemia, gallbladder disease, hypertension, liver disease, osteoarthritis, respiratory problems, sleep apnea, stroke, type-2 diabetes, and gynecological problems such as abnormal menses and infertility (Krebs, 2009; www.cdc.gov/obesity/causes/health.html). An additional health hazard resulting from the desire for sweet-tasting foodstuffs is the role of sugar in the formation of dental caries

(Sreebny, 1982).

The aversion to bitter tastants, however, can also cause negative health effects, especially in the rejection of healthy foods (Roy, 1990; Drewnowski and Gomez-

Carneros, 2000). This is especially a concern for bitter “supertasters” and the young

(Roy, 1990; Krebs, 2009). Additionally, the bitter taste of some medicines is of concern as it is a cause of poor patient compliance, especially among the young (Binello et al.,

2004; Ley, 2008).

1. Bitterness and "functional foods"

A paradoxical health issue related to bitter taste is in the area of “functional foods,” as the very same plant secondary metabolites that are considered beneficial in many vegetables are also bitter and therefore considered undesirable in many food preparations (Table 1.1) (Drewnowski and Gomez-Carneros, 2000). In fact, domesticated plants are often selected for the removal of bitter (and in many cases toxic)

13 secondary metabolites through breeding, such as the steroidal alkaloids in potatoes, quinolizidine alkaloids in lupins, and cyanogenic glycosides in almonds (Wink 1985;

Johns, 1986; Roy, 1990; Drewnowski and Gomez-Carneros, 2000; Binello et al., 2004).

There are many methods to remove bitter compounds from foodstuffs aside from traditional breeding techniques, including the use of transgenic organisms as in the removal of limonoids from oranges, and chemical modification techniques such as the enzymatic and microbial transformation of flavones in orange juice (Drewnowski and

Gomez-Carneros, 2000). The phenolic components of wine may be removed using a variety of techniques such as precipitation (as in the process of aging), adsorption onto proteins such as isinglass or casein, and the use of cyclodextrin beads (Drewnowski and

Gomez-Carneros, 2000). Bitter components of foods may also be masked during food preparation by dilution with other flavors such as fats, sugars, or salt (Drewnowski and

Gomez-Carneros, 2000). Of course, the addition of fat, sugar, or salt can render an otherwise healthy meal choice unhealthy. Although the removal of lethal bitter compounds from food is necessary, putative beneficial substances may also be selected out, thereby placing health and taste at odds with one another. When the bitter component is necessary for the function of the product as in certain medicines and

“functional foods,” removal is not an option (Drewnowski and Gomez-Carneros, 2000).

14

Functional Compound Activities Supported by References Food Classes of Bitter Human Clinical Trials Constituents Brassica Glucosinolates, Cancer Terry et al., 2001; vegetables isothiocyanates chemoprevention Johnson, 2003 (breast, colorectal) Citrus Juices Flavonoids, Cancer Norell et al., 1986; Puri limonoids, chemoprevention et al., 1996 (gastric, pancreatic) Cocoa Products Catechins, Cancer Bayard et al., 2007; diketopiperazines, chemoprevention Cooper et al., 2008 xanthines (general), cardiovascular protective effects, prevention of type 2 diabetes Soy Products Isoflavones, Anti-osteoporosis, Potter et al., 1998; Guha soyasaponins cancer et al., 2009; Yang et al., chemoprevention 2009; Yan and (breast, colorectal, Spitznagel, 2005 prostate) Table 1.1. Purported activities of selected “functional foods.”

It should be noted that bitterness is not perceived universally as a negative attribute and may even be preferred in products such as beer, bitter melon, chocolates, coffee, grapefruit, lemon, and wine (Ley, 2008). However, the bitterness of caffeine and the ability of ethanol to enhance the bitter taste of the components of alcoholic beverages, indicates that the acceptance of these products may be linked to their perceived beneficial attributes rather than their flavor profile (Drewnowski and Gomez-Carneros, 2000).

These “accepted” bitter foods and beverages are not accepted carte blanche, and, indeed, recent work on correlations between taste receptors and alcoholism have shown that the

15 non-functioning bitterness receptors hTAS2R16 and hTAS2R38 are linked to alcoholism especially among African-Americans (Duffy et al., 2004; Hinrichs et al., 2006; Wang et al., 2007). From such information it can be gleaned that the bitter taste of some alcoholic beverages may lead to a rejection of these drinks. In fact, very traditional preparation techniques such as aging, and the use of proteins such as egg whites or isinglass, and the admixture with sweeteners, have been applied in order to counteract the bitterness of alcoholic beverages (Drewnowski and Gomez-Carneros, 2000). Examples of compounds present in “functional foods” are given in Figure 1.3.

HO O HO O

H COOH OH O O COOH OH OH HO OH O 1 O 2 H HOH2C O O O HO O HOH C O HOH2C O O 2 H O HO O O 3 OH OH O HO O H O O OH O OH OH O H O O H3C OH H 4 OH OH 5 O OH HO HO S O HO O HOH2C O S S S SCN HO HO N N HO OH OSO3H HO3SO OH 6 7 OH 8 O HOH2C O S HO O OH HN N N HO OH OSO H OH O 3 N N OH OH 9 O 10 11 O NH HN OH R2 O 12 13 Figure 1.3. Structures of bitter-tasting compounds from selected “functional foods.” 16

1.1. Examples of “functional foods” containing bitter components – soy.

Genistein (1) and dadzein (2), isoflavones from soy [Glycine max Merr.

(Leguminosae)], have been shown to contribute directly to an unacceptably poor taste in soy milk (Drewnowski and Gomez-Carneros, 2000; Matsuura et al., 1989). A substantial amount of evidence has been obtained on genistein, in particular, suggesting its benefit as a potential cancer chemopreventive agent, particularly in the area of breast cancer (e.g.

Lamartiniere, 2000; Polkowski et al., 2000; Jerome-Morais et al., 2011), as well as evidence for possible use in the prevention of osteoporosis (Bitto et al., 2010). A technique proposed to remove genistein and dadzein functionally from soy milk by blocking their enzymatic hydrolytic production from the respective glucosides has been published, and such an approach would allow for the inclusion of soy isoflavones in the diet while bypassing flavor issues (Matsuura et al., 1989). This method does not, however, completely remove the bitter isoflavones 1 and 2 (Matsuura et al., 1989). Other processing techniques have also been found to remove soy isoflavones in various products, including a 76% loss of the isoflavone content in the production of tempeh, a meat substitute, a 67% loss in the production of tofu, and an 82% loss in the production of soy protein isolate (Wang and Murphy, 1996).

It has been contended that the isoflavones are not the sole contributors to soy product bitterness, since the soyasaponins, particularly soyasaponin I (3), also are found as a major cause of product bitterness (Aldin et al., 2006). These bitter saponins, while selected against for taste reasons, exhibit growth inhibitory activity for cancer cell lines

(Hsu et al., 2005; Ellington et al., 2006).

17

1.2. Examples of "functional foods" containing bitter components – citrus juices.

Citrus juices may contain two major bitter compound classes, flavonoids and limonoids (Puri et al., 1996; Drewnowski and Gomez-Carneros, 2000). Indeed, the bitterness of limonin (4) breakdown products and flavonoids has been termed “a major problem for the citrus industry” (Drewnowski and Gomez-Carneros, 2000). The flavonoid naringin (5) is considered to be the primary bitter constituent of fresh juices, while the limonoid, limonin, results in the delayed bitterness of citrus products, due to its formation from the tasteless precursor limonate-A-ring lactone over a period of hours after juicing (Figure 1.4) (Puri et al., 1996).

O O

O O OH O O + O COOH H H H O O H O O H Limonoid D-ring O O lactone hydrolase O H H

limonate-A-ring lactone limonin (4) (non-bitter) (bitter)

Figure 1.4. Production of the bitter compound limonin in orange juice.

Certain limonoids have been found to possess a variety of activities germane to cancer chemoprevention, inclusive of the induction of the phase II enzymes glutathione

S-transferase and quinone reductase, which increases the removal of xenobiotics,

18 inclusive of some carcinogens. Selected limonoids inhibit the initiation and promotion of carcinogenesis, and alter estrogen metabolism, and inhibit both estrogen-dependent and - independent cell growth in vitro (Ejaz et al., 2006; Perez et al., 2010). Additionally, compounds of this type have been found to inhibit biofilm formation of the shiga toxin- producing Escherichia coli O157:H7, while lowering the expression of the shiga toxin gene and various other genes associated with virulence (Vikram et al., 2010).

The bitter flavonoid, naringin, has been shown to be a superoxide dismutase and catalase activator, affording antioxidant properties via a mechanism that does not deplete vitamin E levels (Jeon et al., 2001). Naringin was also found to lower cholesterol levels, which in light of its antioxidant effects suggests a possible role of this compound in preventing atherosclerosis (Kim et al., 2006). Many of the techniques devised to debitter citrus juices remove these potentially beneficial compounds, once again placing taste preferences at odds with optimum health (Puri et al., 1996).

1.3. Examples of "functional foods" containing bitter components – cruciferous vegetables

Food plants in the family Brassicaceae (e.g., broccoli, Brussels sprouts, cabbage, cauliflower, kale, and mustard greens) produce bitter, sulfur-containing glucosinolates, which are broken down to isothiocyanates upon plant tissue damage, due to the release of the enzyme myrosinase (Drewnowski and Gomez-Carneros, 2000; Johnson, 2003). An exmple of this is the formation of sulforaphane (6) from glucoraphanin (7) shown in

Figure 1.5 (Fimognari et al., 2002). Levels of the glucosinolates progoitrin (8) and

19 sinigrin (9) have been found to correlate directly to reports of Brussels sprouts being perceived as “too bitter” or “having a poor taste” (Drewnowski and Gomez-Carneros,

2000). Although produced by cutting during harvesting, isothiocyanates are only formed at the site of plant injury, leaving much of the glucosinolates present in the plant unconverted to isothiocyanates (Johnson, 2003). While the human gut microflora is able to produce isothiocyanates from glucosinolates, the main factor in this conversion is the plant enzyme myrosinase, as evidenced by a 1.2-7.3% conversion rate for cooked watercress (thus possessing an inactivated myrosinase) as compared to a 17.2-77.7% conversion level for raw watercress (Johnson, 2003). The amount of isothiocyanates in producing vegetables varies greatly based on variety of factors and is indeed influenced directly by the whole supply chain from plant breeding, to harvesting, storage, and ultimately food preparation techniques (Johnson, 2003).

O HO S O HO O myrosinase S S HO SCN HO N HO3SO glucoraphanin sulforaphane

Figure 1.5. Conversion of glucoraphanin to sulforaphane.

20

1.4. Examples of "functional foods" containing bitter components – cocoa

Products from the seeds of cocoa tree [Theobroma cacao L. (Malvaceae)] may seem an odd class of healthy functional food due to the high fat and sugar levels in chocolate. However, the health benefits of cocoa consumption have been noted in various experimental and human trials (Weisburger 2001; Azam et al., 2003; Lee et al.,

2003; Bayard et al., 2007; Cooper et al., 2008). In fact, cocoa has been found to contain more antioxidants present than tea or red wine (Lee et al., 2003), although it should be mentioned that tea can be consumed with minimal caloric intake (Weisburger 2001). The primary bitter constituent of cocoa is somewhat controversial with (-)-catechin (10) and closely related flavans, theobromine and other xanthines (11), and amino acid dimers known as diketopiperazines (12) are all believed to contribute to the bitterness of cocoa

(Pickenhagen et al., 1975, Drewnowski and Gomez-Carneros, 2000; Kofink et al., 2007).

It has been reported that a bitterness-enhancement effect occurs between theobromine and compounds of the diketopiperazine class, which complicates the clear delineation of any compound class being the main bitter-tasting component of cocoa (Pickenhagen et al., 1975).

Regardless of which compound class represents the primary bitter constituents, both the catechins and the xanthines present in cocoa have been found to possess antioxidant activity, and hence may confer an effect against mutagenesis and atherosclerosis via a free-radical scavenging mechanism (Azam et al., 2003; Lee et al.,

2003). The xanthines, however, have also been found to have a prooxidant activity, albeit only in extreme and biologically irrelevant conditions (Azam et al., 2003).

21

Although cocoa possesses many compounds that may provide health benefits, it is difficult to recommend chocolate consumption to promote health due to its high saturated fat and the added sugar present (Roy, 1990; Weisburger 2001; Lee et al., 2003; Cooper et al., 2008). The use of cocoa beverages over chocolate partially avoids this issue, due to its comparatively lower fat and sugar content (Lee et al., 2003; Cooper et al., 2008). The sugar content of chocolates is largely due to the desire to mask the inherent bitterness of cocoa products, in yet another example of the conflict between health and taste preference (Roy, 1990).

2. The Search for Healthy Sugar Substitutes

Largely due to the health concerns brought about by excessive sugar intake, mankind has searched for minimally calorigenic and cariogenic sugar substitutes. In the

United States this has led to the approval as food addivites by the United States Food and

Drug Administration (FDA) of five artificial high potency sweeteners (acesulfame-K, aspartame, saccharin, sucralose, and neotame), which have a low taste threshold when compared with sucrose (Bloomgarden, 2011). Additionally many sweet compounds have been isolated from plants representing numerous chemotypes (Kinghorn et al., 2010). Of these, rebaudioside A isolated from Stevia rebaudiana (Bertoni) Bertoni (Asteraceae), was self-designated as Generally Regarded As Safe (GRAS) status by several companies

(Blue California, Cargill, McNeil Nutritionals, Sweet Green Fields, and Whole Earth

Sweetener/Merisant), which was not challenged by the FDA (Tarantino, 2008a;

Tarantino, 2008b; Tarantino, 2009; Cheeseman, 2009; Cheeseman, 2010b; Behrens et al.,

22

2011). This includes the use of rebaudioside A as a table top sweetener, and in low- calorie beverages (Tarantino, 2009). Other sweet plant natural products and semisynthetic derivatives have been approved as sweeteners or flavoring agents in various countries throughout the world (Kinghorn et al., 2010; Behrens et al., 2011). Aside from the use of high potency sweeteners, bulk sweeteners comprised mainly of sugars and sugar alcohols, and generally possessing a sweetness potency close to that of sucrose, are used alone or in conjunction with high-potency sweeteners as sugar substitutes and due to limited metabolism are often lower calorie alternatives to sugars (Levin et al., 1995,

Embuscado, 2006).

2.1. High-potency sweeteners

Of the high-potency sweeteners used in the United States, the vast majority have been discovered by serendipity. The sweetness of saccharin (14) (Figure 1.6) was reported in 1879 by Constantin Fahlberg, who, while working in the laboratory of Ira

Remsen at Johns Hopkins University studying the chemistry of coal tar, spilled a solution containing saccharin on his hands, and, after not thoroughly washing his hands, noted the uncharacteristic sweetness of bread he consumed with his evening meal (Morrison, 1979;

Walters, 1991). The sweetener sodium cyclamate (15), which was formerly used in the

United States and remains in use in much of the world, was discovered by Michael Sveda while working as a graduate student at the University of Illinois, when some of the compound was picked up on a cigarette placed on his laboratory bench (Walters, 1991).

James Schlatter discovered the sweet taste of aspartame (16) in 1965 while working on

23 gastric peptides at G.D. Searle & Company, Skokie, Illinois (Crosby and Beidleri, 1976;

Walters, 1991). A solution containing the compound bumped, contacting his hand, and later in the day he licked his fingers in order to pick up a piece of paper (Crosby and

Beidleri, 1976; Walters, 1991). Noting that the sweetness could not be due to sugar,

Schlatter then tasted the dipeptide assuming it was safe due to its structure, and verified its sweet taste (Crosby and Beidleri, 1976). Aspartame could very well have been purposefully designed as the sweetness of certain amino acids was already known, indeed even being indicated in the naming of glycine. The sweetness of acesulfame-K (17) (ace-

K), was discovered in 1967 by Karl Clauss at Hoechst AG (Hoechst Germany), who also licked his fingers after handling the substance in order to pick up a piece of weighing paper (Walters, 1991). In retrospect, the sweet tastes of acesulfame K and cyclamate are not surprising given their structural similarities to saccharin (Walters, 1991). The sweetness of sucralose (18) was discovered in 1976 at the University of London when

Shashikant Phadnis misunderstood his adviser Leslie Hough's suggestion to “test” the substance as a request to taste the substance (Crosby and Beidleri, 1976; Walters, 1991)!

Once again, it seems as though a rational approach in finding a non-metabolically activate sucrose derivative could have led to the discovery of this compound; especially in light of the well known sweetness of chlorinated compounds.

Neotame (19) was discovered in 1994 by Claude Nofre and Jean Marie Tinti of

Lyon, France, while exploring a proposed hydrophobic region in the aspartame binding pocket (Nofre and Tinti, 1994). Although based on a skeleton discovered serendipitously this is the only artificial sweetener on the United States market that was intentionally

24 discovered. Rebaudioside A (20) is a secondary metabolite isolated from the leaves of S. rebaudiana and was first obtained in pure form and determined structurally by the

Tanaka group at Hiroshima University in Japan (Kohda et al., 1976). However, this substance was not approved for use in the United States until 2008, after arduous safety screening was carried out, and its use was supported by a review by the Joint FAO/WHO

Expert Committee on Food Additives (Carakostas et al., 2008; Tarantino, 2008a).

O O OH O NH S S N O Na O O O N K H O O O H O H N N S 2 O O 14 15 16 O 17 OR2 Cl OH OH O OH HO O H O O HO H O O OH HN N H Cl COOR1 Cl O 18 19 20 R1=-glc 2 R2=-glc --glc 3 -glc

Figure 1.6. Structures of cyclamate and current high-potency sweeteners used in the United States.

2.2. Commercially useful natural high-potency sweeteners

Several sweet-tasting natural products and natural product derivative high- potency sweeteners are used throughout the world as sweeteners or flavoring agents 25

(Table 1.2, Figure 1.7). Rebaudioside A (20) is not the sole sweet constituent of the producing plant, and stevioside (21) the parent compound is itself highly sweet and is present in extracts of Stevia rebaudiana as the most abundant constituent (Kennelly,

2002). Stevioside was first crystallized in 1931 (Bridel and Lavielle, 1931). In fact, a self designation of GRAS for stevioside has been received by the FDA with “no questions” (Cheeseman, 2011a). This designation, while weaker than the FDA-approved

GRAS status, allows for the marketing of stevioside in the U.S. as a food additive

(Cheeseman, 2011a).

Natural High-Potency Examples of Countries where Examples of Countries Sweetener Utilized as a Sweetener where Utilized as a Food Additive but not a Sweetener Rebaudioside A (20) France, Japan, United States Stevioside (21) Argentina, Brazil, Japan, United States Paraguay, People’s Republic of China, South Korea Glycyrrhizin (22) Japan European Union, United States Monoglucuronide of Japan glycyrrhetinic acid (MGGR) (23) Phyllodulcin (24) Japan Neohesperidin European Union, Switzerland, United States dihydrochalcone (25) Turkey Thaumatin Australia, European Union, United States Japan Mogroside V (26) Japan United States

Table 1.2. Internationally used natural and semi-synthetic high-potency sweeteners. (Varying degrees of purity are employed for applications of the above mentioned compounds).

26

COOH OR 2 OCH3 O H H OH O H COOR 1 RO OH O 2 21 22 R = -glcA --glcA 24 R1 R 2  23 R = -glcA OR2 -glc -glc2--glc OH OH RO OH OCH3 HO

OH O R1O

R1 R2 25 R = -glc2--rha 26 -glc6--glc -glc2--glc 6 -glc

Figure 1.7. Structures of some commercially used natural high potency sweeteners.

The triterpenoid glycyrrhizin (22), also named glycyrrhizic acid, is a sweet component of Glycyrrhiza glabra L. (Fabaceae) and other species with the common name “licorice.” A crude plant extract containing glycyrrhizin was used initially as a sucrose substitute in Japan in the early part of the 20th century (Mizutani and Tanaka,

2002). Although having a long history of human use, glycyrrhizin consumption does raise health concerns due to its known adrenocorticomimetic effects, which can lead to pseudoaldosteronism (Conn et al., 1968; Kinghorn et al., 2010; Kinghorn and Compadre,

2012). Investigation into derivatives of glycyrrhizin with better flavor profiles yielded a monoglucuronide derivative [MGGR (23)] as well as an ammoniated derivative, monoammonium glycyrrhizinate (Mitzutani et al., 1994; Anonymous, 2002; Kinghorn 27 and Compadre, 2012). MGGR, which is produced using microbial transformation, was found to be sweeter then the parent compound and is used to flavor chocolate milk and soft drinks in Japan (Mitzutani et al., 1994; Kinghorn and Compadre, 2012). Licorice products containing glycyrrhizin and monoammonium glycyrrhizinate have both attained

GRAS status in the United States when used as flavoring agents or surfactants, although they are not approved for use as sweeteners (Anonymous, 2002; www.fda.gov/food/foodingredientspackaging/foodadditives/ucm191033.htm).

Phyllodulcin (24), an isocoumarin, is responsible for the sweet taste of a ceremonial Japanese tea prepared from the leaves of Hydrangea macrophylla (Thunb.)

Ser. (Hydrangeaceae) (Asahina and Ueno, 1916). This compound is produced upon crushing or fermentation of the plant leaves, which allows for the hydrolysis of the glycosidic bond in the parent non-sweet glycoside to afford the sweet aglycone. The commercial use of phyllodulcin is restricted due to its poor water solubility and less than optimal taste properties (Kinghorn et al., 2010; Kinghorn and Compadre, 2012).

Attempts to produce derivatives that do not have these disadvantages have been unsuccessful thus far (Kinghorn et al., 2010; Kinghorn and Compadre, 2012).

Neohesperidin dihydrochalcone (25) is a highly sweet compound produced from the bitter flavonoid glycoside neohesperidin. Compound 25 was originally produced in order to lower levels of the latter bitter compound in the Seville orange, Citrus auranticum L. (Rutaceae) (Horowitz and Gentili, 1977; Dubois et al., 1977). Compound

25 displays a delayed onset in its sweetness and an aftertaste (Dubois et al., 1977).

Compound 25 is approved as a sweetener in the European Union and is used as a

28 flavoring agent in the United States, although the U.S. FDA has not considered an official toxicological literature review on the compound (Kinghorn and Compadre, 2011; http://www.accessdata.fda.gov/scripts/fcn/fcnDetailNavigation.cfm?rpt=eafusListing&id

=2985).

The thaumatins are a group of closely related intensely sweet proteins (1600 x the sweetness of sucrose on a weight basis) isolated from the fruits of West African plant

Thaumatococcus danielli Benth. (Maranaceae), with the major proteins thaumatin I and II isolated in 1972 (van der Wel and Loeve, 1972). The sweetness of the thaumatins was found to be dependent on the disulfide bonds present in the proteins, with six thaumatins

I, II, III, a, b, and c isolated and identified thus far (van der Wel and Loeve, 1972,

Crammer, 2008). Thaumatin is an approved sweetener in many countries and is used in the U.S. as a flavoring agent and a sweetener in a variety of products (Kinghorn and

Compadre, 2011). Thamautin is offically on the FDA EAF (Everything Added to Food) list, indicating that while it is a currently used food additive in the United States a toxicological literature search has yet to be considered by the FDA

(http://www.accessdata.fda.gov/scripts/fcn/fcnDetailNavigation.cfm?rpt=eafusListing&id

=2784).

The dried fruits of Siraitia grosvenorii (Swingle) C. Jeffrey ex A.M. Lu & Zhi Y.

Zhang (Cucurbitaceae), commonly known as “lo han guo,” produces a variety of sweet compounds, with the most abundant sweet constituent present being mogroside V (26).

Extractives of S. grosvenorii have been used as a sweetener in Japan for some time

(Kinghorn and Compadre, 2011). Also, a self-designated GRAS status for an extract

29

with enriched mogroside V content was received by the FDA with “no questions”

(Cheeseman, 2010a; Cheeseman, 2011b).

Additionally, a variety of plant natural products have been found to be sweet that

are not used commercially, representing many chemotypes, with most of these

compounds being of the flavonoid, isoprenoid, and protein classes (Kinghorn et al., 2010;

Kinghorn and Compadre, 2012).

2.3. Methods to discover new high-potency sweeteners.

While the sweet receptor most likely arose to monitor foods for simple

carbohydrate content, “the list of sweet compounds is vast and taxes the imagination to

derive any common feature” (Roper, 2007). This fact made defining the sweet receptor,

based on a consideration of its structurally diverse ligands extremely difficult. This was

the only possible method to probe the structure of the sweetness receptor before its recent

identification as Tas1R2::Tas1R3 (Roper, 2007). Initial work attempting to explain the

structural requirements for sweetness was proposed in 1919 (Oertly and Meyers, 1919).

This borrowed heavily from the then current theory of color, and proposed the existence

of the “auxogluc” and the “glucophore,” corresponding to the presence of a proton

attached to either a hydroxylated or halogenated carbon (Oertly and Meyers, 1919). An

updated version of this model was proposed in 1967 that stated the need for a hydrogen

bond donor or electron acceptor placed about 3 Å from an electron-donating group

(Shallenberger and Acree, 1967). This “AH-B” theory has been modified several times

subsequently in order to include the increasing number of compound classes of

30 sweeteners discovered with a recent proposal positing eight binding regions probed mainly with sweet guanidine and urea derivatives (Nofre and Tinti, 1996). This method of defining sweeteners, although useful for closely related compounds (Walters, 1991), has been largely overthrown by the discovery that various sweeteners act on sites quite distal to one another (Figure 1.8) (Temussi, 2002; Xu et al., 2004; Nie et al., 2005; Cui et al., 2006; Winnig et al., 2007; Slack, 2009).

Venus Flytrap Domains

Sugars, Sucralose Aspartame, Neotame

Cyclamate Brazzein Neohesperidin

Cell Membrane dihydrochalconeCell Membrane

Lactisole Perillartine hTas1R2 hTasT1R3

Transmembrane Domains

Figure 1.8. Diagram of the sweetness receptor and binding sites. (Binding sites based on Temussi, 2002; Xu et al., 2004; Nie et al., 2005; Morini, 2005; Cui et al., 2006; Winnig et al., 2007; Slack, 2009). *Thaumatin forces the venus flytrap domain to adapt the active configuration.

31

As has been stated previously, the only FDA-approved sweetener discovered by a ligand based design is neotame (Acree and Lindley, 2008). However, the properties of the sweetness inhibitor lactisole (27) (Figure 1.9) were discovered using this method (Acree and Lindley, 2008). It should be noted that although obtained by synthesis, lactisole occurs naturally in coffee beans (Coffea spp.) (Ley et al., 2008a).

O O OH

MeO 27

Figure 1.9. Structure of the naturally occurring sweetness-inhibitor lactisole.

As a result of the elucidation of the sweetness receptor, computational models of

Tas1R2::Tas1R3 to probe the binding pockets of various ligands have become more widespread (Morini et al., 2005). Also, site-directed mutagenesis can be used to investigate key interactions (Assadi-Porter et al., 2008). These approaches are a great boon in designing new sweeteners, while the crystal structure of these transmembrane proteins is being elucidated.

Another method for the identification of sweet compounds is the screening of compound libraries for sweetness. Due to the need to verify the safety of chemicals used in taste testing in human volunteers, and the fact that many compounds do not show the 32 same sweetness profile in animal models, this method has been impractical until the recent development of cell-based assays for sweetness.

2.4. Bulk sweeteners

The bulk caloric sweeteners are generally employed as sugar substitutes when the amount of sweetener is a factor in a food or beverage formulation (Kroger et al., 2006).

These are also included in table top sweetener compositions in order to provide an amount of sweetener comparable to sucrose for the same sweetness intensity (Kroger et al., 2006). In the United States, erythritol (28), isomalt (29), lactitol (30), maltitol (31), mannitol (32), sorbitol (33), tagatose (34), trehalose (35) and xylitol (36) (Figure 1.10) are currently used officially in food preparations, with mannitol, sorbitol, and xylitol being the most commonly employed (Kroger et al., 2006). Several of these sweeteners possess a cooling effect that is not acceptable for some foodstuff applications, but can be congruent with certain desired flavors (Levin et al., 1995). These compounds generally are poorly absorbed in comparison to sucrose and therefore provide fewer calories (Levin et al., 1995). This limited absorption results in the main side effect of bulk sweetener use, namely, diarrhea, so some products containing bulk sweetners must display the warning label “excess consumption may have a laxative effect” (Levin et al., 1995;

Kroger et al., 2006).

33

CH2O glc CH2OH CH2OH HO OH HO

CH2OH HO gal- O HO OH OH OH O -glc OH OH OH HO

CH2OH CH2OH CH2OH CH2OH 28 29 30 31

CH2OH CH2OH CH2OH HO HO O CH OH O 2 HO HO HO O O OH HO OH OH OH HO HO OH HO OH HO OH HO OH OH CH2OH CH2OH CH2OH OH OH CH2OH 32 33 34 35 36 Figure 1.10. Structures of selected bulk sweeteners.

It should be noted that although the bulk sweeteners generally provide fewer calories per gram than sucrose, their sweetness per gram calorie (here defined as sweetness efficiency) is nearly identical to sucrose for most of the bulk sweeteners, necessitating their use in combination with high-potency sweeteners (Table 1.3) (Levin et al., 1995; Kroger et al., 2006). A sweetness efficiency greater than one shows a bulk sweetener could replace sucrose in a preparation while lowering the caloric value of the foodstuff. Erythritol (28) is an exception to this trend and has a sweetness efficiency somewhat greater than sucrose (Levin et al., 1995; Kroger et al., 2006). D-Tagatose (29) seems to be up taken efficiently in the small intestine, with the remaining compound broken down by gut flora in the lower intestine (Levin et al., 1995; Levin, 2002).

Tagatose has an FDA-approved caloric content of 1.5 calories per gram, however inefficiencies in its breakdown may result in a net negative caloric value (Levin, 2002).

Most of the bulk sweeteners are non-cariogenic, with xylitol (30) possessing anti- 34 cariogenic activity, while tagatose has shown promising anti-diabetic effects (Levin,

2002; Kroger et al., 2006).

Sweetener Sweetness relative to sucrose Cal/g Sweetness efficiency Erythritol (28) 0.75 0.2 15 Lactitol (30) 0.4 2 0.8 Maltitol (31) 0.9 2.1 1.7 Mannitol (32) 0.5 1.6 1.25 Sorbitol (33) 0.6 2.6 0.92 Tagatose (34) 0.9 1.5 2.4 Xylitol (36) 1 2.4 1.7 Sucrose 1 4 1 Table 1.3. Comparison of some bulk sweeteners relative to sucrose. (Sweetness efficiency is a measure of sweetness per calorie normalized to sucrose).

3. Efforts to Modulate Taste Perception

3.1. Sweetness enhancers

Another strategy in producing low-calorie foods is the use of sweetness enhancers, as compounds that enhance the sweet taste of sugars would allow for their use at lower sweeteners concentrations while producing the same effect (Slack, 2007).

Additionally, sweetness enhancers could be used to improve the flavor profile of artificial sweeteners by lowering their levels to sub-threshold amounts for their off-tastes (Slack,

2007). Utilization of the synergistic effect of various sweeteners has been known already for a variety of sweeteners for quite some time (Moskowitz, 1973).

Indeed, all of the FDA-approved artificial high potency sweeteners discussed are known to display synergy with other sweeteners, including the bulk sweeteners (Morini et al., 2005). Additionally, rebaudiosides A (14), C (37), and D (38) have all been 35 patented recently for this purpose (Palmer and Salemme, 2010; Salemme et al., 2011).

While the goal of sweetness enhancement research generally is to find a compound that is itself tasteless and enhances the sweet taste of a given sweetener, sweet compounds active at sub-threshold levels are also acceptable.

This emphasis on sub-threshold activity is largely due to the difference in FDA safety requirements for flavoring agents versus sweeteners. Many classes of primary metabolites have been found to possess sweetness-enhancing effects including sugars, nucleic acids, peptides, and fatty acid esters (Prakash et al., 2009). The structures of a selection of such compounds are shown in Figure 1.11. The natural dihydrochalcone, trilobatin (39) isolated from Lithocarpus polystachyus (Wall. ex A. DC.) Rehder

(Fagaceae), and Malus trilobata C. K. Schneid. (Rosaceae), which has long been known as a sweetener, has also appeared in the patent literature as a sweetness enhancer along with the closely related compound, hesperetin dihydrochalcone 4"-O-β-D- glucopyranoside (40) (Jia et al., 2010). S-Ethyl-L-cysteine (41) was identified as a

“kokumi,” or flavor impact enhancing product from the Maillard reaction of Allium vegetables, and this reaction occurs commonly during cooking (George et al., 2011). This compound also slightly increased the sweetness of chicken broth, while not being perceived as sweet itself (George et al., 2011). Aladapcin (42), a bacterial metabolite from a Nocardia species, has been demonstrated as a sweetness enhancer using a cell- based assay (Krohn and Zinke, 2009). S-Alapyridaine (43), a product from the Maillard reaction between L-alanine and glucose, is also a sweetness enhancer (Ottinger et al.,

2003). Hesperetin (44), a flavonoid from the higher plant Eriodictyon californicum

36

Decne. (Hydrophyllaceae), is a patented sweetness enhancer as a single agent (Ley et al.,

2008c), or when used along with dihydrochalcones in a mixture with alkylamines (Ley et al., 2008b). (+)-Catechin (45), gallocatechin (46), (-)-epicatechin (47), and epigallocatechin (48) have also been patented as sweetness enhancers (Kashket, 1990).

OR2 OH OMe

RO OH RO OH H OH H COOR1 OH O OH O R1 R2 37 -glc -glc2--rha 39 R=-glc 40 R=-glc 3 -glc O NH2 38 -glc2--glc -glc2--glc 3 HO NH2 Me -glc S O H N COOH N OH H2N COOH H2N N H O O OH OH OH OH 41 42 43 OMe OH OH HO O HO O HO O OH

OH OH OH O OH OH OH OH 44 OH 45 OH 46 HO O HO O OH

OH OH OH OH 47 48 Figure 1.11. Structures of selected sweetness-enhancing compounds.

37

It is believed that compounds displaying synergy when used in combination occupy different binding sites on the sweetness receptor (Morini et al., 2005). In fact, the synergy profile of stevioside (21) has led to the hypothesis that it interacts with a yet-to- be discovered binding site (Morini et al., 2005). Figure 1.8 shows the possible sweetness enhancing capabilities of various sweeteners that might occur if used in combination.

3.2. Bitterness maskers

The most common industrial means of masking a bitter taste is by creating a physical barrier to the tastant, although this is not practical for all food applications or some liquid medicines (Ley, 2008). Physical barriers can include encapsulation (on both the micro and macro levels), forming suspensions, emulsions, coatings, and the use of complexing agents, such as ion-exchange resins. Strong flavors can be used to overpower bitter tastants, and congruent flavors such as chocolate, coffee, grapefruit, and mint can be used to place bitterness in a more favorable context (Ley, 2008; Sharma and

Lewis, 2010). Salts (NaCl and LiCl) have been shown generally to inhibit the bitterness of various bitter tastants, while sour tastants may effect bitterness perception unpredictably, with lower concentrations masking bitterness and higher concentrations showing enhancement (Breslin, 1996). Sweet substances are well known to suppress the intensity of bitter compounds (Ley, 2008). Umami substances have also been shown to mask bitter taste, but monosodium glutamate [MSG (49)] (Figure 1.12) also has salty and sour attributes owing to its component ions, which tends to obscure the relationship between umami and bitter taste perception (Fuke and Ueda, 1996).

38

OH OMe OH OH

HO O HO O NH Na+ 2 -O OH

O O OH O OH O OH O 43 44 OH 45 OH OH HO OH MeO O HO O

OH O OH O 46 47 48 OH O OH O OMe OH

MeO MeO MeO NH HO HO OH O 49 50 51 O O MeO H OH N OH HO MeO O OMe 52 53 54 55 OH OR2 OMe HO RO O OH OH HO OH

R1O OH O O O OH R R1 R2 O 6 56-glc2--rha OH 57 -glc -glc --glc O O OH HO O OH O O O O O HO O O OH OH HO O O O O OH OH HO O O O HO O HO O OH O R O R O HO OH OH OH OPO3H2 OH 58 59

Figure 1.12. Structures of selected of bitterness-masking compounds. (Structures for 58 and 59 are generic as the structures were not provided by Nakamura et al., 2002)

39

One potential method of developing palatable foods with beneficial secondary metabolites is to mask their bitter off-tastes with other compounds at concentrations below their taste threshold (Ley, 2008). The most desirable bitterness masker would be a compound with bitterness inhibition against a wide range of tastants that does not interfere with other tastes and does not have a taste itself at active concentrations (Sohi et al., 2004). A single receptor masker, however, can be important commercially if it blocks undesired bitterness of a food, food additive, beverage, or medicine.

Mining the scientific literature for examples of bitterness maskers active below their taste thresholds is a cumbersome task. A search of the available scientific literature for the concept of bitterness masking reveals that roughly ten percent of the approximately 560 distinct references on the subject could be construed as reporting examples of this method of masking bitter taste. However, this assessment may indeed cover compounds used above their limit of detection as many documents did not disclose the taste threshold of the compound used.

A particularly productive class of compounds for this method of bitterness masking has been the flavonoids, especially those isolated from the United States native plant Eriodictyon californicum. Thus, prior to the research reported in this dissertation, work by Ley et al. showed eriodictyol (50), homoeriodictyol (51) and its sodium salt, and sterubin (52), leaf flavonoids of E. californicum, to be bitter-masking agents (Ley et al.,

2005). The sodium salt of homoeriodicyol was of special interest, since it masked the bitterness of a variety of compounds including caffeine, guaifenisin, paracetamol, and salicin (Ley et al., 2005). This led to the investigation of several structural analogues,

40 including the dihydrochalcone floretin (53), a closely related flavan [7,4'-dihydroxy-3'- methoxyflavan (54)], various gingerdione derivatives {e.g., [2]-dehydrogingerdione (55) and [3]-gingerdione (56)}, and benzamide analogues such as vanillylamide (57) (Ley et al., 2006; Ley et al., 2008d, Ley et al., 2010; Wessjohann et al., 2010 ). A compound only tangentially related to the benzamide derivatives of homoeriodictyol, L-menthane carboxylic acid-N-(4-methoxyphenyl)-amide (58), was found to mask the bitterness of the similar compound, menthol (59) (Oertling et al., 2011). The relevant preliminary work could be the observation that the sodium salt of ferulic acid (60) masked the bitterness of acesulfame K, caffeine, quinine, and saccharin (Riemer, 1994). Earlier research on para-methoxycinnamaldehyde (61) as a sweetener led to a report that this compound has bitterness-masking capabilities when used in compositions containing vanillin and saccharin (Neely and Thompson, 1975). Another relevant observation is that neodiosmin (62) has been demonstrated to have bitterness-masking properties against limonin, caffeine, quinine, and saccharin (Guadagni et al.; 1976; Guadagni et al., 1979).

The previously mentioned mogroside V (26) from Siraitia grosvenorii has been patented for use below its taste threshold to block many bitter tastants including coffee, grapefruit, organ meat, and potassium chloride (Finley et al., 2011). In addition, the reportedly tasteless mogroside III (63), from the same plant source, also displayed bitterness-masking properties (Finley et al., 2011). These results indicate that while sweetness is known to mask bitter taste, high-potency sweeteners may be active in this regard at levels below their taste threshold. The activity of mogroside III, however,

41 implies that such activity may be particular to the mogroside class rather than a general activity of high-potency sweeteners.

Bitterness-maskers may also have multiple mechanisms of action, as inferred by work on phosphatidic acid (640 and tannic acid (65) (Nakamura et al., 2002). These compounds may be categorized as barriers to the tastant by their ability to adsorb bitter compounds (Nakamura et al., 2002). However, investigation of their bitterness-masking properties has indicated these compounds reduce the bitterness of quinine to below the amount scavenged from solution (Nakamura et al., 2002). Interestingly, tannic acid, while acting as a bitterness-masking agent at lower concentrations, was found to be bitterness-enhancing at higher concentrations, with this posited as being due to its astringent properties (Nakamura et al., 2002).

4. The Use of Cell-based Approaches in the Discovery of Flavor Modulators and

Sweetener Discovery

Besides furnishing knowledge of the binding sites of tastants to their receptors, current knowledge of the sweet and bitter receptors has also led to their utilization in cell- based assays to enhance the discovery of new taste-modulation agents and sweeteners. A cell-based approach to new sweetener screening has many benefits, such as reproducibility, especially by removing taster bias and allowing for the collection of more data points (the number of receptors activated vs. the number of human volunteers), which leads to better quantification of the results obtained. Additionally, this approach renders high-throughput screening approaches to taste testing possible, uses a minimal

42 amount of sample, by-passes most solubility issues, and reduces the need for safety tests necessary in conjunction with taste testing until after compounds are established as potential leads (Sheridan, 2004; Slack et al., 2010). The minimal use of sample and the ability to test compounds not readily soluble in water is a particular boon to the potential discovery of new sweeteners and taste-modifying agents of natural origin. This is readily evident when one considers the total amount of sample consumed over a series of chromatographic fractions in traditional taste tests. The ability to test compounds that are largely water insoluble allows for a greater amount of experimental data to be obtained, to more fully probe the structure-activity relationships of various compounds. Also, compounds can be identified as possessing potentially interesting taste properties before attempts at formulation to enhance solubility occur. There is an added advantage of employing a cell-based approach in identifying potential bitterness-masking agents, since due to the existence of multiple bitterness receptors, targeted bitterness-maskers can be created. High-throughput screening methods in the field of taste have indeed been a long time in development, as one recent paper touts a double-blinded screening method developed in 1966 as “state of the art” (Ley et al., 2008d). The use of human volunteers in taste panels, however, allows for a true taste profile to be recorded (Ley et al., 2008d).

43

CHAPTER 2: REVIEW OF ERIODICTYON CALIFORNICUM AND THE GENUS

ERIODICTYON

A. Background on Eriodictyon

1. The family Hydrophyllaceae

The family Hydrophyllaceae, commonly referred to as the waterleaf family, is a

Western hemisphere family mainly occurring in North America, with one genus, Codon,

restricted to South America (Bacon et al., 1986). In fact, the family is found in the

United States of America with sixteen native genera to the U.S., of which fourteen are

located mainly in the western states (Bacon et al., 1986). Hydrophyllaceae was formerly

known as Hydroleae and Hydroleaceae and is sometimes placed in the family

Boraginaceae as a subfamily called the Hydrophylloideae (Watson and Dallwitz, 1992;

http://www.mobot.org). The Hydrophyllaceae is known for the production of flavonoids,

with those produced by Eriodictyon reportedly being more structurally diverse than those

from other genera (Bacon et al., 1986). There is some confusion as to the number of

genera and species comprising the Hydrophyllaceae, with reports of seventeen or

eighteen genera inclusive of 250 or 300 species (Bacon et al., 1986, Watson and

Dallwitz, 1992). Hydrophyllaceae was been described as follows (Watson and Dallwitz,

1992): 44

Habit and leaf form. Herbs (usually), or shrubs (sometimes spiny); non- laticiferous and without coloured juice. Annual, biennial, and perennial; with a basal aggregation of leaves, or with neither basal nor terminal aggregations of leaves. Mesophytic. Leaves alternate, or opposite; spiral; petiolate; not connate; non-sheathing; not gland-dotted; without marked odour, or foetid; simple, or compound; when compound, pinnate, or palmate (rarely). Lamina dissected, or entire; when simple/dissected, pinnatifid; pinnately veined, or palmately veined (rarely); cross-venulate. Leaves exstipulate; without a persistent basal meristem. Leaf anatomy. Stomata present; mainly confined to one surface, or on both surfaces; anomocytic. Lamina dorsiventral, or isobilateral. Minor leaf veins with phloem transfer cells (5 genera), or without phloem transfer cells (Wigandia). Stem anatomy. Cork cambium present; initially superficial. Nodes unilacunar. Secondary thickening developing from a conventional cambial ring. Xylem with tracheids, or without tracheids. Vessel end-walls simple. Wood parenchyma apotracheal (diffuse). Reproductive type, pollination. Fertile flowers hermaphrodite. Plants hermaphrodite. Pollination entomophilous; via Hymenoptera. Inflorescence, floral, fruit and seed morphology. Flowers aggregated in ‘inflorescences’ (usually), or solitary (rarely); in cymes. The ultimate inflorescence unit cymose. Inflorescences ‘boragoid’ cincinni. Flowers often ebracteolate; regular; usually 5 merous; cyclic; tetracyclic. Free hypanthium absent. Hypogynous disk absent (usually), or present. Perianth with distinct calyx and corolla; (8–)10(–20); 2 whorled; isomerous. Calyx (4–)5(–10); 1 whorled; gamosepalous, or polysepalous (sometimes ‘lobes divided to the base’); regular; imbricate. Epicalyx present (as appendages between the calyx lobes), or absent. Corolla (4– )5(–10); 1 whorled; appendiculate (often having scales inside the tube, alternating with the stamens), or not appendiculate; gamopetalous; imbricate, or contorted; rotate, or campanulate, or funnel-shaped; regular; blue, or purple, or white. Androecium (4–)5(–10) (as many as C). Androecial members adnate (to the corolla tube, and usually with basal appendages also united to the corolla, which in Hydrophyllum form tubes leading to the nectar); all equal, or markedly unequal; free of one another; 1 whorled. Androecium exclusively of fertile stamens. Stamens (4–)5(–10); inserted near the base of the corolla tube; isomerous with the perianth; oppositisepalous (alternating with the petals); alternating with the corolla members. Filaments variously, basally appendiculate. Anthers dorsifixed; versatile; dehiscing via longitudinal slits; introrse; appendaged. Microsporogenesis simultaneous. Tapetum glandular. Pollen grains aperturate; 3 aperturate, or 5–6 aperturate; colpate, or colporate, or rugate; 2-celled (in 3 genera). Gynoecium 2 carpelled. The pistil 1 celled, or 2 celled. Gynoecium 45

syncarpous; synovarious to synstylovarious; superior (usually), or partly inferior (sometimes). Ovary 1 locular, or 2 locular. Gynoecium median; stylate. Styles 1, or 2; free, or partially joined; attenuate from the ovary; apical. Stigmas dry type; papillate; Group II type. Placentation when unilocular, parietal; when bilocular, axile. Ovules in the single cavity when unilocular, 2–100 (i.e. to ‘many’); when bilocular 2–50 per locule (i.e. to ‘many’); funicled, or sessile; pendulous (when funicled); epitropous (the micropyle directed upwards and outwards); non-arillate; anatropous, or amphitropous; unitegmic; tenuinucellate. Endothelium differentiated. Embryo-sac development Polygonum-type. Polar nuclei fusing prior to fertilization. Antipodal cells formed, or not formed (then the three nuclei degenerating early — Hydrolea); when formed, 3; not proliferating; ephemeral. Synergids sometimes with filiform apparatus. Endosperm formation cellular, or nuclear, or cellular to nuclear. Embryogeny solanad. Fruit non-fleshy; dehiscent (usually), or indehiscent; a capsule, or capsular-indehiscent. Capsules when dehiscent, loculicidal (usually), or splitting irregularly, or septicidal (rarely). Seeds copiously endospermic. Endosperm oily. Cotyledons 2. Embryo chlorophyllous (1/1), or achlorophyllous (1/2); straight (spathulate or linear). Seedling. Germination phanerocotylar. Geography, cytology. Temperate to tropical. Widespread. X = 5–13(+).

2. Background on the genus Eriodictyon and species studied phytochemically

The genus Eriodictyon is comprised of up to ten species located mainly in the

American West and Baja California, Mexico (Bacon et al., 1986). Plants of the genus

Eriodictyon have unappendaged bell-shaped flowers, with most species having the inflorescence occur as flowers on alternating sides (Bacon et al., 1986). The fruits of the plants are capsular and are split lengthwise and on the septa present in the fruit (Bacon et al., 1986). The leaves vary among the species in the genus in the density of the leaf trichomes (plant hairs) as well as the leaf shape and size and flower size (Bacon et al.,

1986). Phytochemical work on this genus, though not extensive, has generally shown that leaf resin flavonoids are the main secondary plant metabolite constituents of these

46 plants (Bacon et al., 1986; Bohm and Constant, 1990).

2.1. Eriodictyon angustifolium Nutt.

Aside from Eriodictyon californicum, E. angustifolium is the most thoroughly studied member of the genus Eriodictyon on a phytochemical basis. The natural range of the plant is of note as it is the only member of the genus that occurs naturally beyond the

American West with an eastward range as far as Maine (http://plants.usda.gov). The initial study of the leaves of E. angustifolium led to a report of the isolation of homoeriodictyol (51), 3-hydroxy-o-toluic acid (60), L-rhamnose (61), and salicylic acid

(62) (Figure 2.1) (Hadley and Gisvold, 1944). The identification of these compounds was performed by mass spectrometric determinations, chemical reactions, and melting point analysis, with the purity of salicylic acid confirmed by mixture melting point determination using a pure standard (Hadley and Gisvold, 1944). The constituents of the flowers, leaves, and stems of E. angustifolium were also compared in a chemotaxonomic study along with those of E. tomentosum Benth. and E. californicum (Bacon et al., 1986).

This study led to reports of apigenin (63), chrysoeriol (64), hispidulin (65), homoeriodictyol (51), jaceosidin (66), and nepetin (67) as constituents of E. angustifolium (Bacon et al., 1986). In another study, an extract of the E. angustifolium was screened against the microbes Staphylococcus aureus, Bacillus subtilis, Klebsiella pneumoniae, and Candida brassicae, and benzyl-trans-4-coumarate (68) and the flavanones naringenin (69), homoeriodictyol (51), 4'-methoxyhomoeriodictyol (70), 7- methoxyhomoeriodictyol (71), and sakuranetin (72) were isolated (Dentali and 47

Hoffmann, 1992). Compound 68 was found to be weakly active against S. aureus, B. subtilis, and C. brassicae (25-100 μg/mL) (Dentali and Hoffmann, 1992).

OMe OH

HO O OH HO OH HO HO OH HO O OMe OMe O HO O OH O OH OH OH OH 61 62 51 60 HO O HO O HO O HO O

MeO MeO OH O OH O OH O OH OH O OMe 65 66 63 OH 64 OH OMe O HO O HO O HO O O MeO HO OH O OH O OH O 68 67 OMe 69 70 OH OH O OH MeO O MeO O O OH HO OH O OH O OH 71 O OH 72 O OH 73

O O OH HO OH OH O OH 74 O OH 75 O OH

O O O OH

O O OH HO HO 76 77 78 Figure 2.1. Compounds isolated from Eriodictyon angustifolium (relative configuration depicted as shown in the original reports).

48

The most recent study on E. angustifolium was on the flavor properties of erionic acids A-F (73-78), with erionic acids C-E being reported as bitter among other tastes and erionic acid F reported as “slightly bitter, sweet, phenolic and woody” (Reichelt et al.,

2010). Erionic acid C (75), when mixed with the sodium salt of homoeriodictyol (51) in equal amounts (100 mg/kg in aqueous solution), did not modify the bitterness of a 500 mg/kg solution of caffeine (-14% bitterness) significantly, while the homoeriodictyol salt was determined to be bitterness-masking (-38% bitterness) and erionic acid C (75) increased the bitterness of the solution (+31% bitterness) (Reichelt et al., 2010). The mixture of bitter-tasting compounds and bitterness-masking compounds in this plant suggests that it is unlikely that a simple extract of the plant can be used as a bitterness- masking formulation if the erionic acids are present.

2.2. Eriodictyon sessilifolium Greene

Eriodictyon sessilifolium, which is native to Baja California, Mexico, was studied also in order to better understand the phytochemical composition of the genus (Arriaga-

Giner et al., 1988). In this investigation three benzoic acid derivatives (79-81) (Figure

2.2) were found to be constituents of the plant (Arriaga-Giner et al., 1988). Compound

81 could only be inferred as a constituent as due to chromatographic limitations only acetylated and methoxylated derivatives of the compound could be isolated, and its identity was proposed by comparing the methoxylation and acetylation patterns to determine natural substitution patterns (Arriaga-Giner et al., 1988). The flavonoids acacetin (82), apigenin (63), chrysin (83), genkwanin (84), hispudulin (65), isorhamnetin 49

(85), kaempferid (86), pectolinarigenin (87), and velutin (88) were also identified in this study, but it should be noted that the occurrence of these flavonoids was reported solely on the basis of thin-layer chromatography (TLC) and ultraviolet-visible light spectroscopy (Arriaga-Giner et al., 1988).

OH OH

HO O HO O O

MeO OH MeO OH OH O OH O OH 63 65 79 OMe

O O HO O OH MeO MeO OH O O OH O 80 81 82 OMe OH OH HO O MeO O HO O

OH OH O OH O OH O OMe 83 OH 84 OMe 85 OH

HO O HO O MeO O

OH MeO OH O OH O OH O 86 87 88

Figure 2.2. Compounds identified from Eriodictyon sessilifolium (relative configuration depicted as shown in the original report).

50

2.3. Eriodictyon tomentosum Benth.

Eriodictyon tomentosum is a plant species restricted to California and is commonly known as “wooly yerba santa” due it its abundant plant hairs (Bacon et al.,

1986; plants.usda.gov). The only phytochemical study of the plant is the previously mentioned chemosystematic work on E. angustifolium, E. californicum, and E. tomentosum (Bacon et al., 1986). In this work, it was determined that apigenin (63), cirsimartin (89), hispidulin (65), kaempferol 3-O-glucoside (90), nepetin (67), pectolinargenin (91), and quercetin 3-O-glucoside (92) (Figure 2.3) are constituents of E. tomentosum (Bacon et al., 1986).

OH OH OH OH

HO O HO O HO O

MeO MeO OH O OH O OH O 63 OH 65 OH 67 MeO O HO O

MeO O -glc OH O OH O 89 90 OH OMe OH

HO HO O

MeO O -glc OH O OH O 91 92 Figure 2.3. Compounds isolated from Eriodictyon tomentosum (relative configuration depicted as shown in the original report).

51

2.4. Eriodictyon trichocalyx var. trichocalyx A. Heller

Eriodictyon trichocalyx is another Eriodictyon species that occurs primarily in

California and also has the common name “hairy yerba santa,” once again due to the presence of plant hairs (plants.usda.gov). This plant was studied in order to further advance the understanding of the chemotaxonomy of the genus Eriodictyon (Bohm and

Constant, 1990). Several flavonoids were identified from the plant, namely, apigenin

(63), chrysoeriol (64), eriodictyol (50), hispidulin (65), homoeriodictyol (51), isorhamnetin (85), luteolin (93), naringenin (69), nepetin (67), and velutin (88) (Figure

2.4) (Bohm and Constant, 1990).

OH OMe OMe OH OH OH OH

HO O HO O HO O HO O

OH O OH O OH OH O OH O 50 OH 51 OH 63 OH 64 HO O HO O HO O

MeO MeO OMe OH OH O OMe OH O OH O 65 OH 67 OH 69 OH HO O MeO O HO O

OH OH O OH O OH O 85 88 93 Figure 2.4. Compounds isolated from Eriodictyon trichocalyx var. trichocalyx (relative configuration depicted as shown in the original report).

52

2.5. Eriodictyon californicum Decne.

2.5.1. Background on Eriodictyon californicum

Eriodictyon californicum is the most intensely investigated member of the genus thus far with over 50 research papers and patents focusing on this plant having appeared.

It should also be noted that Eriodictyon glutinosum was a formerly used synonym for this species (Ritter, 1895). E. californicum is a highly branched evergreen shrub native to

California and Oregon in the U.S., and to northern Mexico, which can reach a height of

2.4 m (Johnson, 1983; Howard, 1992). E. californicum has a well developed root system and can reproduce sexually or vegetatively through its rhizomes (Howard, 1992). The plant was classified originally as Wigandia californica (Hydroleaceae) by Hooker and

Arnott, albeit with questions due to the absence of fully formed fruits in the collection made (Hooker and Arnott, 1841). It should be mentioned that Wigandia and Eriodictyon along with the monotypic Turricula are the only woody members of the plant family

Hydrophyllaceae (Carlquist et al., 1983, Bacon et al., 1986).

The leaves of E. californicum develop a resinous covering rich in flavonoid aglycones that have been implied as having various beneficial roles for the plant, including the prevention of damage due to ultraviolet light, the deterrence of predation, prevention of desiccation, and the promotion of wildfires, which plays a role in the reproduction cycle of the plant (Johnson, 1983, Howard, 1992). It has also been proposed that the flavonoids play a mechanical role in the plant by modifying the viscosity of the leaf resin (Johnson, 1983). The leaves of E. californicum have a long history of medicinal use in the United States, including by Native Americans, and the 53 plant appeared in the United States Pharmacopeia as early as 1890 (Ritter, 1895; Power, and Tutin 1906). Oral administration of the plant has been used for the treatment of respiratory problems such as coughs, colds, asthma, bronchitis, and, in the past, tuberculosis; a topical balm prepared from the plant has been used for bruises, sprains, cuts, poison ivy, and insect bites (Ritter, 1895; Follin and Turkington, 2004). These properties have resulted in many common names for E. californicum, including

“consumptive’s weed”, “holy weed”, “mountain balm”, “yerba santa” and “sacred herb”

(Tutin and Power, 1906; Follin and Turkington, 2004). In 1887, an extract of E. californicum leaves was reported to mask the bitterness of quinine, and according to the author “a syrup prepared from Eriodictyon leaves is extensively used for the administration of quinine in a bitterless form” (Rother, 1887). Indeed by 1917, it was reported that “its preparations are principally used, however, as vehicles to disguise the taste of disagreeable medicines like quinine” (Sayre, 1917). The plant was also reported to possess an initial bitter taste that turns sweetish, further indicating its interesting flavor properties (Anonymous, 1916).

2.5.2. Early phytochemical investigations of Eriodictyon californicum

Quite early on in the phytochemical investigation of the plant, the flavonoids eriodictyol (50) and homoeriodictyol (51) were isolated, along with chrysoeriol (64)

(Figure 2.5), which was mentioned in the report as a phenolic compound with the formula

“C16H41O6” but with an editorial note suggesting C16H14O6 as the correct molecular formula (Power and Tutin, 1906). This typographical error of the true molecular formula, 54

OH OH OMe OH OMe OH OH OH

HO O HO O HO O MeO O

OH O OH O OMe OH O OH O OMe 44 OH 50 OH 51 OH 52 OH

HO O HO O HO O HO O

MeO MeO OH O OH OH O OH O OMe OH O 63 OH 64 OH 65 OH 66 OH HO O HO O MeO O MeO O

MeO OH OH O OH O OH OH O OH OH O OH 67 69 71 72 HO O MeO O HO O HO O

MeO O -glc O -glc OH O OH OH O OH OH O OMe OH O 83 89 90 92 OH OMe O OMe

HO O MeO O HO O O HO O

OH O OMe OH O OH O OH O 93 94 95 96 OH OH O OH

HO O MeO O HO O OH

MeO MeO HO OH O OH O O OH OH O OMe 97 98 99 100 OH

HO OH O OH 101 O OH

OH OH

OH OMe 102 103

Figure 2.5. Compounds isolated from Eriodictyon californicum (relative configuration depicted as shown in the original reports).

55

C16H12O6, was corrected in a second report on E. californicum, with the compound identified as chrysoeriol (64) (Tutin and Clewer, 1909). Although earlier reports exist on the chemical examination of E. californicum none of the reported components was verified later with modern spectroscopic techniques (Power and Tutin, 1906). Indeed,

Power and Tutin discounted the earlier compound identifications as needing revision

(Power and Tutin, 1906). The initial phytochemical investigation on E. californicum led to a report of the presence of natural rubber, sugars, wax, and a lack of alkaloids

(Holzhauer, 1878). Power and Tutin also reported the substances triacontane and pentatriaconane, which are possible components of the leaf resin of the plant (Power and

Tutin, 1906). A phloroglucinol (formerly referred to as “phloroglucin”) derivative was also cited as isolated in a previous work by other authors with the formula, C14H18O5

(Power and Tutin, 1906). Flavanones were also classified as phloroglucinol derivatives, due to their degradation with potassium hydroxide (Figure 2.6) (Tutin, 1910, Whiting,

2001). Chemical interconversions were used as a primary method of structure elucidation at the time, and perhaps this was an impure flavonoid thus producing an incorrect molecular formula (Tutin, 1910, Whiting, 2001). Other compounds proposed in early work on E. californicum include “eriodonol” and “xanthoeridol” with chemical formulas assigned as C19H18O7 and C18H14O7, respectively (Tutin, and Clewer, 1909).

However, the identification of these compounds has not been determined by more recent phytochemical investigation.

56

OH OH OH HO O KOH HO OH OH + O2 OH O OH phloroglucinol

Figure 2.6. Degradation leading to the identification of flavonoids as phloroglucinol derivatives (Whiting, 2001).

Quite impressively, the structures of eriodictyol (50) and homoeriodictyol (51) were nearly correctly elucidated by degradation soon after their initial isolation, with the only error being that they were assigned a chalcone skeleton instead of a flavanone skeleton (Tutin, 1910). By analogy, the structures of the flavanones hesperetin (44) and naringenin (69), not known earlier to be constituents of E. californicum, were also proposed in this nearly correct chalcone form (Tutin, 1910). In fact, the elucidation of flavanones (and chalcones) from natural sources with a correct structure was only newly reported at that time (Perkin and Hummel, 1904). Moreover, the conversion of chalcones to flavanones by chemical means was also only recently proposed, as shown in Figure 2.7

(Perkin and Hummel, 1904).

There were two major factors that lead to the misclassification of compounds 44,

50, 51, and 69 as chalcones. The first was that the methods used by Tutin and co-workers to produce methoxy and acetyl derivatives of these flavonoids were harsh, e.g., by boiling in excess acetic anhydride (Tutin, 1910, Asahina and Inubuse, 1928). This led to the

57 fragmentation of the C ring in the flavanones thus treated, indicating erroneously an additional hydroxy group to be present in the parent molecule (Tutin, 1910, Asahina and

Inubuse, 1928). Indeed, the number of hydroxy functional groups incorporated into the structure of hesperetin (44) was contested, with Perkin obtaining a triacetylated derivative, thereby indicating the presence of three hydroxy groups (Perkin, 1898, Tutin

1910). The second misleading piece of information pertinent to the structure of these flavanones was whether or not a chiral center was present. The basic conditions used in the serial partitioning employed in order to obtain these compounds most probably caused their racemization in the work performed by Tutin et al. (Figure 2.8) (Krause and

Galensa, 1991). However, Mossler isolated homoeriodictyol (51) and reported the compound as having a chiral center (Mossler, 1908; Tutin, 1910). It would be nearly twenty years later until sufficient evidence was obtained that these compounds are indeed of the flavanone class (Asahina and Inubuse, 1928).

OH OH + HO OH [H ] HO O OH OH

O O

Figure 2.7. The interconversion between flavanones and chalcones discovered by Perkin and Hummel (Perkin and Hummel, 1904).

58

O O O

H OH O O H2O O Figure 2.8. Proposed mechanism for the base-catalyzed racemization of a flavanone.

The first study investigating the biological activity of the secondary metabolites of E. californicum was a screening of fractions from the leaves and eriodictyol (50) and homoeriodictyol (51) against a wide array of microbes (Salle et al., 1951). This work may have been inspired by the ethnobotanical knowledge of the use of E. californicum as a treatment for tuberculosis. Eriodictyol (50) was determined as possessing a weak activity against a series of microbes, while a fraction named “ericolin” was presented as the most active “constituent” of the plant (Salle et al., 1951). “Ericolin” was determined to inhibit the growth of Bacillus subtilis, Chromobacterium violaceum, Micrococcus pygenes var. aureus, Micrococcus roseus, Micrococcus ureae, Mycobacterium tuberculosis 599, Sacrina lutea, Sacrina ureae, and Streptococcus pyogenes (Salle et al.,

1951). Interestingly, this is the only peer-reviewed report in which an attempt was made to identify active constituents of E. californicum for an ethnobotanical use, although a report from the same period did show an ethanol extract of the leaves of E. californicum to be active against Mycobacterium tuberculosis and Escherichia coli (Gottshall et al.,

1949).

59

2.5.3. Recent studies on Eriodictyon californicum

Research on the constituents of E. californicum lulled for over thirty years, until phytochemical work in the 1980s. These studies have focused on the roles of leaf flavonoids in plant survival and in chemotaxonomy. The flavonoids chrysoeriol (64), luteolin (93), hesperetin (44), homoeriodictyol (51), 7-methoxyhesperetin (94), and 7- methoxyhomoeriodictyol (71) were found in an investigation on the role of the leaf resin flavonoids of E. californicum as potential antifeedants and in serving as screens for ultraviolet light (Johnson, 1983). It was determined that different constituents of the plant are responsible for each activity (Johnson, 1983). Additional roles of the flavonoids in the plant were proposed, namely in modulating the viscosity of the leaf resin and in water retention (Johnson, 1983). A study on the chemotaxonomy of Eriodictyon revealed apigenin (63), homoeriodictyol (51), hispidulin (65), jaceosidin (66), kaempferol 3-O- glucoside (90), luteolin (93), nepetin (67), and quercetin 3-O-glucoside (92) as constituents (Bacon et al., 1986).

The next published study on this plant was performed in the laboratory of John

Cassady at The Ohio State University, who investigated the chemopreventive potential of the flavonoids from this species. This study led to the identification of 12 flavonoids, namely, eriodictyol (50), chrysin (83), chrysoeriol (64), cirsimaritin (89), hispidulin (65), homoeriodictyol (51), 4'-isobutyrylhomoeriodictyol (95) (a new compound), isosakuranetin (96), 6-methoxyhomoeriodictyol (97), 6-methoxysakuranetin (98), pinocembrin (99), and sakuranetin (72) (Liu et al., 1992). These compounds were tested against benzo[α]pyrene metabolism, considered to be a step in activating carcinogens,

60 and for their ability to prevent the binding of derivatives to DNA (Liu et al., 1992). It was determined that chrysoeriol (64) and cirsimartin (89) were the most active of the compounds isolated in this study (Liu et al., 1992).

The most recent work on Eriodictyon californicum has focused on the flavor modulatory properties of the flavonoids of the plant, as conducted by a group at Symrise

GmbH & Co. KG, Holzminden, Germany. The historical use of the leaf extracts of E. californicum to mask the bitterness of quinine was the inspiration for the initial work

(Ley et al., 2005). The monosodium salt of homoeriodictyol (51) and sterubin (52) were isolated in this work, and additionally eriodictyol (50), hesperetin (44), homoeriodictyol

(51), and naringenin (69) were also obtained for sensory evaluation (Ley et al., 2005). It was determined that the isolated flavonoids were racemic mixtures (Ley et al., 2005).

This loss of chirality and the presence of a phenolic sodium salt are probably due to the use of sodium carbonate in the isolation procedure (as shown in Figure 2.8). It was determined that eriodictyol (50) and the sodium salt of homoeriodictyol (51) were the most active compounds tested for blocking the bitterness of caffeine. The sodium salt of homoeriodictyol was found to also mask the bitterness of guaifenesin and paracetamol significantly (Ley et al., 2005). Eriodictyol (50), homoeriodictyol (51), and the sodium salts of homoeriodictyol were also patented as bitterness maskers (Ley et al., 2002). The isovanillic feature of the B-C rings of homoeriodictyol (51) served as the basis for the design of several synthetic bitterness-masking derivatives, as reviewed in the previous chapter.

61

Hesperetin (44) was also isolated from the leaves of E. californicum and determined to be a sweetness-enhancing compound (Ley et al., 2008c). A careful investigation of the earlier literature on hesperetin (44) indicates such a potential application for this compound. Upon tasting the compound neat, hesperetin (44) was found to be flavorless, but it was demonstrated as being intensely sweet in alcoholic solution (Tutin, 1910). These results could be interpreted as hesperetin (44) enhancing the sweetness of alcohol.

The most recent work on E. californicum by the Symrise group was conducted alongside E. angustifolium, and compared the flavor properties of benzoic acid derivatives from these plants (Reichelt et al., 2010). The prenylated benzoic acid derivatives eriolic acids A-D (100-103) were isolated in this study, with eriolic acids A

(100), B (101), and D (103) determined to be bitter (Reichelt et al., 2010).

Additionally, various leaf wax components have been identified as constituents of the plant using gas chromatography-mass spectrometery (GC-MS) (Ley et al., 2005).

This work identified tetradecyloctadecanoate, and compounds with the corresponding formulas C27H56, C28H58, and C30H62, C31H64, and two compounds with the formula

C29H60 (Ley et al., 2005).

62

CHAPTER 3: ISOLATION, IDENTIFICATION, AND IN VITRO EVALUATION

OF FLAVOR-MODULATING FLAVONOIDS FROM ERIODICTYON

CALIFORNICUM

A. Statement of Problem

Previous studies on Eriodictyon californicum have revealed some constituents of this plant to have bitterness-masking or sweetness-enhancing properties, as has been discussed in Chapters 1 and 2. Additionally, the older literature on the plant, as also previously discussed, has indicated interesting flavor properties of the plant inclusive of bitterness-masking and sweetness-enhancing effects. However, relatively few of the flavonoids from the plant have actually been investigated pertaining to their role in taste modulation, and all such studies have involved sensory testing by human volunteer subjects.

In collaboration with the Givaudan Flavors Corporation, Cincinnati, OH, the flavor properties of flavonoids isolated from E. californicum were investigated in some newly developed cell-based assays. As discussed in Chapter 1, the availability of cell- based screens is a very recent advance in the field of taste, and the presented work probably represents the first use of these methods in an academic research project. This

63 work also represents the most comprehensive investigation of the bitterness-masking and sweetness-enhancing capabilities of the leaf flavonoids of E. californicum to date. In this chapter, their isolation, identification, and of taste-modulatory evaluation are described.

B. Experimental

1. General Experimental Procedures

Melting points were measured using a Fisher-Johns melting point apparatus.

Optical rotations were procured with a Perkin-Elmer model 343 polarimeter. Circular dichroism (CD) spectra were recorded on a JASCO J-810 spectropolarimeter. Infrared

(IR) spectra were obtained on a Nicolet 6700 FT-IR spectrometer. Nuclear magnetic resonance (NMR) spectra were taken at room temperature on Bruker Avance DRX-300

MHz and DRX-400 MHz NMR spectrometers. High-resolution mass spectra were recorded on a LCT-TOF mass spectrometer. High-performance liquid chromatography

(HPLC) was performed using a semi-preparative reversed-phase phenol column (YMC pack-Ph, 150 mm x 20 mm i.d.), with Hitachi Prep-36 pumps, a Hitachi L-2200 Elite

LaChrom autosampler, and a Hitachi L-2400 Elite LaChrom UV detector. The silica gel employed for column chromatography in this study and was 65x250 mesh silica gel

(Sorbent Technologies, Inc., Norcross, GA). Analytical thin-layer chromatography

(TLC) was performed on 200 μm-thick aluminum-backed silica gel plates (Sorbent

Technologies, Inc.) with shortwave UV (254 nm) light and a vanillin-sulfuric acid spray

64 reagent (5% concentrated H2SO4 in ethanol with 10% vanillin) employed for visualization.

2. Plant Material

Leaves of Eriodictyon californicum Decne. (Hydrophyllaceae) were collected by

Dr. Richard Spjut at Whiskeytown Lake, California, in September, 2002, and a voucher specimen is deposited at the World Botanical Associates Herbarium, Laurel, Maryland, under the accession number WBA-4400-33.

3. Extraction of the leaves of Eriodictyon californicum

The dried leaves of E. californicum (2.0 kg) were ground and extracted exhaustively with methanol (30 x 4 L), yielding 525 g of extract. This dried methanol extract was then partitioned with ca. 375 mL of 9:1 methanol-water and 375 of mL hexane, in aliquots of approximately 60 g at a time. The methanol/water layer was evaporated to a thick tar and then partitioned with approximately 375 mL of chloroform and 375 mL of water. The aqueous layer afforded 200 g in total of a water-soluble extract, and, at the interface between the solvent layers, a viscous syrup (170 g) formed that was found to be insoluble in a wide variety of solvents, while only partially soluble in methanol. The resulting chloroform layer was partially detannified with 1% NaCl in water to give a final chloroform-soluble extract of 110 g. The fractionation tree employed is depicted in Figure 3.1.

65

Methanol extract

Hexanes Methanol: Water (9:1)

Aqueous Chloroform

1% NaClaq Chloroform extract

Figure 3.1. Fractionation tree employed for Eriodictyon californicum leaves. (Scheme adapted from Wall et al., 1996)

4. Chromatography of the chloroform-soluble extract of the leaves of Eriodictyon californicum

The chloroform-soluble extract of Eriodictyon californicum was separated using vacuum-liquid chromatography (VLC) over silica gel (180 g) employing a step gradient from hexanes, to ethyl acetate, to methanol; the fractions obtained were pooled into a total of five major fractions. The step gradient employed was as follows: 2 L hexanes,

66

500 mL 90% hexanes in ethyl acetate, 6.5 L 80% hexanes in ethyl acetate, 500 mL 60% hexanes in ethyl acetate, 4.5 L 40% hexanes in ethyl acetate, 1 L ethyl acetate, 1.5 L 50% ethyl acetate in methanol, and 3 L methanol. Collections with the exception of the initial

2 L hexane collection were taken in 500 mL increments. Fraction 2 (73.8 g), which eluted with 20%-60% ethyl acetate in hexane (collections 9-29), was selected for further workup. This fraction was then subjected to Diaion® HP-20 column chromatography to remove chlorophylls, yielding 48.5 g of a flavonoid-rich fraction, as indicated by the formation of densely packed impure pale yellow crystals, in a solid yellow matrix upon drying the fraction. This fraction was obtained with 90:1 methanol-water as the mobile phase. This fraction was then subjected to open column chromatography with a silica gel stationary phase (510 g) and a step-gradient of hexane to ethyl acetate as the mobile phase, yielding ten sub-fractions. The gradient used was as follows: 500 mL hexanes,

500 mL 99% hexanes in ethyl acetate, 1 L 95% hexanes in ethyl acetate, 7 L 90% hexanes in ethyl acetate, 8 L 85% hexanes in ethyl acetate, 8.5 L 80% hexanes in ethyl acetate, 5.5 L 75% hexanes in ethyl acetate, 4.5 L 70% hexanes in ethyl acetate, 6 L 60% hexanes in ethyl acetate, 4 L 50% hexanes in ethyl acetate, 4.5 L 30% hexanes in ethyl acetate, and 4 L 100% ethyl acetate. These were collected in 500 mL aliquots and combined as follows: aliquots 1-8, 9-12, 13-21, 22-32, 33-40, 41-49, 50-65, 66-91, 92-

100, and 101-110. Sub-fraction 7 was of particular note as it afforded a precipitate much like the flavonoid-rich fraction from the Diaion® HP-20 separation, as well as a syrupy mother liquor. These were separated and treated as separate fractions, as described below.

67

Pinocembrin (99) (4.8 mg) was isolated via HPLC from subfraction 4 (tR 46.1 min) using a water to methanol gradient (30% from 0-20 min, 50% MeOH 40 min, 70%

MeOH 70 min). Sakuranetin (72) (300 mg) was isolated via HPLC from subfraction 6 (tR

51.6 min) using a water to methanol gradient (30%-40% MeOH 25 min, 40%-45%

MeOH 80 min), and 4′-isobutyrylhomoeriodictyol (95) (16.0 mg) was also obtained from this subfraction using the same HPLC method (tR 78.3 min). The precipitate yielded homoeriodictyol (51) (10 g), which was obtained by repeatedly washing the precipitate with dichloromethane, while naringenin (69) (22.5 mg) and hesperetin (44) (6.9 mg) were also isolated from the dichloromethane wash of this precipitate, employing HPLC with a water to methanol gradient (0%-5% MeOH 5 min, 5%-30% MeOH 15 min, 30%-40%

MeOH 40 min, 40%-100% MeOH 120 min), with tR values of 43.5 and 49.9 min, respectively. The supernatant of subfraction 7 yielded 6-methoxyhomoeriodictyol (97)

(6.0 mg), 6-methoxyhesperetin (104) (8.0 mg), and 6-methoxysakuranetin (98) (6.0 mg), which were then separated via HPLC with a water to methanol gradient (10%-100%

MeOH over 90 min), eluting at 49.3 min, 51.1 min, and 55.8 min, respectively.

Jaceosidin (66) (7.7 mg) was isolated via silica gel chromatography from subfraction 9 employing dichloromethane as the mobile phase. The structures of the compounds isolated from E. californicum in the present study are shown in Figure 3.2.

68

OH OMe OMe OMe OH OH

HO O HO O HO O

MeO OH O OH O OH O OMe 44 51 66 OH OH OH

HO O MeO O HO O

MeO OH OH O OH OH O OH O OMe 69 72 97 MeO O HO O HO O

MeO MeO OH O OH O OH O 98 99 104

Figure 3.2. Compounds isolated from Eriodictyon californicum in the present investigation.

5. Characterization of isolated compounds

5.1. Notes on the characterization of isolated compounds

NMR assignments marked with an asterisk may be considered interchangeable.

This is particularly an issue with the carbons at the C-7 and C-9 positions of the isolated compounds as they are in very similar chemical environments. Also, the chemical shifts of at the C-6 and C-8 on the 5,7-dioxygenated compounds tended to nearly overlap, in both the 1H and 13C NMR spectra. The numbering system used for the isolated flavonoids is given in Figure 3.3, with the isobutyryl group of 4′-isobutyrylhomoeriodictyol (95) receiving a double prime (′′) designation with numbering starting at the ester carbon.

69

5' 6' 4' 8 B 7 9 O 2 1' 3' A C 2' 6 3 10 4 5 O

Figure 3.3. Numbering system employed for flavonoids isolated from Eriodictyon californicum.

5.2. Characterization of hesperetin (44)

25 Hesperetin (44). White needles, mp 226-227 ºC; [α] D -29 (c 1.0, MeOH); CD

(MeOH, Δε) 327.6 nm (2.9 degree•L•mol-1•m-1) 290.2 nm (-11.4 degree•L•mol-1•m-1); IR

(NaCl) υmax 3376, 1641, 1620, 1514, 1462, 1453, 1274, 1183, 1183, 1161, 1087, 1067

-1 1 cm ; H NMR (300 MHz, acetone-d6) δH 2.77 (1H, dd, J = 3.0, 17.1 Hz, H-3α), 3.17 (1H dd, J = 12.3, 17.1 Hz, H-3β), 3.88 (3H, s, OMe-4′), 5.46 (1H, dd, J = 12.9, 3.0 Hz, H-2),

5.96 (1H, d, J = 1.2 Hz, H-8*), 5.99 (1H, d, J = 1.2 Hz, H-6*), 7.00 (2H, m, H-5′, H-6′),

13 7.06 (1H, s, H-2′), 7.75 (1H, s, OH), 9.17 (1H, s, OH), 12.31 (1H, s, OH-5); C NMR δC

197.1 (C-4), 167.4 (C-7**), 165.3 (C-9**), 164.2 (C-5**), 148.7 (C-4′), 147.6 (C-3′),

132.8 (C-1′), 118.2 (C-6′), 114.4 (C-2′), 112.3 (C-5′), 103.2 (C-10), 96.8 (C-6***), 95.9

(C-8***), 79.8 (C-2), 56.3 (OMe-4′), 43.5 (C-3); HRESIMS m/z 309.0686 [M+Na]+

+ (calcd for C16H14O6Na , 309.0688). The purity of hesperetin (44) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane- ethyl acetate, Rf = 0.59; 6:1 dichloromethane-acetone, Rf = 0.64; 19:1 dichloromethane- methanol, Rf = 0.57).

70

5.3. Characterization of homoeriodictyol (51)

25 Homoeriodictyol (51). Light yellow needles, mp 224-226 ºC; [α] D +17 (c 1.0,

-1 -1 CHCl3); CD (MeOH, Δε) 327.2 nm (2.4 degree•L•mol •m ) 292.6 nm (-12.3

-1 -1 degree•L•mol •m ); IR (NaCl) υmax 3359, 1610, 1641, 1519, 1464, 1435, 1342, 1272,

-1 1 1184, 1160, 1087, 1067 cm ; H NMR (300 MHz, acetone-d6) δH 2.74 (1H, dd, J = 3.0,

17.1 Hz, H-3α), 3.21 (1H dd, J = 12.9, 17.1 Hz, H-3β), 3.89 (3H, s, OMe-3′), 5.43 (1H, dd, J = 12.9, 2.7 Hz, H-2), 5.98 (1H, s, H-6*), 5.98 (1H, s, H-8*), 6.86 (1H, m, H-5′),

13 6.98 (1H, m, H-6′), 7.16 (1H, d, J = 1.2 Hz, H-2′), 12.16 (1H, s, OH-5); C NMR δC

196.3 (C-4), 166.6 (C-7**), 164.4 (C-9**), 163.5 (C-5**), 147.6 (C-4′), 147.0 (C-3′),

130.4 (C-1′), 119.7 (C-6′), 114.9 (C-5′), 110.3 (C-2′), 102.3 (C-10), 96.0 (C-6***), 95.1

(C-8***), 79.3 (C-2), 55.5 (OMe-3′), 42.7 (C-3); HRESIMS m/z 309.0688 [M+Na]+

+ (calcd for C16H14O6Na 309.0688). The purity of homoeriodictyol (51) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane- ethyl acetate, Rf = 0.61; 6:1 dichloromethane-acetone, Rf = 0.57; 19:1 dichloromethane- methanol, Rf = 0.57).

5.4. Characterization of jaceosidin (66)

Jaceosidin (66). Dark yellow needles, mp 229-230 ºC; IR (NaCl) υmax 3354,

2921, 2851, 1658, 1613, 1514, 1495, 1463, 1430, 1361, 1272, 1164, 1107 cm-1; 1H NMR

(300 MHz, DMSO-d6) δH 3.75 (3H, s, OMe-6), 3.89 (3H, s, OMe-3′), 6.62 (1H, s, H-8),

6.90 (1H, s, H-3), 6.93 (1H, d, J = 9.0 Hz, H-5′), 7.56 (2H, m, H-2′, H-6′), 10.04 (1H, s,

13 OH-4′*), 10.66 (1H, s, OH-7*), 13.10 (1H, s, OH-5); C NMR (400 MHz, DMSO-d6) δC

71

182.2 (C-6), 163.7 (C-2), 157.4 (C-7**), 152.7 (C-9**), 152.4 (C-5**), 150.7 (C-4′),

148.0 (C-3′), 131.4 (C-6), 121.5 (C-1′), 120.3 (C-6′), 115.7 (C-5′), 110.1 (C-2′), 104.0 (C-

10), 102.7 (C-3), 94.3 (C-8), 59.9 (OMe-6), 56.0 (OMe-3′); HRESIMS m/z 353.0653

+ + [M+Na] (calcd for C17H14O7Na 353.0638). The purity of jaceosidin (66) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane- ethyl acetate, Rf = 0.42; 6:1 dichloromethane-acetone, Rf = 0.65; 19:1 dichloromethane- methanol, Rf = 0.55).

5.5. Characterization of naringenin (69)

25 Naringenin (69). White needles, mp 256-257 ºC (227-229 ºC ); [α] D -12 (c 1.0,

MeOH); CD (MeOH, Δε) 326.4 nm (2.0 degree•L•mol-1•m-1) 288.4 nm (-7.6

-1 -1 degree•L•mol •m ); IR (NaCl) υmax 3383, 2959, 2923, 1650, 1634, 1615, 1508, 1463,

-1 1 1435, 1341, 1246, 1162, 1081, 1065 cm ; H NMR (300 MHz, acetone-d6) δH 2.74 (1H, dd, J = 3.0, 17.1 Hz, H-3α), 3.24 (1H dd, J = 12.9, 17.1 Hz, H-3β), 5.46 (1H, dd, J =

12.9, 3.0 Hz, H-2), 5.97 (2H, m, H-6, H-8), 6.91 (2H, d, J = 8.7 Hz, H-3′, H-5′), 7.41

13 (2H, d, J = 8.4 Hz, H-2′, H-6′), 9.08 (2H, OH), 12.19 (1H, s, OH-5); C NMR δC 196.4

(C-4), 166.5 (C-7*), 164.4 (C-9*), 163.5 (C-5*), 157.8 (C-4′), 129.9 (C-1′), 128.1 (C-2′,

C-6′), 115.3 (C-3′, C-5′), 102.3 (C-10), 96.0 (C-6**), 95.0 (C-8**) 79.1 (C-2), 42.6 (C-

+ + 3); HRESIMS m/z 295.0582 [M+Na] (calcd for C15H12O3Na 295.0582). The purity of naringenin (69) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf = 0.64; 6:1 dichloromethane-acetone, Rf =

0.60; 19:1 dichloromethane-methanol, Rf = 0.42).

72

5.6. Characterization of sakuranetin (72)

25 Sakuranetin (72). White needles, mp 140-141 ºC; [α] D -18 (c 1.0, CHCl3); CD

(MeOH, Δε) 326.6 nm (2.1 degree•L•mol-1•m-1) 288.0 nm (-8.2 degree•L•mol-1•m-1); IR

-1 1 (NaCl) υmax 3383, 1637, 1519, 1441, 1377, 1334, 1305, 1205, 1191, 1156, 1089 cm ; H

NMR (300 MHz, acetone-d6) δH 2.74 (1H, dd, J = 3.0, 17.1 Hz, H-3α), 3.24 (1H dd, J =

12.6, 17.1 Hz, H-3β), 3.83 (3H, s, OMe) 5.46 (1H, dd, J = 3.0, 12.9 Hz, H-2), 6.04 (1H, d, J = 2.1 Hz, H-8*), 6.03 (1H, d, J = 2.1 Hz, H-6*), 6.89 (2H, d, J = 8.7 Hz, H-3′, 5′),

7.39 (2H, d, J = 8.4 Hz, H-2′,6′), 8.61 (1H, s, OH), 12.13 (1H, s, OH-5); 13C NMR (400

MHz, acetone-d6) δC 197.0 (C-4), 167.4 (C-7), 163.2 (C-5), 162.9 (C-7), 157.8 (C-4′),

128.7 (C-1′), 128.4 (C-2′, 6′), 115.2 (C-3′, 5′), 102.6 (C-10), 94.6 (C-6**), 93.8 (C-8**)

78.6 (C-2), 55.9 (OMe), 42.0 (C-3); HRESIMS m/z 309.0731 [M+Na]+ (calcd for

+ C16H14O5Na , 309.0739). The purity of sakuranetin (72) was verified by employing thin- layer chromatography (TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf =

0.66; 6:1 dichloromethane-acetone, Rf = 0.83; 19:1 dichloromethane-methanol, Rf =

0.76).

5.7. Characterization of 4′-isobutyrylhomoeriodictyol (95)

25 4′-Isobutyrylhomoeriodictyol (95). White needles, mp 149-150 ºC; [α] D -87 (c

1.0, MeOH); CD (MeOH, Δε) 326.8 nm (4.4 degree•L•mol-1•m-1) 288.0 nm (-20.5

-1 -1 degree•L•mol •m ); IR (NaCl) υmax 3382, 2973, 2934, 1758, 1738, 1642, 1609, 1512,

-1 1 1467, 1423, 1344, 1313, 1272, 1182, 1160, 1088 cm ; H NMR (300 MHz, acetone-d6)

δH 1.30 (6H, d, J = 7.2 Hz, H-3′′a, H-3′′b), 2.85 (2H, m, H-3α, H-2′′), 3.23 (1H, dd, J =

73

12.9, 17.1 Hz, H-3β), 3.88 (3H, s, OMe-4′), 5.58 (1H, dd, J = 12.9, 3.0 Hz, H-2), 5.98

(1H, d, J = 2.1 Hz, H-6), 6.02 (1H, d, J = 2.1 Hz, H-8), 7.16 (2H, m, H-5′, H-6′), 7.36

13 (1H, s, H-2′), 12.15 (1H, s, OH-5); C NMR δC 196.8 (C-4), 175.1 (C-1′′), 167.5 (C-

7**), 165.3 (C-5), 164.1 (C-9**), 152.4 (C-3′), 141.2 (C-4′), 138.7 (C-1′), 123.7 (C-5′),

119.4 (C-6′), 111.8 (C-2′), 103.1 (C-10), 96.8 (C-6), 95.9 (C-8), 79.7 (C-2), 56.3 (OMe-

4′), 43.6 (C-3), 34.5 (C-2′′), 19.3 (C-3′′a, C-3′′b); HRESIMS m/z 395.1090 [M+Na]+

+ (calcd for C20H20O7Na , 395.1107). The purity of 4′-isobutyrylhomoeriodictyol (95) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf = 0.73; 6:1 dichloromethane-acetone, Rf = 0.85; 19:1 dichloromethane-methanol, Rf = 0.78).

5.8. Characterization of 6-methoxyhomoeriodictyol (97)

25 6-Methoxyhomoeriodictyol (97). White needles, mp 161-162 ºC; [α] D -14 (c

1.0, MeOH); CD (MeOH, Δε) 326.6 nm (2.9 degree•L•mol-1•m-1) 284.2 nm (-10.5

-1 -1 degree•L•mol •m ); IR (NaCl) υmax 3383, 2925, 2851, 1644, 1506, 1462, 1436, 1342,

-1 1 1296, 1275, 1159, 1090, cm ; H NMR (400 MHz, acetone-d6) δ 2.74 (1H, dd, J = 2.8,

17.2 Hz, H-3α), 3.22 (1H dd, J = 13.0, 17.2 Hz, H-3β), 3.77 (3H, s, OMe-6), 3.88 (3H, s,

OMe-3′), 5.43 (1H, dd, J = 13.0, 2.8 Hz, H-2), 6.02 (1H, s, H-8), 6.87 (2H, d, J = 8.1 Hz,

H-5′), 6.99 (1H, d, J = 8.1 Hz, H-6′), 7.18 (1H, s, H-2′), 7.79 (1H, s, OH), 9.24 (1H, s,

OH), 12.32 (1H, s, OH-5); 13C NMR δ 198.1 (C-4), 159.9 (C-7*), 159.6 (C-9*), 156.3

(C-5), 148.4 (C-4′), 147.9 (C-3′), 131.3 (C-1′), 129.9 (C-6), 120.5 (C-6′), 115.7 (C-5′),

111.1 (C-2′), 103.3 (C-10), 95.7 (C-8), 80.2 (C-2), 60.7 (OMe-6), 56.3 (OMe-3′), 43.7

74

+ + (C-3); HRESIMS m/z 355.0797 [M+Na] (calcd for C17H16O7Na 355.0794). The purity of 6-methoxyhomoeriodictyol (97) was verified by employing thin-layer chromatography

(TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf = 0.51; 6:1 dichloromethane-acetone, Rf = 0.89; 19:1 dichloromethane-methanol, Rf = 0.68).

5.9. Characterization of 6-methoxysakuranetin (98)

6-Methoxysakuranetin (98). White needles, mp 155-156 ºC (lit. 173-174 ºC);36

25 -1 -1 [α] D -8 (c 1.0, MeOH); CD (MeOH, Δε) 335.8 nm (2.1 degree•L•mol •m ) 289.8 nm (-

-1 -1 8.0 degree•L•mol •m ); IR (NaCl) υmax 3418, 2922, 2850, 1642, 1574, 1499, 1454,

-1 1 1309, 1287, 1202, 1112, 1087 cm ; H NMR (300 MHz, acetone-d6) δH 2.75 (1H, d, J =

17.4 Hz, H-3α), 3.20 (1H dd, J = 13.2, 17.4 Hz, H-3β), 3.70 (3H, s, OMe-6), 3.89 (3H, s,

OMe-7), 5.46 (1H, d, J = 12.6 Hz, H-2), 6.18 (1H, s, H-8), 6.89 (2H, d, J = 8.7 Hz, H-3′,

H-5′), 7.39 (2H, d, J = 8.7 Hz, H-2′, H-6′), 8.58 (1H, s, OH-4′), 11.99 (1H, s, OH-5); 13C

NMR δC 198.3 (C-4), 162.0 (C-7*), 159.5 (C-9*), 158.7 (C-4′), 155.9 (C-5), 131.2 (C-6),

130.7 (C-1′), 129.0 (C-2′, C-6′), 116.2 (C-3′, C-5′), 103.7 (C-10), 92.6 (C-8), 80.2 (C-2),

60.5 (OMe-6), 56.6 (OMe-7), 43.5 (C-3); HRESIMS m/z 339.0825 [M+Na]+ (calcd for

+ C17H16O6Na , 339.0845). The purity of 6-methoxysakuranetin (98) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane- ethyl acetate, Rf = 0.51; 6:1 dichloromethane-acetone, Rf = 0.84; 19:1 dichloromethane- methanol, Rf = 0.69).

75

5.10. Characterization of pinocembrin (99)

25 Pinocembrin (99). White needles, mp 189-190 ºC; [α] D -45 (c 1.0, MeOH); CD

(MeOH, Δε) 326.6 nm (6.5 degree•L•mol-1•m-1) 284.2 nm (-25.9 degree•L•mol-1•m-1); IR

(NaCl); υmax 3355, 2956, 2925, 1640, 1465, 1455, 1342, 1306, 1275, 1183, 1161, 1088,

-1 1 1067 cm ; H NMR (300 MHz, acetone-d6) δH 2.80 (1H, dd, J = 3.0, 17.1 Hz, H-3α),

3.24 (1H dd, J = 12.6, 17.1 Hz, H-3β), 5.57 (1H, dd, J = 12.9, 3.0 Hz, H-2), 5.60 (1H, d, J

= 1.8 Hz, H-6), 5.99 (1H, d, J = 1.8 Hz, H-8), 7.47 (3H, m, H-3′,-4′,-5′), 7.56 (2H, d, J

13 = 6.9 Hz, H-2′,6′), 9.64 (1H, s, OH), 12.16 (1H, s, OH-5); C NMR δC 196.8 (C-4),

167.4 (C-7*), 165.3 (C-5), 164.2 (C-9*), 140.1 (C-1′), 129.5 (C-3′, -5′), 129.4 (C-4′),

127.3 (C-2′,-6′), 103.3 (C-10), 97.0 (C-6), 95.9 (C-8) 80.0 (C-2), 43.6 (C-3); HRESIMS

+ + m/z 279.0633 [M+Na] (calcd for C15H12O4Na , 279.0643). The purity of pinocembrin

(99) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf = 0.78; 6:1 dichloromethane-acetone, Rf = 0.81;

19:1 dichloromethane-methanol, Rf = 0.78).

5.11. Characterization of 6-methoxyhesperetin (104)

25 6-Methoxyhesperetin (104). White needles, mp 234-235 ºC; [α] D -11 (c 1.0,

MeOH); CD (MeOH, Δε) 334.6 nm (2.5 degree•L•mol-1•m-1) 293.0 nm (-9.7

-1 -1 degree•L•mol •m ); IR (NaCl) υmax 3374, 2926, 2843, 1643, 1586, 1503, 1461, 1342,

-1 1 1296, 1276, 1160, 1090 cm ; H NMR (300 MHz, acetone-d6) δH 2.76 (1H, dd, J = 3.0,

17.2 Hz, H-3α), 3.16 (1H dd, J = 12.6, 17.2 Hz, H-3β), 3.78 (3H, s, OMe-6), 3.86 (3H, s,

OMe-4′), 5.42 (1H, dd, J = 12.6, 3.0 Hz, H-2), 6.02 (1H, s, H-8), 6.98 (2H, m, H-5′, H-

76

6′), 7.04 (1H, s, H-2′), 7.75 (1H, s, OH), 9.17 (1H, s, OH), 12.31 (1H, s, OH-5); 13C NMR

δC 198.0 (C-4), 159.9 (C-7*), 159.5 (C-9*), 156.3 (C-5), 148.7 (C-4′), 147.6 (C-3′), 132.9

(C-1′), 129.9 (C-6), 118.8 (C-6′), 114.4 (C-2′), 112.3 (C-5′), 103.4 (C-10), 95.7 (C-8),

79.9 (C-2), 60.8 (OMe-6), 56.4 (OMe-4′), 43.6 (C-3); HRESIMS m/z 355.0794 [M+Na]+

+ (calcd for C17H16O7Na , 355.0794). The purity of 6-methoxyhesperetin (104) was verified by employing thin-layer chromatography (TLC) with three solvent systems (1:1 hexane-ethyl acetate, Rf = 0.51; 6:1 dichloromethane-acetone, Rf = 0.76; 19:1 dichloromethane-methanol, Rf = 0.68).

C. Discussion

1. Identification of hesperetin (44)

The molecular weight of hesperetin (44) was determined by high-resolution mass spectrometry with a sodiated molecular ion peak appearing at m/z 309.0686, consistent with the elemental formula, C16H14O6. The degree of unsaturation necessary to achieve this mass indicated that the compound is highly ringed with several double bonds. This was also supported by the 1H NMR and 13C NMR data obtained for this compound. The

1H NMR spectrum (Figure 3.4) of hesperetin (44) revealed several resonances in the aromatic region [δH 5.96 (1H, d, J = 1.2 Hz, H-8*), 5.99 (1H, d, J = 1.2 Hz, H-6*), 7.00

(2H, m, H-5′, H-6′), 7.06 (1H, s, H-2′)]. The magnitude of the coupling constants of the doublets observed indicated meta coupling. Two typical broad hydroxy signals were observed [δH 7.75 (1H, s, OH), 9.17 (1H, s, OH)] as well as a resonance at δH 12.31 (1H,

77 s, OH-5), which occurred as a sharp strong signal indicating an alcoholic proton chelated by a ketone. The 13C NMR spectrum of hesperetin (44) (Figure 3.5) showed several resonances in the aromatic region [δC 167.4 (C-7**), 165.3 (C-9**), 164.2 (C-5**), 148.7

(C-4′), 147.6 (C-3′), 132.8 (C-1′), 118.2 (C-6′), 114.4 (C-2′), 112.3 (C-5′), 103.2 (C-10),

96.8 (C-6***), 95.9 (C-8***)]. Several of these were in the typical region of oxygenated aromatic carbons. The carbonyl suggested by the 1H NMR spectrum was confirmed by

13 the C NMR resonance observed at δC 197.1 (C-4). Overall, the data were consistent with hesperetin (44) being of the flavanone class. A coupled doublet of doublets system, representing the protons at positions C-2 and C-3, resonated at δH 2.77 (1H, dd, J = 3.0,

17.1 Hz, H-3α), 3.17 (1H dd, J = 12.3, 17.1 Hz, H-3β), and 5.46 (1H, dd, J = 12.9, 3.0

Hz, H-2). The signals at δH 7.00 (2H, m, H-5′,-6′) and 7.06 (1H, s, H-2′) indicated the presence of three B-ring protons. From the observation of a singlet in the B-ring it could be deduced that C-3′ and C-4′ are substituted with the signal corresponding to the relatively isolated C-2′ position. The two similar meta-coupled resonances at 5.96 (1H, d,

J = 1.2 Hz, H-8*) and 5.99 (1H, d, J = 1.2 Hz, H-6*) corresponded to the C-6 and C-8 positions, allowing for a downfield C-5 hydroxy group. HMBC, HSQC and NOESY correlations (Figures 3.6, 3.7, and 3.8) were used to confirm the presented NMR assignments. The methoxy group position could be determined due to the NOESY correlation between the methoxy protons and a proton in the C-5′, C-6′ proton multiplet.

The stereocenter at C-2 was resolved utilizing circular dichroism spectroscopy (CD). A positive Cotton effect was observed at 327.6 nm and a negative value at 290.2 nm, which is typical for 2S flavanones (Figure 3.9) (Gaffield, 1970). The physical and spectroscopic

78 data obtained were consistent with the cited literature for hesperetin (44) (Arakawa and

Nakazaki, 1960; Chalia, et al., 1965; Vasconcelos et al., 1998; Heneczkowski et al.,

2001; Giorgioa et al., 2004; Lee et al., 2006).

1 Figure 3.4. H NMR spectrum of hesperetin (44) in acetone-d6

13 Figure 3.5. C NMR spectrum of hesperetin (44) in acetone-d6

79

Figure 3.6. HMBC correlation NMR spectrum of hesperetin (44)

Figure 3.7. HSQC correlation NMR spectrum of hesperetin (44)

80

Figure 3.8. NOESY correlation NMR spectrum of hesperetin (44)

40

30

20

10

0 250 260 270 280 290 300 310 320 330 340 350 -10

-20

-30 -40 Figure 3.9. CD spectrum of hesperetin (44)

81

2. Identification of homoeriodictyol (51)

The high-resolution mass spectrum of homoeriodictyol (51) (m/z 309.0688

+ 1 [M+Na] ) indicated the same molecular formula as hesperetin (44) (C16H14O6). The H

NMR and 13C NMR spectra of homoeriodictyol (51) (Figures 3.10 and 3.11) also indicated that the compound is very similar in structure to compound 44. The most striking differences in either the 1H NMR or 13C NMR spectrum of homoeriodictyol (51) relative to flavanone 44 were observed in the B-ring proton signals [δH 6.86 (1H, m, H-

5′), 6.98 (1H, m, H-6′), 7.16 (1H, d, J = 1.2 Hz, H-2′)]. The single meta-coupled proton

(C-2′) and the observed resonances indicated the same general substitution pattern present in the hesperetin (44) B-ring. The B-ring of compound 51 was therefore determined to have the opposite methoxylation pattern of that found in hesperetin (44).

The HSQC spectrum (Figure 3.12) revealed that, while the 13C NMR spectrum of homoeriodictyol (51) appeared to be closely comparable to that of hesperetin (44), in fact the C-2′ and C-5′ chemical shifts were transposed. HMBC correlations (Figure 3.13) were used to finalize the NMR assignments for compound 51. The stereocenter at C-2 was resolved utilizing circular dichroism spectroscopy (CD) (Figure 3.14). Like hesperetin, a positive Cotton effect was observed at 327.2 nm and a negative value at

292.6 nm, so this compound could again be assigned as a 2S flavanone (Gaffield, 1970).

The physical and spectroscopic data obtained for homoeriodictyol (51) were consistent with the cited literature (Tatuta, 1940; Gaffield, 1970; Liu et al., 1992; Ibrahim et al.,

2003).

82

1 Figure 3.10. H NMR spectrum of homoeriodictyol (51) in acetone-d6

13 Figure 3.11. C NMR spectrum of homoeriodictyol (51) in acetone-d6

83

Figure 3.12. HSQC correlation spectrum of homoeriodictyol (51)

Figure 3.13. HMBC correlation spectrum of homoeriodictyol (51) 84

15

10

5

0 250 260 270 280 290 300 310 320 330 340 350 -5

-10

-15 Figure 3.14. CD spectrum of homoeriodictyol (51)

3. Characterization of jaceosidin (66)

The high-resolution mass spectrum of jaceosidin (66) revealed the molecular formula to be C17H14O7 (m/z 353.0653 [M+Na]). This is consistent with a structure similar to hesperetin (44) and homoeriodictyol (51), but incorporating an additional methoxy group and an additional level of unsaturation. Accordingly, the 1H NMR and

13C NMR spectra of jaceosidin (66) (Figures 3.15 and 3.16) did not show the typical flavanone resonances at the C-2 and C-3 positions. However, there were still typical A- and B-ring signals in 1H NMR spectrum, indicating the compound to be a flavone. The signals at δH 6.93 (1H, d, J = 9.0 Hz, H-5′) and 7.56 (2H, m, H-2′, H-6′) revealed three

13 protons situated in the B-ring. The observed C NMR B-ring chemical shifts [δC 150.7

(C-4′), 148.0 (C-3′), 121.5 (C-1′), 120.3 (C-6′), 115.7 (C-5′), 110.1 (C-2′)] indicated that the B-ring is similar to those of hesperetin (44) and homoeriodictyol (51). The observed coupling constant of 9.0 Hz resonating at δH 6.93 indicated an ortho coupling. This confirmed a C-3′, C-4′ disubstitution, with a well-defined coupling constant (J = 9.0 Hz) assigned to the H-5′ proton. The coupling constants for the other two protons could not

85 be determined with accuracy due to the overlap of the resonances. The signal at δH 6.62

(1H, s, H-8) indicated the presence of only one A-ring proton. Three hydroxy group

1 signals were also seen in the H NMR spectrum [δH 10.04 (1H, s, OH-4′*), 10.66 (1H, s,

OH-7*), 13.10 (1H, s, OH-5)], with the latter, sharp peak indicating that the C-5 position is hydroxylated. The 13C NMR spectrum of jaceosidin (66) confirmed the presence of a flavone carbon skeleton for this molecule, with a characteristic C-ring carbonyl resonance at δC 182.2 (C-4) and olefinic signals δC 163.7 (C-2), and 102.7 (C-3). This spectrum also showed two methoxy group resonances at δC 59.9 (OMe-6) and 56.0 (OMe-7).

HMBC, HSQC, and NOESY correlations (Figures 3.17, 3.18, and 3.19) were used to finalize the NMR assignments, and determine the methoxy group positions. The spectroscopic and physical data were determined to be consistent with the literature values recorded for jaceosidin (66) (Wollenweber and Mann, 1989; Nakasugi et al.,

2000; Min et al., 2009).

1 Figure 3.15. H NMR spectrum of jaceosidin (66) in DMSO-d6 86

13 Figure 3.16. C NMR spectrum of jaceosidin (66) in DMSO-d6

Figure 3.17. HMBC correlation NMR spectrum of jaceosidin (66)

87

Figure 3.18. HSQC NMR correlation spectrum of jaceosidin (66)

Figure 3.19. NOESY NMR correlation spectrum of jaceosidin (66)

88

4. Identification of naringenin (69)

High-resolution mass spectrometry of naringenin (69) revealed the molecular

+ formula C15H12O3 (m/z 295.0582 [M+Na] ). As the compound was isolated from the same subfraction as hesperetin (44) and homoeriodictyol (51), it was considered likely to be an additional trihydroxylated flavanone. The 1H NMR spectrum of naringenin (69)

(Figure 3.20) displayed a coupled doublet of doublets system, representing H-2 and H-3 protons, which were observed at δH 2.74 (1H, dd, J = 3.0, 17.1 Hz, H-3α), 3.24 (1H dd, J

= 12.9, 17.1 Hz, H-3β), and 5.46 (1H, dd, J = 12.9, 3.0 Hz, H-2). The B-ring of naringenin (69) was found to be monosubstituted at the para position, as indicated by the signals at δH 6.91 (2H, d, J = 8.7, H-3′, H-5′) and 7.41 (2H, d, J = 8.4, H-2′, H-6′).

Additionally, two overlapping signals at δH 5.97 (2H, m, H-6, H-8) could be assigned to the C-6 and C-8 positions, which are in very similar environments, due to the oxygenation evident at the C-5, C-7, and C-9 positions. The assignment of three hydroxy

1 groups from the elemental formula was supported by the H NMR spectrum [δH 9.08

(2H, OH), 12.19 (1H, s, OH-5)]. The 13C NMR spectrum of naringenin (69) (Figure

3.21) confirmed the presence of a flavanone carbon skeleton for this molecule, with characteristic C-ring resonances at δC 196.4 (C-4), 79.1 (C-2), and 42.6 (C-3). HMBC and HSQC correlations (Figures 3.22 and 3.23) were used to verify the overall NMR assignments. Circular dichroism (CD) spectroscopy (Figure 3.24) was employed to determine the stereochemistry as 2S due to the Cotton effects observed at 326.4 nm and

288.4 nm (Gaffield, 1970). This compound exhibited similar spectroscopic data to the

89 literature values for naringenin (69) (Giorgioa et al., 2004; Selenski and Pettus, 2006;

Jeon et al., 2008).

1 Figure 3.20. H NMR spectrum of naringenin (69) in acetone-d6

13 Figure 3.21. C NMR spectrum of naringenin (69) in acetone-d6 90

Figure 3.22. HSQC correlation NMR spectrum of naringenin (69)

Figure 3.23. HMBC correlation NMR spectrum of naringenin (69) 91

30

20

10

0 250 270 290 310 330 350 370 -10

-20

-30 Figure 3.24. CD spectrum of naringenin (69)

5. Identification of sakuranetin (72)

A determination of the molecular formula of sakuranetin (72) by high-resolution mass spectrometry revealed this to be C16H14O5, displaying a sodiated molecular ion peak at m/z 309.0731 [M+Na]+. This elemental formula along with the known predominance of flavonoid secondary metabolites from E. californicum strongly suggested the presence of a methoxylated flavanone. The 1H NMR spectrum of sakuranetin (72) (Figure 3.25) was found to be very similar to that of naringenin (69). Like naringenin (69), the signals at δH 6.89 (2H, d, J = 8.7 Hz, H-3′, 5′) and 7.39 (2H, d, J = 8.4 Hz, H-2′,6′) revealed four protons on the B-ring with a line of symmetry, indicating monosubstitution at the C-4′ position. Likewise, the signals at δH 6.04 (1H, d, J = 2.1 Hz, H-8), and 6.03 (1H, d, J =

2.1 Hz, H-6) were used to infer that there are two A-ring protons present in very similar

1 positions. Two hydroxy group signals were also seen in the H NMR spectrum [δH 8.61

(1H, s, OH-4′), 12.13 (1H, s, OH-5)], with a sharp downfield singlet indicative of a hydroxy substituent at the C-5 position. The 13C NMR spectrum of sakuranetin (72)

(Figure 3.26) confirmed the presence of a flavanone carbon skeleton for this molecule,

92 with characteristic C-ring resonances at δC 197.0 (C-4), 78.6 (C-2), and 42.0 (C-3). This spectrum also showed one methoxy group signal at δC 55.9 (OMe). HMBC and HSQC correlations (Figures 3.27 and 3.28) were used to finalize the NMR assignments. HMBC correlations from the C-5 hydroxy signal to the C-5, and C-6 resonances allowed for their assignments. The methoxy group position could be determined due to the HMBC correlation of the methoxy protons to the C-7 carbon, which, in turn, was determined due to the HMBC correlations between it and the protons at the C-6 and C-8 positions. As with the other flavanones isolated in this study, CD spectroscopy (Figure 3.29) was used to establish the absolute configuration, with the Cotton effects at 326.6 nm and 288.0 nm revealing a 2S configuration (Gaffield, 1970). The assignment of compound 72 as sakuranetin was finalized due to its physical and spectroscopic data matching the literature values for this compound (Agrawal, 1989; Ichino et al., 1998; Vasconcelos et al., 1998; Jerz et al., 2005).

1 Figure 3.25. H NMR spectrum of sakuranetin (72) in acetone-d6 93

13 Figure 3.26. C NMR spectrum of sakuranetin (72) in CDCl3

Figure 3.27. HMBC spectrum of sakuranetin (72)

94

Figure 3.28. HSQC spectrum of sakuranetin (72)

40

30

20

10

0 250 260 270 280 290 300 310 320 330 340 350 -10

-20 -30 Figure 3.29. CD spectrum of sakuranetin (72)

6. Identification of 4′-isobutyrylhomoeriodictyol (95)

The determination of the molecular formula for compound 95 as C20H20O7 was carried out by high-resolution mass spectrometry (m/z 395.1090 [M+Na]+). The addition of five carbons to the typical flavonoid skeleton of fifteen carbons, indicated that the compound is somewhat atypical. Like the previously discussed compounds the 1H NMR

95 spectrum of 4′-isobutyrylhomoeriodictyol (95) suggested that the compound is of the flavanone class (Figure 3.30). A coupled doublet of doublets system, representing the protons affixed to positions C-2 and C-3, was observed at δH 2.85 (2H, m, H-3α, H-2′′),

3.23 (1H, dd, J = 12.9, 17.1 Hz, H-3β), and 5.98 (1H, d, J = 2.1 Hz, H-6), respectively.

As the mass spectrum revealed a level of unsaturation beyond that present in the basic flavonoid skeleton, either an additional ring or an additional double bond was determined to be present. The 1H NMR spectrum also displayed unusual resonances, such as additional peaks at δH 2.85 and the resonance at δH 1.30 (6H, d, J = 7.2 Hz, H-3′′a, H-

3′′b). The presence of a geminal dimethyl group coupled to a proton, and the presence of a methoxy substituent [δH 3.88 (3H, s, OMe-4′)], showed that the additional unsaturation must be due to a double bond. The signals for the B-ring protons at δH 7.16 (2H, m, H-5′,

6′) and 7.36 (1H, s, H-2′) revealed three protons at this site. However, these chemical shifts occurred downfield relative the previously discussed flavanones, indicating this ring to be unusually substituted. One hydroxy group signal was detected in the 1H NMR spectrum [δH 12.15 (1H, s, OH-5)] with a sharp downfield singlet indicative of a hydroxy substituent at the C-5 position. The 13C NMR spectrum of 4′-isobutyrylhomoeriodictyol

(95) (Figure 3.31) also exhibited unusual resonances at δC 175.1 (C-1′′), 34.5 (C-2′′), and

19.3 (C-3′′a, -3′′b), demonstrating an ester linkage and further validating evidence for the presence of a gem-dimethyl group. The HSQC spectrum (Figure 3.32) explained the resonances detected at δH 2.85 by revealing two protons attached to different carbons, namely, δC 43.6 (C-3) and 34.5 (C-2′′). Both the NOESY (Figure 3.33) and HMBC

(Figure 3.34) spectra revealed that the unusual peaks in the spectrum are due to an

96 isobutyryl substituent. HMBC correlations were used to indicate the position of a methoxy group present and allowed for the unambiguous assignment of the C-5 and C-6 carbons. The methoxy group position could be determined due by the HMBC correlation of these protons to the C-7 carbon, which was determined due to the correlations between it and the protons at the C-6 and C-8 positions. By displaying a positive Cotton effect at

326.8 nm and a negative Cotton effect at 288.0 nm in the CD spectrum (Figure 3.35), the previously unidentified absolute configuration of 4′-isobutyrylhomoeriodictyol (95) was determined to be 2S. The spectroscopic and physical data obtained were consistent with the only other report of this compound by Liu et al. (Liu et al., 1992).

1 Figure 3.30. H NMR spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6

97

13 Figure 3.31. C NMR spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6

Figure 3.32. HSQC correlation spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6

98

Figure 3.33. NOESY correlation spectrum of 4′-isobutyrylhomoeriodictyol (95) in acetone-d6

Figure 3.34. HMBC correlation spectrum of isobutyrylhomoeriodictyol (95) in acetone- d6

99

60

40

20

0 250 260 270 280 290 300 310 320 330 340 350 -20

-40

-60

Figure 3.35. CD spectrum of 4′-isobutyrylhomoeriodictyol (95)

7. Identification of 6-methoxyhomoeriodictyol (97)

A molecular formula of C17H16O7 was determined for 6- methoxyheomoeriodictyol, as obtained by high-resolution mass spectrometry (m/z

355.0797 [M+Na]+). Taken together with its 1H NMR spectrum (Figure 3.36), it could be determined that this compound is a dimethoxylated, trihydroxylated flavanone. The 1H

NMR spectrum showed resonances typical for protons at positions C-2 and C-3, which were observed at δH 2.74 (1H, dd, J = 2.8, 17.2 Hz, H-3α), 3.22 (1H dd, J = 13.0, 17.2

Hz, H-3β), and 5.43 (1H, dd, J = 13.0, 2.8 Hz, H-2). The signal at δH 6.02 (1H, s, H-8) was indicative of single proton substituent in the A-ring, consistent with the extensive oxygenation of this ring. The signals at δH 6.87 (2H, d, 8.1 Hz, H-5′), 6.99 (1H, d, 8.1

Hz, H-6′), and 7.18 (1H, s, H-2′) revealed three unsubstituted positions on the B-ring. The

B-ring proton splitting pattern and the chemical shift of the protons were found to be closely comparable to that of homoeriodictyol (51), indicating the same B-ring substitution pattern. The 13C NMR spectrum of 6-methoxyhomoeriodictyol (97) (Figure

3.37) confirmed this similarity to homoeriodictyol. HMBC and HSQC correlations 100

(Figures 3.38 and 3.39) were employed to finalize the assignments. HMBC correlations were used to place one of the methoxy groups at C-6 (δC 60.7) and the other on the B-ring

(δC 56.3). HMBC correlations from the hydroxy proton at C-5 were employed to verify the assignments for both C-5 and C-6. The CD spectrum (Figure 3.40) of 6- methoxyhomoeriodictyol (97) revealed the substance to be of the 2S configuration with optima at 326.6 nm (2.9 degree•L•mol-1•m-1) and 284.2 nm (-10.5 degree•L•mol-1•m-1)

(Gaffield, 1970). The absolute configuration of this compound has not been assigned previously. Through comparison with cited work, the structural determination of compound 97 as 6-methoxyhomoeriodictyol was confirmed (Li et al., 1988; Liu et al.,

1992).

1 Figure 3.36. H NMR spectrum of 6-methoxyhomoeriodictyol (97) in acetone-d6 101

13 Figure 3.37. C NMR spectrum of 6-methoxyhomoeriodictyol (97) in acetone-d6

Figure 3.38. HMBC correlation NMR spectrum of 6-methoxyhomoeriodictyol (97) 102

Figure 3.39. HSQC correlation NMR spectrum of 6-methoxyhomoeriodictyol (97)

50

40

30

20

10

0 250 260 270 280 290 300 310 320 330 340 350 -10

-20 -30 Figure 3.40. CD spectrum of 6-methoxyhomoeriodictyol (97)

8. Identification of 6-methoxysakuranetin (98)

The high-resolution mass spectrum of 6-methoxysakuranetin (98) was consistent

+ with the formula, C17H16O6 (m/z 339.0825 [M+Na] ). Like that for 6- methoxyhomoeridictyol (97), the 1H NMR spectrum of 6-methoxysakuranetin (98)

(Figure 3.41) revealed the presence of one A-ring proton. As evident for sakuranetin (72), 103 the B-ring was shown to be para-monosubstituted from the 1H NMR spectrum, with ortho-coupled resonances at δH 6.89 (2H, d, J = 8.7 Hz, H-3′, H-5′), and 7.39 (2H, d, J =

8.7 Hz, H-2′, H-6′). Two hydroxy group resonances were also detected in the 1H NMR spectrum [δH 8.58 (1H, s, OH-4′), 11.99 (1H, s, OH-5)] with the C-5 position hydroxylated as in the previously discussed flavanones. The 13C NMR spectrum of 6- methoxysakuranetin (98) (Figure 3.42) verified the similarity of this compound to sakuranetin (72) and 6-methoxyhomoeridictyol (97). HMBC and HSQC correlations

(Figures 3.43 and 3.44) were used to finalize the NMR assignments, and determine the methoxy groups as A-ring substituents. The CD spectrum of 6-methoxysakuranetin (98)

(Figure 3.45) was consistent with a 2S configuration, as in the other flavonoids isolated

[335.8 nm (2.1 degree•L•mol-1•m-1) 289.8 nm (-8.0 degree•L•mol-1•m-1)] (Gaffield,

1970). The measured physical and spectroscopic data were in accord with those by other groups for 6-methoxysakuranetin (98) (Uchida et al., 1980; Fernandez et al., 1988; Liu et al., 1992).

1 Figure 3.41. H NMR spectrum of 6-methoxysakuranetin (98) in acetone-d6 104

13 Figure 3.42. C NMR spectrum of 6-methoxysakuranetin (98) in acetone-d6

Figure 3.43. HMBC correlation NMR spectrum of 6-methoxysakuranetin (98) 105

Figure 3.44. HSQC correlation NMR spectrum of 6-methoxysakuranetin (98)

30 25 20 15 10 5 0 -5250 270 290 310 330 350 370 -10 -15 -20 -25 Figure 3.45. CD spectrum of 6-methoxysakuranetin (98)

9. Identification of pinocembrin (99)

The high-resolution mass spectrum of pinocembrin (99) revealed a molecular

+ 1 formula of C15H12O4 (m/z 279.0633 [M+Na] ). As the H NMR spectrum of pinocembrin

(99) (Figure 3.46) indicated that the compound is a flavanone [δH 2.80 (1H, dd, J = 3.0,

106

17.1 Hz, H-3α), 3.24 (1H dd, J = 12.6, 17.1 Hz, H-3β), and 5.57 (1H, dd, J = 12.9, 3.0

Hz, H-2)], it could be determined that the molecule is based on flavanone skeleton with two hydroxy group substituents. The signals at δH 7.47 (3H, m, H-3′, H-4′, H-5′), and

7.56 (2H, d, J = 6.9 Hz, H-2′, H-6′) revealed five protons on the B-ring, showing this ring to be otherwise unsubstituted. The signals at δH 7.56 (2H, d, J = 6.9 Hz, H-2′, H-6′) indicated that there are two A-ring protons in very similar positions, as discussed previously, thereby demonstrating a C-5, C-7 dioxygenation pattern. Two hydroxy group

1 signals were also seen in the H NMR spectrum [δH 9.64 (1H, s, OH), 12.16 (1H, s, OH-

5)] further validating a dihydroxylated flavone structure. These assignments were corroborated by the 13C NMR spectrum of compound 99 (Figure 3.47). HMBC and

HSQC correlations (Figures 3.48 and 3.49) were used to finalize the NMR assignments, especially in the A-ring positions due to correlations of the C-5 hydroxy proton. Circular dichroism (Figure 3.50) spectroscopy [326.6 nm (6.5 degree•L•mol-1•m-1) 284.2 nm (-

25.9 degree•L•mol-1•m-1)] revealed the compound as having the 2S configuration

(Gaffield, 1970). The obtained physical and spectroscopic data were consistent with literature values for pinocembrin (99) (Gaffield, 1970; Jung et al., 1990; Ichino et al.,

1998; Melliou and Chinou, 2004).

107

1 Figure 3.46. H NMR spectrum of pinocembrin (99) in acetone-d6

13 Figure 3.47. C NMR spectrum of pinocembrin (99) in acetone-d6

108

Figure 3.48. HMBC NMR spectrum of pinocembrin (99)

Figure 3.49. HSQC NMR spectrum of pinocembrin (99)

109

50

40

30

20

10

0 250 260 270 280 290 300 310 320 330 340 350 -10

-20

-30 Figure 3.50. CD spectrum of pinocembrin (99)

10. Structure elucidation of 6-methoxyhesperetin (104)

Although 6-methoxyhesperetin (104) has been proposed as a constituent of

Populus balsamifera L. (Salicaceae) and Iris tectorum Maxim. (Iridaceae) previously, it was reported as a planar structure without supporting spectroscopic data provided

(Seitembetova, 2000, Shu et al., 2010). While the molecular weight of this compound was found to be the same as 6-methoxyhomoeriodictyol (97) [HRESIMS m/z 355.0794

+ + [M+Na] (calcd for C17H16O7Na 355.0794)], there were differences in the NMR data measured. 6-Methoxyhesperetin (104) exhibited the characteristic peaks for a flavanone in its 1H NMR spectrum (Figure 3.51), with a coupled doublet of doublets system, representing protons at positions C-2 and C-3, observed at δH 2.76 (1H, dd, J = 3.0, 17.2

Hz, H-3α), 3.16 (1H dd, J = 12.6, 17.2 Hz, H-3β), and 5.42 (1H, dd, J = 12.6, 3.0 Hz, H-

1 2). The H NMR spectrum also revealed the presence of two methoxy substituents at δH

3.78 (3H, s) and 3.86 (3H, s). The signal at δH 6.02 (1H, s, H-8) was indicative of single proton on the A-ring, constistent with extensive oxygenation of this ring. The signals at

δH 6.98 (2H, m, H-5′,-6′) and 7.04 (1H, s, H-2′) revealed three unsubstituted positions on

110

1 the B-ring. Three hydroxy group signals were also seen in the H NMR spectrum [δH 7.75

(1H,), 9.17 (1H), 12.31 (1H, s, OH-5)] with a sharp downfield singlet indicative of a hydroxy substituent at the C-5 position. A B-ring singlet in the 1H NMR spectrum was assigned to the H-2′ proton, while the multiplet at δH 6.98 could be attributed to splitting between the nearly overlapping peaks of H-5′ and H-6′. The B-ring proton splitting pattern and the chemical shift of the protons were found to be closely comparable to those of hesperetin (44), indicating the same B-ring substitution pattern. This is of importance since the A and B-rings of flavanones are known to generally be independent of one another in their NMR spectra (Agrawal, 1989). The 13C NMR spectrum of compound 104 (Figure 3.52) confirmed the presence of a flavanone skeleton for this molecule, with characteristic C-ring resonances at δC 198.0 (C-4), 79.9 (C-2), and 43.6

(C-3). HMBC and HSQC correlations (Figures 3.53 and 3.54) were used to place one of the methoxy groups at C-6 (δC 60.8) and the other on the B ring (δC 56.4). The position of the B-ring methoxy group was determined by an NOE correlation to the H-5′, H-6′ signals (Figure 3.55). These data, in conjunction with the similarity of the proton spectrum to hesperetin (44), confirmed the structure as 6-methoxyhesperetin (104)

(Figure 3.56). Key 2D-NMR correlations for compound 104 are shown in Figure 3.57.

The absolute configuration was determined to be 2S via CD [334.6 nm (2.5 degree•L•mol-1•m-1) 293.0 nm (-9.7 degree•L•mol-1•m-1)] (Figure 3.58).

111

1 Figure 3.51. H NMR spectrum of 6-methoxyhesperetin (104) in acetone-d6

13 Figure 3.52. C NMR spectrum of 6-methoxyhesperetin (104) in acetone-d6

112

Figure 3.53. HSQC NMR spectrum of 6-methoxyhesperetin (104)

Figure 3.54. HMBC NMR spectrum of 6-methoxyhesperetin (104)

113

Figure 3.55. 1H-1H NOESY NMR spectrum of 6-methoxyhesperetin (104)

Figure 3.56. B-ring 1H NMR signals of 6-methoxyhesperetin (104) (left) and hesperetin (44) (right)

OCH H 3 HO O OH H H3CO OH O Figure 3.57. Selected correlations from HMBC (single headed arrows) and NOESY (double headed arrows) for 6 methoxyhesperetin (104) 114

20 15 10 5 0 -5250 270 290 310 330 350 370 -10 -15 -20 -25 -30 Figure 3.58. CD spectrum of 6-methoxyhesperetin (104)

11. Discussion of the potential sweetness-enhancing activity of compounds evaluated in vitro

Four of the flavonoids isolated from E. californicum, hesperetin (44), homoeriodictyol (51), naringenin (69), and sakuranetin (72), were tested in a cell-based sweetness-enhancing bioassay at a concentration of 25 μM. (The bioassay employed, as carried out by our collaborators at Givaudan Flavors Corporation, Cincinnati, OH is described in Appendix A). Additionally, a naturally occurring chalcone (105) (Figure

3.59), previously isolated from Merrillia caloxylon Swingle (Rutaceae) (Fraser and

Lewis, 1974), was purchased from Indofine Chemical Company, Inc. (Hillsborough, NJ) and tested in the chimeric sweetness receptor bioassay.

OH MeO OMe OMe

OH O

Figure 3.59. Structure of a chalcone (105) evaluated in the sweetness-enhancing assay.

115

Compounds with activity corresponding to at least 50% of the activity of cyclamate (15), the positive control, were considered active (Table 3.1). By comparing the activity of homoeriodictyol (51) to hesperetin (44), it appears that an isovanillic substitution pattern of the B-ring is favored over the vanillic substitution pattern by the sweetness receptor. Also the C-7-methoxylated compound sakuranetin (72) was found to be active in the cell-based system. The non-methoxylated analogue of sakuranetin, naringenin (69), was found to be inactive in the bioassay employed, indicating that the methoxy group at the C-7 position is necessary for the activity of sakuranetin (72). A chalcone (105) tested in this study, although representative of a different class of flavonoid contains the isovanillic substitution pattern of hesperetin (44), as well as a methoxy group that can rotate to a position analogous to the C-7 position like that found in sakuranetin (72). Indeed, this chalcone (105) was determined to be the most active of the compounds tested in this study. The activity of chalcone 105 encourages the testing of flavonoids of other classes with substitution patterns analogous to the active compounds demonstrated.

Compound Name Percent Activity of Control Hesperetin (44) 59.7 Homoeriodictyol (51) -0.8 Naringenin (69) 0.6 Sakuranetin (72) 53.9 3,2'-Dihydroxy-4,4',6'-trimethoxychalcone (105) 68.4 Cyclamate (15) 100 Table 3.1 Evaluation of compounds from E. californicum in an in vitro chimeric sweetness-enhancing assay. 116

It must be noted that while these results are promising, final evaluation by screening in a human sensory panel is optimal. While solubility issues limited the ability to test these compounds in vivo, hesperetin (44) is a known sweetness enhancer in man

(Ley et al., 2008c). These data lend credibility to the in vitro results obtained in the chimeric sweetness-enhancing screen as an accurate predictor of in vivo activity.

12. Discussion of the bitterness-masking activity of compounds evaluated in vitro

The bitterness receptor employed in the cell-based assay used for evaluating bitterness-masking activities was hTAS2R31, the receptor responsible for detecting the bitterness of several tastants including saccharin (14). (Further details for the bitterness- masking bioassay carried out by our collaborators at Givaudan Flavors Corporation,

Cincinnati, OH are given in Appendix A). Of the ten flavonoids from the leaves of E. californicum evaluated against hTAS2R31, three [jaceosidin (66), sakuranetin (72), 6- methoxysakuranetin (98)] were shown at 25 μM to decrease the receptor response to saccharin (14) by more than 50% (Table 3.2). The IC50 values (Table 3.2) for these compounds were roughly equivalent with sakuranetin being the most active antagonist in this study. The activity of sakuranetin (72) and 6-methoxysakuranetin (98) revealed that the C-7 position may play an important role in antagonistic activity at the hTAS2R31 receptor. The hTAS2R31 data also imply that methoxylation at the C-6 position does not greatly affect activity at the receptor, as evidenced by the values obtained for the previously mentioned compounds sakuranetin (72) and 6-methoxysakuranetin (98), which were both found to be selective receptor antagonists. This insight regarding the C-

117

6 position is supported also by comparison of the activities of hesperetin (44) and homoeriodictyol (51) and their 6-methoxy analogues, 6-methoxyhomoeriodictyol (97) and 6-methoxyhesperetin (104), which were not found to suppress the response of hTAS2R31 to saccharin (14) significantly at a concentration level of 25 μM.

Compound Name Percent Inhibition of IC50 (μM) at the the Response to Receptor hTAS2R31 Saccharin Hesperetin (44) -8.1±12.2 - Homoeriodictyol (51) 8.2±13.8 - Jaceosidin (66) 59.9±33.6 11.7±6.0 Naringenin (69) -5.7±7.1 - Sakuranetin (72) 82.4±23.4 5.5±2.5 4′-Isobutyrylhomoeriodictyol (95) -55.0±31.7 - 6-Methoxyhomoeriodictyol (97) 12.5±14.0 - 6-Methoxysakuranetin (98) 69.5±33.1 10.2±5.4 Pinocembrin (99) 25.5±24.3 - 6-Methoxyherperetin (104) 11.0±11.9 - Table 3.2. Evaluation of compounds from E. californicum in an in vitro cell-based bitterness-masking assay.

The simple flavanone, pinocembrin (99), was determined to be slightly active. In light of this activity, the C-7 methoxlyated derivative of this compound (the naturally occurring pinostrombin) may be an intriguing lead for future bitterness-masking tests.

Another inference that can be made from these data is that flavones may be more active than flavanones, as evidenced by the bitterness-masking activity of jaceosidin (66) and the inactivity of 6-methoxyhomoeriodictyol (97). Although only one flavone was tested in the present investigation, it is possible that its functional group substitution 118 pattern plays a different role in flavones when compared to flavanones in regard to bitterness-masking applications. It is known, however, that flavanones and flavones possessing the same substitution pattern can display quite different taste properties

(Benavente-García et al., 1997).

The high negative value associated with the activity of 4′- isobutyrylhomoeriodictyol (95) should also be mentioned. This compound appears to be a promiscuous GPCR activator, and, by activating GPCRs in the cells used to host hTAS2R31, displays false agonistic activity for the receptor.

As was noted in the previous section on sweetness-enhancing flavonoids from E. californicum, the optimal evaluation of these compounds would require a follow-up human sensory study, employing a panel of trained volunteer human subjects. However, once again, solubility concerns proved to be a limiting factor towards sensory evaluation of these flavonoid derivatives.

D. Conclusions

Investigation of Eriodictyon californicum for bitterness-masking and sweetness- enhancing constituents in this dissertation work has led to the isolation and characterization of nine known flavonoids (44, 51, 66, 69, 72, 95, 97, 98, 99) and one flavonoid (104) having a structure that had been proposed previously with no supportive spectroscopic or physical data provided. With the exception of 104, all of the compounds isolated in this work have been obtained previously as constituents of E. californicum. Of

119 the previously defined compounds, two had been published without determination of the

2S absolute configuration, 4′-isobutyrylhomoeriodictyol (95) and 6- methoxyhomoeriodictyol (97). In fact, the only report on flavanones from E. californicum attempting to assign the absolute configuration at C-2 (Ley et al. 2005) determined that the compounds generated occur as racemates. This was most likely due to the use of base during the purification stages allowing for racemization to occur, in the manner proposed in Figure 2.8. Although naturally occurring flavanones tend to be of the 2S configuration, there are examples of natural 2R flavanones, inclusive of sakuranetin (72) (Gaffield, 1970). Also in the report by Ley et al. (2005), the sodium salt of a phenol was reported as an isolate from E. californicum, a result that is likely to be caused by the extraction method used, rather than being a reflection of the biochemistry of the plant. It should also be mentioned that the presence of sodium in the homoeriodictyol salt generated plays a role in its flavor modulation profile (Ley et al.,

2005). Of the flavonoids isolated in this dissertation research, two of the four substances tested for in vitro sweetness-enhancing activity [hesperetin (44) and sakuranetin (72)] and three of the ten compounds evaluated for in vitro bitterness-masking activity [jaceosidin

(66), sakuranetin (72), and 6-methoxysakuranetin (98)] were considered active. These results are most interesting in light of the fact that they were not obtained with the benefit of bioactivity-guided fractionation, since the bioassay methodology used had not been developed at the time this dissertation work commenced. This research confirms E. californicum as being an important source of lead compounds for taste modulatory research.

120

The present body of work affirms the feasibility of reviewing the historical phytochemical literature for leads in the field of flavor research, since this was the impetus for the results presented herein. Additionally, the benefit of employing cell-based techniques in the field of flavor research can be readily seen, although the compounds isolated in this study could not be evaluated in follow-up sensory studies due to either the small amounts isolated or solubility limitations. Although E. californicum has been investigated as a source of flavor modulatory compounds by Ley et al., only five flavonoids isolated from this source (as racemates) were evaluated for flavor modulatory activity. While hesperetin (44) and homoeriodictyol (51) were evaluated previously for sweetness-enhancing and bitterness-masking activities (Ley et al., 2005; 2008a), and naringenin (69) for bitterness-masking activities (Ley et al., 2005), the other compounds isolated in the present study have not been evaluated before in any flavor modulation testing procedures.

This study of the flavor modulatory properties of the secondary metabolites of E. californicum has shown that many of the compounds obtained potentially possess interesting flavor properties. The activity of sakuranetin (72) coupled with the inactivity of naringenin (69), in both sweetness-enhancing and bitterness-masking screens, indicates a significant role of the flavonoid A-ring in flavor modulation. The B-ring substitution pattern has been probed earlier, due to homoeriodictyol (55) being the major bitterness-masking agent from the plant, while hesperetin (44) is a patented sweetness- enhancing compound differing from 55 only in the B-ring. When the activities of 3,2'- dihydroxy-4,4',6'-trimethoxychalcone (105) as an in vitro sweetness-enhancer and of the

121 flavone jaceosidin as an antagonist of hTAS2R31 are taken into account, the conclusion can be made that structural subclasses of flavonoids other than flavanones may be useful lead compounds for flavor modulatory applications.

While the sodium salt of the most abundant E. californicum flavonoid constituent, homoeriodictyol (51), has been found active as a bitter-masking agent for a variety of substances (including caffeine, denatonium benzoate, and quinine) it was not tested against saccharin (Ley et al. 2005). In fact, homoeriodictyol (51) was not determined to be a bitterness-masking agent in the in vitro assay against hTAS2R31, which used saccharin (14) as a positive control in the present dissertation research. Although quinine also activates this receptor, the bitterness of quinine is elicited through multiple receptors

(Meyerhof et al., 2010). Therefore, it is feasible that homoeriodictyol (51) would have no effect as a bitterness-masking agent for saccharin. This indicates that homoeriodictyol

(51) is not responsible alone for the full bitterness-masking profile of E. californicum, since other components of the plant may play an important role in the complete bitter- masking profile. In fact, the antagonistic activity of sakuranetin (72) against hTAS2R31 may be partially responsible for the historical use of E. californicum to mask the bitterness of quinine. It is to be hoped that the results achieved in this dissertation research will stimulate future workers to continue the evaluation of E. californicum flavonoid constituents in taste modulation, once additional new methods of biological evaluation become available.

122

References

Acree, T. E.; Lindley, M. Structure-activity relationship and AH-B after 40 years. In

Sweetness and Sweeteners: Biology, Chemistry and Psychophysics; Weerasinghe,

D. K.; Dubois, G. E., Eds.: ACS Symposium Series, American Chemical Society;

Washington, DC, 2008; vol. 979, pp 96-108.

Adler, E.; Hoon, M. A.; Mueller, K. L.; Chandrashekar, J.; Ryba, N. J.; Zuker, C. S. A

novel family of mammalian taste receptors. Cell 2000, 100, 693-702.

Agrawal, P. K., Ed. Carbon-13 NMR of Flavonoids; Elsevier Science Publishers;

Amsterdam, 1989.

Aldin, E.; Reitmeier, C. A.; Murphy, P. Bitterness of soy extracts containing isoflavones

and saponins. J. Food Sci. 2006, 71, S211-S215.

Anonymous. The Pharmacopeia of the United States of America, 9th Decennial Revision;

P. Blakiston's Son & Co.: Philadelphia, 1916; pp 148-149.

Anonymous, Food and Drugs. Fed. Reg. 2002, 67, 512-513.

Arakawa, H.; Nakazaki, M. Absolute configuration of the optically active flavanones. J.

Liebigs Ann. Chem. 1960, 636, 111-117.

Arriaga-Giner F. J.; Wollenweber, E.; Schober, I.; Yatskievych, G. Three new benzoic

acid derivatives from the glandular excretion of Eriodictyon sessilifolium

(Hydrophyllaceae). Z. Naturforsch., C: Biosci. 1988, 43, 337-340.

123

Asahina, Y.; Ueno, S. Phyllodulcin, a chemical constituent of Amacha (Hydrangea

thunbergii Sieb). J. Pharm. Soc. Jpn. 1916, 36, 146.

Asahina, Y.; Inubuse, M. Flavanone glucosides. II. Constitution of naringenin. Ber.

Dtsch. Chem. Ges. B. 1928, 61, 1514-1516.

Assadi-Porter, F. M.; Tonelli, M.; Maillet, E.; Hallenga, K.; Benard, O.; Max, M.;

Markley, J. L. Direct NMR detection of the binding of functional ligands to the

human sweet receptor, a heterodimeric family 3 GPCR. J. Am. Chem. Soc. 2008,

130, 7212-7213.

Azam, S.; Hadi, N.; Khan, N. U.; Hadi, S. M. Antioxidant and prooxidant properties of

caffeine, theobromine and xanthine. Med. Sci. Monit. 2003, 9, BR325-BR330.

Bacon, J. D.; Hannan, G. L.; Fang, N.; Mabry, T. J. Chemosystematics of the

Hydrophyllaceae: flavonoids of three species of Eriodictyon. Biochem. Sys. Ecol.

1986, 14, 591-595.

Bartoshuk, L. M. The biological basis of food perception and acceptance. Food Qual.

Prefer. 1993, 4, 21-32.

Bayard, V.; Chamorro, F.; Motta, J.; Hollenberg, N. K. Does flavanol intake influence

mortality from nitric oxide-dependent processes? Ischemic heart disease, stroke,

diabetes mellitus, and cancer in Panama. Int. J. Med. Sci. 2007, 4, 53-58.

Behrens, M.; Meyerhof, W.; Hellfritsch, C.; Hofmann, T. Sweet and umami taste: natural

products, their chemosensory targets, and beyond. Angew. Chem. Int. Ed. 2011,

50, 2220-2242.

124

Benavente-García, O.; Castillo, J.; Marin, F. R.; Ortuño, A.; Del Río, J. A. Uses and

properties of citrus flavonoids. J. Agric. Food Chem. 1997, 45, 4505-4515.

Bernhardt, S. J.; Naim, M.; Zehavi, U.; Lindemann, B. Changes in IP3 and cytosolic

Ca2+ in response to sugars and non-sugar sweeteners in transduction of sweet

taste in the rat. J. Physiol. 1996, 490, 325-336.

Binello, A.; Cravotto, G.; Nano, G. M.; Spagliardi, P. Synthesis of chitosan–cyclodextrin

adducts and evaluation of their bitter-masking properties. Flavour Frag. J. 2004,

19, 394-400.

Bitto, A; Polito, F.; Squadrito, F.; Marini, H.; D’Anna, R.; Irrera, N.; Minutoli, L.;

Granese, R.; Altavilla, D. Genistein aglycone: a dual mode of action anti-

osteoporotic soy isoflavone rebalancing bone turnover towards bone formation.

Curr. Med. Chem. 2010, 17, 3007-3018.

Bloomgarden, Z. T. Nonnutritive sweeteners, fructose, and other aspects of diet. Diabetes

Care 2011, 34, e46-e51.

Bohm, B. A.; Constant, H. Leaf surface flavonoids of Eriodictyon trichocalyx. Biochem.

Sys. Ecol. 1990, 18, 491-492.

Borzelleca J. F. Paracelsus: herald of modern toxicology. Toxicol. Sci. 2000, 53, 2-4.

Breslin, P. A. S. Components of flavour interactions among salty, sour and bitter

compounds. Trends Food Sci. Technol. 1996, 7, 390-399.

Bridel, M.; Lavielle, R. Sur le principe sucré des feuilles de Kaâ-hê-é (Stevia rebaundiana B). C. R. Acad. Sci. 1931, 192, 1123–1125.

125

Brockhoff, A.; Behrens, M.; Niv, M. Y.; Meyerhof, W. Structural requirements of bitter

taste receptor activation. Proc. Natl. Acad. Sci. U.S.A. 2010, 107, 11110-11115.

Carakostas, M. C.; Curry, L. L.; Boileau, A. C.; Brusick, D. J. Overview: the history,

technical function and safety of rebaudioside A, a naturally occurring steviol

glycoside, for use in food and beverages. Food Chem. Toxicol. 2008, 46, S1-S10.

Carlquist, S.; Eckhart, V. M.; Michene D. C. Wood anatomy of Hydrophyllaceae. I.

Eriodictyon. Aliso 1983, 10, 397-412.

Chaliha, B. P.; Sastry, G. P.; Rao, P. R. Citrus jambhiri peel. Isolation of a new flavone.

Tetrahedron 1965, 21, 1441-1443.

Chandrashekar, J.; Mueller, K. L.; Hoon, M. A.; Adler, E.; Feng, L.; Guo, W.; Zuker, C.

S.; Ryba, N. J. T2Rs function as bitter taste receptors. Cell 2000, 100, 703-711.

Cheeseman, M. A. Agency Response Letter GRAS Notice No. GRN 000282, 2009.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm185772.htm.

Cheeseman, M. A. Agency Response Letter GRAS Notice No. GRN 000301, 2010a.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm200326.htm.

Cheeseman, M. A. Agency Response Letter GRAS Notice No. GRN 000329. 2010b.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm228941.htm.

126

Cheeseman, M. A. Agency Response Letter GRAS Notice No. GRN 000348, 2011a.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm267232.htm.

Cheeseman, M. A. Agency Response Letter GRAS Notice No. GRN 000359, 2011b

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm253511.htm.

Conn, J. W.; Rovner, D. R.; Cohen, E. L. Licorice-induced pseudoaldosteronism.

Hypertension, hypokalemia, aldosteronopenia, and suppressed plasma rennin

activity. JAMA 1968, 205, 492-496.

Cooper, K. A; Donovan, J. L.; Waterhouse, A. L.; Williamson, G. Cocoa and health: a

decade of research. Br. J. Nutr. 2008, 99, 1-11.

Crammer, B. Recent trends of some natural sweet substances from plants. In Selected

Topics in the Chemistry of Natural Products.; Ikan, R., Ed.; World Scientific:

Hackensack, NJ, 2008; pp 189-207.

Crosby, G. A.; Beidler, L. M. New sweeteners. Crit. Rev. Food Sci. Nutr. 1976, 7, 297-

323.

Cui, M.; Jiang, P.; Maillet, E.; Max, M.; Margolskee, R. F.; Osman, R. The heterodimeric

sweet taste receptor has multiple potential ligand binding sites. Curr. Pharm. Des.

2006, 12, 4591-4600.

Damak, S.; Rong, M.; Yasumatsu, K. Detection of sweet and umami taste in the

absence of taste receptor T1r3. Science 2003, 301, 850-853.

127

Dentali, S. J.; Hoffmann, J. J. Potential antiinfective agents from Eriodictyon

angustifolium and Salvia apiana. Int. J. Pharm. 1992, 30, 223-231.

Drewnowski, A; Gomez-Carneros, C. Bitter taste, phytonutrients, and the consumer: a

review. Am. J. Clin. Nutr. 2000, 72, 1424-1435.

DuBois, G. E.; Crosby, G. A.; Stephenson, R. A.; Wingard, R. E., Jr. Dihydrochalcone

sweeteners. Synthesis and sensory evaluation of sulfonate derivatives. J. Agric.

Food Chem. 1977, 25, 763-772.

Duffy, V. B.; Davidson, A. C.; Kidd, J. R.; Kidd, K. K.; Speed, W. C.; Pakstis, A. J.;

Reed, D. R.; Snyder, D. J.; Bartoshuk, L. M. Bitter receptor gene (TAS2R38), 6-

n-propylthiouracil (PROP) bitterness and alcohol intake. Alcohol. Clin. Exp Res.

2004, 28, 1629-1637.

Ejaz, S.; Ejaz, A.; Matsuda, K.; Lim, C. W. Limonoids as cancer chemopreventive agents

J. Sci. Food Agric. 2006, 86, 339-345.

Ellington, A. A; Berhow, M. A.; Singletary, K. W. Inhibition of Akt signaling and

enhanced ERK1/2 activity are involved in induction of macroautophagy by

triterpenoid B-group soyasaponins in colon cancer cells. Carcinogenesis 2006, 27,

298-306.

Embuscado, M. E. Polyols. In Optimising Sweet Taste in Foods; Spillane, W. J., Ed.;

Woodhead Publishing: Cambridge, UK, 2006; pp 153-174.

Fernandez, C.; Fraga, B. M.; Hernandez, M. G.; Arteaga, J. M. Flavonoid aglycones

from some Canary Islands species of Sideritis. J. Nat. Prod. 1988, 51, 591-593.

128

Fimognari, C.; Nüsse, M.; Cesari, R.; Iori, R.; Cantelli-Forti, G.; Hrelia, P. Growth

inhibition, cell-cycle arrest and apoptosis in human T-cell leukemia by the

isothiocyanate sulforaphane. Carcinogenesis 2002, 23, 581-586.

Finley, J. W.; Holliday, D.; King, J. M.; Prudente, A. D., Jr. Masking bitter flavors by

using mogrosides. PCT Int. Appl. WO 2011046768 A2 20110421, 2011.

Follin S. A.; Turkington, C., Eds. Nursing 2004 Herbal Medicine Handbook, 2nd ed.;

Lippincott Williams & Wilkins: Philadelphia, 2004.

Fraser, A. W.; Lewis, J. R. Rutaceae constituents. IV. Flavonoids from Merrillia

caloxylon. Phytochemistry 1974, 13, 1561-1564.

Fuke, S.; Ueda, Y. Interactions between umami and other flavor characteristics. Trends

Food Sci. Technol. 1996, 7, 407-411.

Gaffield, W. Circular dichroism, optical rotatory dispersion and absolute configuration of

flavanones, 3-hydroxyflavanones and their glycosides. Tetrahedron 1970, 26,

4093-4108.

George, E.; Hofmann, T. F.; Stark, T. D. Natural flavour enhancers and methods for

making same. U.S. Pat. Appl. 2011104351, 2011.

Giorgioa, E.; Parrinellob, N.; Caccamese, S.; Rosini C. Non-empirical assignment of the

absolute configuration of (−)-naringenin, by coupling the exciton analysis of the

circular dichroism spectrum and the ab initio calculation of the optical rotatory

power. Org. Biomol. Chem. 2004, 2, 3602-3607.

129

Go, Y.; Satta, Y.; Takenaka, O.; Takahata, N. Lineage-specific loss of function of bitter

taste receptor genes in humans and nonhuman primates. Genetics 2005, 170, 313-

326.

Gottshall, R. Y.; Lucas, E. H.; Lickfeldt A.; Roberts J. M. The occurrence of antibacterial

substances against Mycobacterium tuberculosis in seed plants. J. Clin. Invest.

1949, 28, 920-923.

Guadagni, D. G.; Maier, V. P.; Turnbaugh, J. G. Effect of neodiosmin on threshold and

bitterness of limonin in water and orange juice. J. Food Sci. 1976, 41, 681-684.

Guadagni, D. G.; Horowitz, R. M.; Gentili, B.; Maier, V. P. Method of reducing

bitterness and off-after-taste. U. S. Pat. Appl. 19794154862, 1979.

Guha, N.; Kwan, M. L.; Quesenberry, C. P.; Weltzien, E. K.; Castillo, A. L.; Caan, B. J.

Soy isoflavones and risk of cancer recurrence in a cohort of breast cancer

survivors: the life after cancer epidemiology study. Breast Cancer Res. Treat.

2009, 118, 395-405.

Hadley, W. J.; Gisvold, O. Phytochemical study of Eriodictyon angustifolium, Nuttley. J.

Am. Pharm. Assoc. 1944, 33, 275-277.

Hart, B. The evolution of herbal medicine: behavioural perspectives. Anim. Behav. 2005,

70, 975-989.

Heneczkowski, M.; Kopacz, M.; Nowak, D.; Kuzniar, A. Infrared spectrum analysis of

some flavonoids. Acta Pol. Pharm. 2001, 58, 415-420.

Hinrichs, A. L.; Wang, J. C.; Bufe, B.; Kwon, J. M.; Budde, J.; Allen, R.; Bertelsen, S.;

Evans, W.; Dick, D.; Rice, J.; Foroud, T.; Nurnberger, J.; Tischfield, J. A.;

130

Kuperman, S.; Crowe, R.; Hesselbrock, V.; Schuckit, M.; Almasy, L.; Porjesz, B.;

Edenberg, H. J.; Begleiter, H.; Meyerhof, W.; Bierut, L. J.; Goate, A. M.

Functional variant in a bitter-taste receptor (hTAS2R16) influences risk of alcohol

dependence. Am. J. Hum. Genet. 2006, 78, 103-111.

Holzhauer, W. C. Eriodictyon californicum. Pharm. J., Trans. 1880, 11, 170-171.

Hooker, W. J.; Arnott, G. A. W. The Botany of Captain Beechey's Voyage; Henry G.

Bohn City London Series: London, 1841.

Hoon, M. A; Adler, E.; Lindemeier, J.; Battey, J. F.; Ryba, N. J.; Zuker, C. S. Putative

mammalian taste receptors: a class of taste-specific GPCRs with distinct

topographic selectivity. Cell 1999, 96, 541-551.

Horowitz, R. M.; Gentili, B. Dihydrochalcone derivatives and their use as sweetening

agents. U.S. Pat. Appl. 19633087821, 1963.

Howard, J. L. 1992. Eriodictyon californicum. In Fire Effects Information System,

website: http://www.fs.fed.us/database/feis/

Hsu, C.-C.; Lin, T.-W.; Chang, W.-W.; Wu, C.-Y.; Lo, W.-H.; Wang, P.-H.; Tsai, Y.-C.

Soyasaponin-I-modified invasive behavior of cancer by changing cell surface

sialic acids. Gynecol. Oncol. 2005, 96, 415-422.

Huang, L.; Shanker, Y. G.; Dubauskaite, J.; Zheng, J. Z.; Yan, W.; Rosenzweig, S.;

Spielman, A. I.; Max, M.; Margolskee, R. F. Gγ13 colocalizes with

gustducin in taste receptor cells and mediates IP3 responses to bitter denatonium.

Nature Neurosci. 1999, 2, 1055-1062.

131

Ibrahim, A.-R. S.; Galal, A. M.; Ahmed, M. S.; Mossa, G. S. O-Demethylation and

sulfation of 7-methoxylated flavanones by Cunninghamella elegans. Chem.

Pharm. Bull. 2003, 51, 203-206.

Ichino, K.; Tanaka, H.; Ito, K. Two novel flavonoids from the leaves of Lindera

umbellata var. lancea and L. umbellata. Tetrahedron 1998, 44, 3251-3260.

Iordachescu, G.; Vlasceanu, G.; Bleoanca, I.; Neagu, C.; Iordachescu, G. Umami taste

and the consumer perception. The Annals of the University of Dunarea de Jos

Galati, Fascicle VI, Food Technology 2008, 2, 58-61.

Jeon, S. H.; Chun, W.; Choi, Y. J.; Kwon, Y. S. Cytotoxic constituents from the bark of

Salix hulteni. Arch. Pharm. Res. 2008, 31, 978-982.

Jeon, S. M.; Bok, S. H.; Jang, M. K.; Lee, M. K.; Nam, K. T.; Park, Y. B.; Rhee, S. J.;

Choi, M. S. Antioxidative activity of naringin and lovastatin in high cholesterol-

fed rabbits. Life Sci. 2001, 69, 2855-2866.

Jerome-Morais, A.; Diamond, A. M.; Wright, M. E. Dietary supplements and human

health: for better or for worse? Mol. Nutr. Food Res. 2011, 55, 122-135.

Jerz, G.; Waibel, R.; Achenbach, H. Cyclohexanoid protoflavanones from the stem-bark

and roots of Ongokea gore. Phytochemistry 2005, 66, 1698-1706.

Jia, Z.; Yang, X.; Hansen, C. A.; Naman, C. B.; Simons, C. T.; Slack, J. P.; Gray K.

Consumables. U.S. Pat. Appl. 20100178389, 2010.

Johns, T. Detoxification function of geophagy and domestication of the potato. J. Chem.

Ecol. 1986, 12, 635-646.

132

Johnson, I. T. Glucosinolates in the human diet. Bioavailability and implications for

health. Phytochem. Rev. 2003, 1, 183-188.

Johnson, N. D. Flavonoid aglycones from Eriodictyon californicum resin and their

implications for herbivory and UV screening. Biochem. Sys. Ecol. 1983, 11, 211-

215.

Johnson, R. K.; Appel L. J.; Brands, M.; Howard, B. V.; Lefevre, M.; Lustig, R. H.;

Sacks, F.; Steffen, L. M.; Wylie-Rosett, J. Dietary sugars intake and

cardiovascular health: a scientific statement for the American Heart Association.

Circulation 2009, 120, 1011-1020.

Jung, J. H.; Pummangura, S.; Chaichantipyuth, C.; Patarapanich, C.; McLaughlin, J. L.

Bioactive constituents of Melodorum fruticosum. Phytochemistry 1990, 29, 1667-

1670.

Kashket, S. Sweetness and flavor enhanced compositions and method of preparing such

compositions. U.S. Pat. 4906480, 1990.

Kennelly, E. J. Sweet and non-sweet constituents of Stevia rebaudiana. In Stevia: the

Genus Stevia; Medicinal and Aromatic Plants – Industrial Profiles, Vol. 19;

Kinghorn, A. D., Ed.; Taylor & Francis: London, 2002; pp 68-85.

Kim, S.-Y.; Kim, H.-J.; Lee, M.-K.; Jeon, S.-M.; Do, G.-M.; Kwon, E.-Y.; Cho, Y.-Y.;

Kim, D.-J.; Jeong, K.-S.; Park, Y. B.; Ha, T. Y.; Choi, M.-S. Naringin time-

dependently lowers hepatic cholesterol biosynthesis and plasma cholesterol in rats

fed high-fat and high-cholesterol diet. J. Med. Food 2006, 9, 582-586.

133

Kinghorn, A. D.; Chin, Y.-W.; Pan, L.; Jia, Z. Natural products as sweeteners and

sweetness modifiers. In Comprehensive Natural Products Chemistry II:

Chemistry and Biology; Vol. 3, Development & Modification of Bioactivity;

Mander, L.; Liu, H.-W., Ser. Eds.; Verpoorte, R., Vol. Ed.; Elsevier: Oxford,

U.K., 2010; pp 269-315.

Kinghorn, A. D.; Compadre, C. M. Less common high-potency sweeteners. In

Alternative Sweeteners, 4th ed.; O’Brien Nabors, L., Ed.; Taylor and Francis:

Boca Raton, FL, 2012; pp 223-246.

Kitagawa, M.; Kusakabe, Y.; Miura, H.; Ninomiya, Y.; Hino, A. Molecular genetic

identification of a candidate receptor gene for sweet taste. Biochem. Biophys. Res.

Commun. 2001, 283, 236-242.

Kofink, M.; Papagiannopoulos, M.; Galensa, R. (-)-Catechin in cocoa and chocolate:

occurrence and analysis of an atypical flavan-3-ol enantiomer. Molecules 2007,

12, 1274-1288.

Kohda, H.; Kasai, R.; Yamasaki, K.; Murakami K.; Tanaka O. New sweet diterpene

glucosides from Stevia rebaudiana. Phytochemistry 1976, 15, 981-983.

Krause, M.; Galensa, R. Analysis of enantiomeric flavanones in plant extracts by high-

performance liquid chromatography on a cellulose triacetate based chiral

stationary phase. Chromatographia 1991, 32, 69-72.

Krebs, J. R. The gourmet ape: evolution and human food preferences. Am. J. Clin. Nutr.

2009, 90, 707-711.

134

Kroger, M.; Meister, K.; Kava, R. Low-calorie sweeteners and other sugar substitutes: a

review of the safety issues. Compr. Rev. Food Sci. Food Safety 2006, 5, 35-47.

Krohn, M.; Zinke, H. Oligopeptides for use as taste modulators. Eur. Pat. EP2093234,

2009.

Kurihara, K. Glutamate: from discovery as a food flavor to role as a basic taste (umami).

Am. J. Clin. Nutr. 2009, 90, 719S-722S.

Lamartiniere, C. A. Protection against breast cancer with genistein: a component of soy.

Am. J. Clin. Nutr. 2000, 71, 1705S-1707S; discussion 1708S-1709S.

Lawless, H. T.; Stevens, D. A.; Chapman, K. T.; Kurtz, A. Metallic taste from electrical

and chemical stimulation. Chem. Senses 2005, 30, 185-194.

Lee, K. W.; Kim, Y. J.; Lee, H. J.; Lee, C. Y. Cocoa has more phenolic phytochemicals

and a higher antioxidant capacity than teas and red wine. J. Agric. Food Chem.

2003, 51, 7292-7295.

Lee, S. Y. H.; Munerol, B.; Pollard, S.; Youdim, K. A.; Pannala, A. S.; Kuhnle, G. G. C.;

Debnam, E. S.; Rice-Evans, C.; Spencer, J. P. E. The reaction of flavanols with

nitrous acid protects against N-nitrosamine formation and leads to the formation

of nitroso derivatives which inhibit cancer cell growth. Free Rad. Biol. Med.

2006, 40, 323-334.

Levin, G. V. Tagatose, the new GRAS sweetener and health product. J. Med. Food 2002,

5, 23-36.

135

Levin, G. V.; Zelmer, L. R.; Saunders, J. P.; Beadle, J. R. Sugar substitutes : their energy

values, bulk characteristics, and potential health benefits. Am. J. Clin. Nutr. 1995,

62, 1161S-1168S.

Ley, J. P. Masking bitter taste by molecules. Chemosens. Percept. 2008, 1, 58-77.

Ley, J. P.; Krammer, G.; Kindel G.; Gatfield, I. L.; Muller, M. Use of hydroxyflavanones

for masking bitter taste. U.S. Pat. Appl. 20020188019, 2002.

Ley, J. P.; Krammer, G.; Reinders, G.; Gatfield, I. L.; Bertram, H.-J. Evaluation of bitter

masking flavanones from Herba Santa (Eriodictyon californicum (H. and A.)

Torr., Hydrophyllaceae). J. Agric. Food Chem. 2005, 53, 6061-6066.

Ley, J. P.; Blings, M.; Paetz, S.; Krammer, G. E.; Bertram, H.-J. New bitter-masking

compounds: hydroxylated benzoic acid amides of aromatic amines as structural

analogues of homoeriodictyol. J. Agric. Food Chem. 2006, 54, 8574-8579.

Ley, J. P.; Blings, M.; Paetz, S.; Kindel, G.; Freiherr, K.; Krammer, G. E.; Bertram, H.-J.

Enhancers for sweet taste from the world of non-volatiles: polyphenols as taste

modifiers. In Sweetness and Sweeteners: Biology, Chemistry and Psychophysics;

Weerasinghe, D. K.; Dubois, G. E. Eds.; ACS Symposium Series, American

Chemical Society: Washington DC, 2008a; vol. 979, pp 400-409.

Ley, J.; Freiherr, K.; Krammer, G.; Reinders G. Aroma compositions of alkamides with

hesperetin and/or 4-hydroxydihydrochalcones and salts thereof for enhancing

sweet sensory impressions. U.S. Pat. Appl. 20080242740 2008b.

136

Ley, J.; Kindel, G.; Paetz, S.; Riess, T.; Haug, M.; Schmidtmann, R.; Krammer, G. Use

of hesperetin for enhancing the sweet taste. U.S. Patent Application

20080305052, 2008c.

Ley, J. P.; Paetz, S.; Blings, M.; Hoffmann-Lücke, P.; Bertram, H.-J.; Krammer, G. E.

Structural analogues of homoeriodictyol as flavor modifiers. Part III: short chain

gingerdione derivatives. J. Agric. Food Chem. 2008d, 56, 6656-6664.

Ley, J. P.; Paetz, S.; Blings, M.; Riess, T.; Krammer, G. E. Structural analogues of

hispolon as flavor modifiers. In Expression of Multidisciplinary Flavour Science,

Proceedings of the 12th Qeurman Symposium, July 1-4 2008, Interlaken,

Switzerland; Blank, I.; Wuest, M.; Yeretzian, C., Eds.; Zuercher Hochschule

fuer Angewandte Wissenschaften, Institut fuer Chemie und Biologische Chemie

Waedenswil, Germany, 2010; pp 402-405.

Lindemann, B. Chemoreception: tasting the sweet and the bitter. Curr. Biol. 1996, 6,

1234-1237.

Lindemann, B. Receptor seeks ligand: on the way to cloning the molecular receptors for

sweet and bitter taste. Nature Med. 1999, 5, 381-382.

Lindemann, B. Receptors and transduction in taste. Nature 2001, 413, 219-225.

Li, R.; Fan, N.; Mabry, T. J. Flavonoids and a coumarin from Gutierrezia

sphaerocephala. Phytochemistry 1988, 27, 1556-1559.

Liu, Y-K.; Ho, D. K.; Cassady, J. M. Isolation of potential cancer chemopreventive

agents from Eriodictyon californicum. J. Nat. Prod. 1992, 55, 357-363.

137

Matsuura, M.; Obata, A.; Fukushima, D. Objectionable flavor of soy milk developed

during the soaking of soybeans and its control. J. Food Sci. 1989, 54, 602-605.

Mattes, R. D. Fat taste and lipid metabolism in humans. Physiol. Behav. 2005, 86,

691-697.

Max, M.; Shanker, Y. G.; Huang, L.; Rong, M.; Liu, Z.; Campagne, F.; Weinstein, H.;

Damak, S.; Margolskee, R. F. Tas1r3, encoding a new candidate taste receptor, is

allelic to the sweet responsiveness locus Sac. Nature Genet. 2001, 28, 58-63.

McKinnon, S. K.; McLaughlin, P. J.; Margolskee, R. F. Gustducin is a taste-cell-specific

G protein closely related to the transducins. Nature 1992, 357, 563-569.

Melliou, E.; Chinou, I. Chemical analysis and antimicrobial activity of Greek propolis.

Planta Med. 2004, 70, 515-519.

Mennella, J. A.; Pepino, M. Y.; Reed, D. R. Genetic and environmental determinants of

bitter perception and sweet preferences. Pediatrics 2005, 115, e216-e222.

Meyerhof, W.; Batram, C.; Kuhn, C.; Brockhoff, A.; Chudobam, E.; Bufe, B.;

Appendino, G.; Behrens, M. The molecular receptive ranges of human TAS2R

bitter taste receptors. Chem. Senses 2010, 35, 157-70.

Min, S.-W.; Kim, N.-J.; Baek, N.-I.; Kim, D.-H. Inhibitory effect of eupatilin and

jaceosidin isolated from Artemisia princeps on carrageenan-induced inflammation

in mice. J. Ethnopharmacol. 2009, 125, 497-500.

Ming, D.; Ninomiya, Y.; Margolskee, R. F. Blocking taste receptor activation of

gustducin inhibits gustatory responses to bitter compounds. Proc. Natl. Acad. Sci.

U.S.A. 1999, 96, 9903-9908.

138

Mizutani, K.; Tanaka, O. The use of Stevia rebaudiana sweeteners in Japan. In Stevia:

the Genus Stevia; Medicinal and Aromatic Plants – Industrial Profiles, Vol. 19;

Kinghorn, A. D., Ed.; Taylor & Francis: London, 2002; pp 178-195.

Mizutani K.; Kuramoto T.; Tamura Y.; Ohtake N.; Doi S.; Nakaura M.; Tanaka O.

Sweetness of glycyrrhetic acid 3-O-β-D-monoglucuronide and the related

glycosides. Biosci. Biotechnol. Biochem. 1994, 58, 554-555.

Montmayeur, J. P.; Liberles, S. D.; Matsunami, H.; Buck, L. B. A candidate taste

receptor gene near a sweet taste locus. Nature Neurosci. 2001, 4, 492-498.

Morini, G.; Bassoli, A.; Temussi, P. A. From small sweeteners to sweet proteins:

anatomy of the binding sites of the human T1R2-T1R3 receptor. J. Med. Chem.

2005, 48, 5520-5529.

Morrison, A. B. Sugar substitutes. Can. Med. Assoc. J. 1979, 120, 633-635.

Moskowitz, H. R. Models of sweetness additivity. J. Exp. Psychol. 1973, 99, 88-98.

Mossler, G. The chemical investigation of Eriodictyon glutinosum. J. Liebigs Ann. Chem.

1907, 351, 233-254.

Nakamura, T.; Tanigake, A.; Miyanaga, Y.; Ogawa, T.; Akiyoshi, T.; Matsuyama, K.;

Uchida, T. The effect of various substances on the suppression of the bitterness of

quinine-human gustatory sensation, binding, and taste sensor studies. Chem.

Pharm. Bull. 2002, 50, 1589-1593.

Nakasugi, T.; Nakashima, M.; Komai, K. Antimutagens in gaiyou (Artemisia argyi Levl.

et Vant.). J. Agric. Food Chem. 2000, 48, 3256-3266.

139

Neely, J. S.; Thompson, J. A. Culinary composition containing

paramethoxycinnamaldehyde as a flavoring agent and sweetener. U.S. Pat.

3,908,026, 1975.

Nelson, G.; Hoon, M. A.; Chandrashekar, J.; Zhang, Y.; Ryba, N. J. P.; Zuker, C. S.

Mammalian sweet taste receptors. Cell 2001, 106, 381-390.

Nelson, G.; Chandrashekar, J.; Hoon, M. A.; Feng, L.; Zhao, G.; Ryba, N. J. P.; Zuker, C.

S. An amino-acid taste receptor. Nature 2002, 416, 199-202.

Nie, Y.; Vigues, S.; Hobbs, J. R.; Conn, G. L.; Munger, S. D. Distinct contributions of

T1R2 and T1R3 taste receptor subunits to the detection of sweet stimuli. Curr.

Biol. 2005, 15, 1948-1952.

Nofre, C.; Tinti, J. M. New dipeptide derivatives and analogs useful as sweetening

agents, and process for their preparation. French Pat. Appl. 2697844 A1

19940513, 1994.

Nofre, C.; Tinti, J.-M. Sweetness reception in man: the multipoint attachment theory.

Food Chem. 1996, 56, 263-274.

Norell, S. E.; Ahlbom, A.; Erwald, R.; Jacobson, G.; Lindberg-Navier, I.; Olin, R.;

Törnberg, B.; Wiechel K.-L. Diet and pancreatic cancer: a case-control study.

Am. J. Epidemiol. 1986, 124, 894-902.

Oertling, H.;, Loges, H.;, Machinek, A.;, Simchen, U.;, Surburg, H. Teeth cleaning

compound containing centhol with reduced bitter sensation. U.S. Patent

20110081303, 2011.

140

Oertly, E.; Myers, R. G. A new theory relating constitution to taste. J. Am. Chem. Soc.

1919, 41, 855- 867.

Ottinger, H.; Soldo, T.; Hofmann, T. Discovery and structure determination of a novel

Maillard-derived sweetness enhancer by application of the comparative taste

dilution analysis (cTDA). J. Agric. Food Chem. 2003, 51, 1035-1041.

Palmer, R. K.; Salemme, F. R. Sweetness enhancers including rebaudioside A or D.

International/U.S. Pat. Appl. 2010/047207, 2010.

Perez, J. L.; Jayaprakasha, G. K.; Cadena, A.; Martinez, E.; Ahmad, H.; Patil, B. S. In

vivo induction of phase II detoxifying enzymes, glutathione transferase and

quinone reductase by citrus triterpenoids. BMC Complement. Altern. Med. 2010,

10, 51-58.

Perkin, A. G. Derivatives of hesperetin. J. Chem. Soc., Trans. 1898, 73, 1031-1038.

Perkin, A. G.; Hummel, J. J. The colouring principle of the flowers of the Butea

frondosa. J. Chem. Soc., Trans. 1904, 85, 1459-1472.

Pickenhagen, W.; Dietrich, P.; Borivoij, K.; Polonsky, J.; Nouaille, F.; Lederer, E.

Identification of the bitter principle of cocoa. Helv. Chim. Acta. 1975, 58, 1078-

1086.

Polkowski, K.; Mazurek, Aleksander, P. Biological properties of genistein. a review of in

vitro and in vivo data. Acta Pol. Pharm. 2000, 57, 135-155.

Potter, S. M.; Baum, J. A.; Teng, H.; Stillman, R. J.; Shay, N. F.; Erdman Jr, J. W. Soy

protein and isoflavones: their effects on blood lipids and bone. Am. J. Clin. Nutr.

1998, 68, 1375-1379.

141

Power, F. B.; Tutin F. Chemical examination of Eriodictyon. Pharm. Rev. 1906, 24, 300-

304.

Prakash, I.; Dubois, G. E.; King, G. A.; San Miguel, R. I. Sweetener compositions having

enhanced sweetness and improved temporal and/or flavor profiles. U.S. Pat. Appl.

2009053378, 2009.

Puri, M.; Marwaha, S. S.; Kothari, R. M.; Kennedy, J. F. Biochemical basis of bitterness

in citrus fruit juices and biotech approaches for debittering. Crit. Rev. Biotechnol.

1996, 16, 145-155.

Reichelt, K. V.; Hartmann, B.; Weber, B.; Ley, J. P.; Krammer, G. E.; Engel, K.-H.

Identification of bisprenylated benzoic acid derivatives from Yerba Santa

(Eriodictyon spp.) using sensory-guided fractionation. J. Agric. Food Chem.

2010, 58, 1850-1859.

Riemer, J. A. Bitterness inhibitors. U.S. Pat. Appl. 5,336,513, 1994.

Ritter, F. W. Eriodictyon glutinosum. Am. J. Pharm. 1895, 67, 565-572.

Roper, S. D. Signal transduction and information processing in mammalian taste buds.

Pflügers Archiv. Eur. J. Physiol. 2007, 454, 759-776.

Rother, R. Some constituents of Yerba Santa. Am. J. Pharm. 1887, 59, 225.

Roy, G. Bitterness: reduction and inhibition. Crit. Rev. Food Sci. Nutr. 1990, 29, 59-71.

Salemme, R. F.; Long, D.; Palmer, R. K.; Brennan, F. X.; Sprous, D. Rebaudioside C and

its stereoisomers as natural product sweetness enhancers. U.S. Pat. Appl.

20110070172, 2011.

142

Salle, A. J.; Jann, G. J.; Wayne, L. G. Studies on the antibacterial properties of

Eriodictyon californicum. Arch. Biochem. Biophys. 1951, 32, 121-123.

Sayre L. E. A Manual of Organic Materia Medica and Pharmacognosy; P. Blakiston's

Son and Co.: Philadelphia, 1917.

Schiffman, S. S. Taste quality and neural coding: Implications from psychophysics and

neurophysiology. Physiol. Behav. 2000, 69, 147-159.

Seitembetova, A. Zh. Relationship between antioxidant activity and molecular structure

of some flavonoids. Izvestiya Ministerstva Obrazovaniya i Nauki Respubliki

Kazakhstan, Natsional'noi Akademii Nauk Respubliki Kazakhstan, Seriya

Khimicheskaya 2000, 2, 107-111. (Chem. Abstr. 2000, 134, 172679).

Selenski, C.; Pettus, T. R. R. (±)-Diinsininone: made Nature’s way. Tetrahedron 2006,

62, 5298-5307.

Shallenberger, R. S.; Acree, T. E. Molecular theory of sweet taste. Nature 1967, 216,

480-482.

Sharma, S.; Lewis, S. Taste masking technologies: a review. Int. J. Pharm. Pharm. Sci.

2010, 2, 6-13.

Sheridan, C. A taste of the future. Nature Biotechnol. 2004, 22, 1203-1205.

Shu, P.; Hong, J.-L.; Wu, G.; Yu, B.-Y.; Qin, M.-J. Analysis of flavonoids and phenolic

acids in Iris tectorum by HPLC-DAD-ESI-MSn. Zhongguo Tianran Yaowu

2010, 8, 202-207

Slack, J. P. Nucleic acid polypeptide and its use. PCT Int. Appl. WO147275, 2007.

143

Slack, J. P.; Simons, C. T.; Hansen, C. A. Method relating to sweetness enhancement.

U.S. Pat. Appl. 20090311686, 2009.

Slack, J. P.; Brockhoff, A.; Batram, C.; Menzel, S.; Sonnabend, C.; Born, S.; Galindo, M.

M.; Kohl, S.; Thalmann, S.; Ostopovici-Halip, L.; Simons, C. T.; Ungureanu, I.;

Duineveld, K.; Bologa, C. G.; Behrens, M.; Furrer, S.; Oprea, T. I.; Meyerhof, W.

Modulation of bitter taste perception by a small molecule hTAS2R antagonist.

Curr. Biol. 2010, 20, 1104-1109.

Sohi, H.; Sultana, Y.; Khar, R. K. Taste masking technologies in oral pharmaceuticals:

recent developments and approaches. Drug Dev. Ind. Pharm. 2004, 30, 429-448.

Spielman, A. I. Gustducin and its role in taste. J. Dent. Res. 1998, 77, 539-544.

Sreebny, L. M. Sugar availability, sugar consumption and dental caries. Commun. Dent.

Oral Epidemiol. 1982, 10, 1-7.

Tarantino L. M. Agency Response Letter GRAS Notice No. GRN 000252, 2008a.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/ GRASListings/ucm154988.htm

Tarantino L. M. Agency Response Letter GRAS Notice No. GRN 000253, 2008b.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm154989.htm

Tarantino L. M. Agency Response Letter GRAS Notice No. GRN 000275, 2009.

http://www.fda.gov/Food/FoodIngredientsPackaging/GenerallyRecognizedasSafe

GRAS/GRASListings/ucm171539.htm

144

Tatuta, H. Asymmetric synthesis. II. Synthesis of optically active hydroxyflavanones

from hydroxychalcones. Nippon Kagaku Kaishi (1921-47), 1940, 61, 1048-1050.

Temussi, P. A. Natural sweet macromolecules: how sweet proteins work. Cell. Mol. Life

Sci. 2006, 63, 1876-88.

Terry, P.; Wolk, A.; Persson, I.; Magnusson, C. Brassica vegetables and breast cancer

risk. J. Am. Med. Assoc. 2001, 285, 2975-2976.

Tutin, F. The constitution of eriodictyol, of homoeriodictyol, and of hesperitin. J. Chem.

Soc., Trans. 1910, 97, 2054-2062.

Tutin, F.; Clewer, H. W. B. Chemical examination of Eriodictyon. Part II. J. Chem. Soc.,

Trans. 1909, 95, 81-87.

Uchida M.; Rüedi, P.; Eugster, C. H. Drüsenfarbstoffe aus Labiaten: Ecklonochinone A

und B, zwei neuartige Dibenzo-p-dioxin-o-chinone aus Plectranthus ecklonii

Benth. Helv. Chim. Acta 1980, 63, 225–231. van der Wel, H.; Loeve, K. Isolation and characterization of thaumatin I and II, the

sweet-tasting proteins from Thamatococcus daniellii Benth. Eur. J. Biochem.

1972, 31, 221-225.

Vasconcelos, J. M.; Silva, A. M. S.; Cavaleiro, J. A. S. Chromones and flavanones from

Artemisia campestris subsp. maritima. Phytochemistry 1998, 49, 1421-1424.

Verhoeven, D. T. H.; Goldbohm, R. A.; Poppel, G. V.; Verhagen, H. Epidemiological

studies on Brassica vegetables and cancer epidemiological studies on Brassica

vegetables and cancer risk. Cancer Res. 1996, 5, 733-748.

145

Vikram, A.; Jesudhasan, P. R.; Jayaprakasha, G. K.; Pillai, B. S.; Patil, B. S. Grapefruit

bioactive limonoids modulate E. coli O157:H7 TTSS and biofilm. Int. J. Food

Microbiol. 2010, 140, 109-116.

Wall, M. E.; Wani, M. C.; Brown, D. M.; Fullas, F.; Oswald, J. B.; Josephson, F. F.;

Thornton, N. M.; Pezzuto, J. M.; Beecher, C. W. W.; Farnsworth, N. R.; Cordell,

G. A.; Kinghorn, A. D. Effect of tannins on screening of plant extracts for enzyme

inhibitory activity and techniques for their removal. Phytomedicine 1996, 3, 281–

286.

Walters, D. E. The rational discovery of sweeteners. In Sweeteners; Walters, D. E.;

Orthoefer, F. T.; DuBois, G. E., Eds.; ACS Symposium Series, American

Chemical Society: Washington, DC, 1991;vol. 450, pp 1-11.

Wang, H.-J.; Murphy, P. A. Mass balance study of isoflavones during soybean

processing. J. Agric. Food Chem. 1996, 44, 2377-2383.

Wang, J. C.; Hinrichs, A. L.; Bertelsen, S.; Stock, H.; Budde, J. P.; Dick, D. M.; Bucholz,

K. K.; Rice, J.; Saccone, N.; Edenberg, H. J.; Hesselbrock, V.; Kuperman, S.;

Schuckit, M. A; Bierut, L. J.; Goate, A. M. Functional variants in TAS2R38 and

TAS2R16 influence alcohol consumption in high-risk families of African-

American origin. Alcohol. Clin. Exp. Res. 2007, 31, 209-215.

Wang, X.; Thomas, S. D.; Zhang, J. Relaxation of selective constraint and loss of

function in the evolution of human bitter taste receptor genes. Hum. Mol. Gen.

2004, 13, 2671-2678.

146

Watson, L.; Dallwitz, M. J. The Families of Flowering Plants: Descriptions, Illustrations,

Identification, and Information Retrieval, 4th Version, available from 1992.

http://delta-intkey.com.

Weisburger, J. H. Chemopreventive effects of cocoa polyphenols on chronic diseases.

Exp. Biol. Med. 2001, 226, 891-897.

Wessjohann, L.; Backes, M.; Ley, J.; Paetz, S.; Reichelt, K. Use of hydroxyflavone

derivatives for modifying taste. Eur. Pat. EP2253226, 2010.

Whiting, D. A. Natural phenolic compounds 1900 – 2000: a bird’s eye view of a.

century’s chemistry. Nat. Prod. Rep. 2001, 18, 583–606.

Wink, M. Chemische Verteidigung der Lupinen: Zur biologischen Bedeutung der

Chinolizidinalkaloide Theor. Appl. Genet. 1988, 75, 225-233

Winnig, M.; Bufe, B.; Kratochwil, N. A; Slack, J. P.; Meyerhof, W. The binding site for

neohesperidin dihydrochalcone at the human sweet taste receptor. BMC Struct.

Biol. 2007, 7, 66-77.

Wollenweber, E.; Mann, K. Exudate flavonoids in three essential oil plants from the

Ciskei (South Africa). Fitoterapia 1989, 60, 249-251.

Yan, L.; Spitznagel, E. L. Meta-analysis of soy food and risk of prostate cancer in men.

Int. J. Cancer 2005, 117, 667-669.

Yan, W.; Sunavala, G.; Rosenzweig, S.; Dasso, M. A. X.; Brand, J. G.; Spielman, A. I.;

Dasso, M. Bitter taste transduced by PLC-β 2-dependent rise in IP 3 and α-

gustducin-dependent fall in cyclic nucleotides. Am. J. Physiol. Cell Physiol. 2001,

280, 742-751.

147

Yang, G.; Shu, X.; Li, H.; Chow, W.; Cai, H.; Zhang, X.; Gao, Y.; Zheng, W.

Prospective cohort study of soy food intake and colorectal cancer risk. Am. J.

Clin. Nutr. 2009, 89, 577-583.

Xu, H.; Staszewski, L.; Tang, H.; Adler, E.; Zoller, M.; Li, X. Different functional roles

of T1R subunits in the heteromeric taste receptors. Proc. Natl. Acad. Sci. U.S.A.

2004, 101, 14258-14263.

148

APPENDIX A

1. Sweetness-enhancing bioassay

Compounds from Eriodictyon californicum and a commercially available chalcone were evaluated in vitro by measuring receptor-dependent calcium flux in cells expressing a modified chimeric sweet receptor heterodimer. This bioassay was conducted at Givaudan Flavors Corp., Cincinnati, OH and largely follows the cited literature (Slack,

2007). For these studies, the modified sweet receptor complex was constructed by fusing the extracellular domain of the human calcium sensing receptor (hCaSR) to the transmembrane heptahelical domains derived from hTAS1R2 and hTAS1R3, respectively. Briefly, HEK-293T cells already expressing a chimeric G-protein α-subunit were stably transfected with the cDNAs (CaSR:T1R2 and CaSR:T1R3) encoding the chimeric sweet receptor complex, thus resulting in a functional sweet receptor complex that responds to substances that bind to sites in the transmembrane domains. Compounds tested were evaluated at 25 μM to determine if they activated the sweet receptor complex.

Putative actives were also evaluated for activity in host cells to ensure that any apparent agonist activity was indeed targeted to the chimeric receptor and not simply due to non- specific effects on intracellular calcium handling. Compounds with activity of 50% of

149 the positive control cyclamate (15) were considered active. A more detailed description of the bioassay employed is described in the patent by Dr. Jay Slack (2007) as follows:

Fluo-4 is a fluorescent indicator for intracellular calcium and allows to determine changes in the calcium concentration, in particular an increase in response to receptor activation occurring after ligand addition. HEK293 cells stably expressing Gα16-gustducin 44 were used as host cells and transfected with various constructs as described in example 4. Black, clear-bottom 96-well plates were used for all assays. They were seeded the day before with 8500 transfected cells per well and maintained at 37° C. overnight in an a growth medium appropriate for the cells used. For HEK293 cells, Dulbecco's Modified Eagle medium containing high glucose, L-glutamine, pyroxidine hydrochloride, and supplemented with 10% fetal bovine serum was used for growth and maintenance of the HEK293 cells. At the time of the assay, the growth medium was discarded and cells were incubated for 1 hour (at 37° C. in the dark) with 50 μl of a calcium assay solution consisting of 1.5 μM Fluo-4 AM (Molecular Probes™, Invitrogen, San Diego, CA) and 2.5 μM probenicid (Sigma-Aldrich, Saint Louis, MO) dissolved in a reduced calcium C1 buffer solution. Reduced calcium C1 buffer solution contains 130 mM NaCl, 5 mM KCl, 10 mM Hepes, 0.5 mM CaCl2 (reduced from 2 mM) and 10 mM glucose (pH 7.4). After the initial 1 hour loading period, the plates were washed 5 times with 100 μl per well of reduced calcium C1 buffer using an automated plate washer (BioTek, Winooski, VT) and after washing, the plate was further incubated for 30 minutes at room temperature in the dark to allow for complete de-esterification of the Fluo-4-AM. The buffer solutions were discarded, the plate was washed 5 times with 100 μl reduced calcium C1 wash buffer and finally the cells were reconstituted in 180 μl of reduced calcium C1 wash buffer. For assay reading, the plate was placed in a FLIPR fluorescence imaging plate reader (FLIPR-Tetra, Molecular Devices, Sunnyvale, CA), and receptor activation was initiated following addition of 20 μl of a 10× concentrated ligand stock solution, which were prepared in reduced calcium C1 buffer. Fluorescence was continuously monitored for 15 seconds prior to ligand addition and for 105 seconds after ligand addition (45-105 sec may be sufficient). Receptor activation is given in relative fluorescence units (RFU) and is defined by the following equation: Fluorescence Increase=Maximum Fluorescence−baseline fluorescence, wherein the baseline fluorescence represents the mean fluorescence calculated for the first 10 to 15 seconds prior to ligand addition. 150

As a negative control, mock transfected cells were exposed to the same concentration of ligand and the concentration of calcium traces not corresponding to a signal was determined. Cells with an activated receptor were identified by the signal (RFU) being significantly above the negative control.

2. Bitterness-masking bioassay

Compounds from Eriodictyon californicum were evaluated in vitro by measuring receptor-dependent calcium flux in HEK-293T cells transfected with the bitterness receptor hTAS2R31, and expressing a chimeric G-protein α-subunit. This bioassay was conducted at Givaudan Flavors Corp., Cincinnati, OH and largely follows the cited literature (Slack et al., 2010). A concentration of 1 mM sodium saccharin (14) was employed in this study as the bitter taste elicitor. These compounds were tested at 25 μM in an initial screen to determine if they inhibit activation of hTAS2R31 by saccharin. Test compounds were also evaluated for inhibition of a non-taste GPCR pathway in order to reduce the potential for false positive results. Compounds able to reduce to response to saccharin to below 50% in this initial screen were considered active. The half-maximal inhibitory concentration (IC50) was then determined for selectively active components in order to assess the potency of the receptor antagonists. Except where differing from the work described above the procedure was described by Slack et al. (2010) as follows:

On Day 0, the hTAS2R31 cell line was pre-plated at a density of 15,000 cells per well in DMEM + 10% FBS in black, clear bottom 96-well plates that had been pre-coated with 0.001 % poly(ethyleneimine) (MW = ~60,000, Acros Organics, Morris Plains, NJ). On day 2, highthroughput screening of antagonists was performed via calcium imaging using Fluo-4. Briefly, growth medium was discarded and the cells were incubated in the dark for 1 hour at 37° C in 50 μl loading buffer consisting of 1.5 μM Fluo- 4 AM (Invitrogen, San Diego, CA) and 2.5 μM probenicid (Sigma- Aldrich, St. Louis, MO) in DMEM (no FBS). After incubation, the plates 151 were washed 5X with 100 μl of assay buffer (described above) and further incubated in the dark at room temperature for 30 minutes. The cells were then washed 5X with 100 μl assay buffer and then calcium responses were measured in a FLIPRTETRA (Molecular Devices, Sunnydale, CA). Test compounds were prepared at a final concentration of 25 μM in the presence of an EC50 concentration of saccharin (500 μM) and assessed for their ability to decrease the hTAS2R31 response to saccharin. Putative hits that showed >50% inhibition of the saccharin response during primary screening were selected and retested for their ability to inhibit hTAS2R31 activation by saccharin as well as acesulfame K. Hits that also caused a decrease in the agonist responses of a non-related GPCR pathway (isoproterenol, β1/β2-adrenergic receptor agonist) were considered as non- specific inhibitors.

152