Prenylated isoflavonoids from soya and licorice

Analysis, induction and in vitro estrogenicity

Rudy Simons

Thesis committee

Thesis supervisor: Prof. Dr. Ir. Harry Gruppen Professor of Food Chemistry Wageningen University

Thesis co-supervisor : Dr. Ir. Jean-Paul Vincken Assistent Professor, Laboratory of Food Chemistry Wageningen University

Other members : Prof. Dr. Renger Witkamp Wageningen University

Dr. Nigel Veitch Royal Botanic Gardens, Kew, United Kingdom

Prof. Dr. Robert Verpoorte Leiden University

Dr. Henk Hilhorst Wageningen University

This research was conducted under the auspices of the Graduate School VLAG ( Voeding, Levensmiddelentechnologie, Agrobiotechnologie en Gezondheid )

Prenylated isoflavonoids from soya and licorice

Analysis, induction and in vitro estrogenicity

Rudy Simons

Thesis submitted in fulfilment of the requirements of the degree of doctor at Wageningen University by the authority of the Rector Magnificus Prof. Dr. M.J. Kropff in the presence of the Thesis Committee appointed by the Academic Board To be defended in public on Tuesday 28 June 2011 at 1.30 p.m. in the Aula.

Rudy Simons

Prenylated isoflavonoids from soya and licorice Analysis, induction and in vitro estrogenicity

Ph.D. thesis, Wageningen University, Wageningen, the Netherlands (2011) With references, with summaries in Dutch and English

ISBN: 978-90-8585-943-7

ABSTRACT

Prenylated isoflavonoids are found in large amounts in soya bean ( Glycine max ) germinated under stress and in licorice ( Glycyrrhiza glabra ). Prenylation of isoflavonoids has been associated with modification of their estrogenic activity. The aims of this thesis were (1) to provide a structural characterisation of isoflavonoids, in particular the prenylated isoflavonoids occurring in soya and licorice, (2) to increase the estrogenic activity of soya beans by a malting treatment in the presence of a food-grade fungus, and (3) to correlate the in vitro agonistic/antagonistic estrogenicity with the presence of prenylated isoflavonoids. A reversed-phase UHPLC-DAD-MS n-based screening method was developed that allowed the tentative identification of all prenylated flavonoids in licorice extracts using neutral losses diagnostic for prenylation. Moreover, both the chain and pyran ring can be discriminated. This method was also employed on extracts from soya seedlings, which were challenged with the food-grade fungus Rhizopus spp. Besides known prenylated pterocarpans, also novel prenylated isoflavones and coumestans were found in soya. Subsequently, a licorice root extract was fractionated and the fractions were screened for estrogenic activity using an in vitro yeast ER bioassay. Several fractions were considered estrogenically active on one or both ER-subtypes (α and β). The estrogenic activity of some fractions was associated with the presence of glabrene, a prenylated isoflavene. The predominant phytoestrogen of licorice root, glabridin, did not show any agonistic activity, but was found to be a potent ERα antagonist. Large-scale Rhizopus -challenging of soya seedlings resulted in a 10 to 12 fold increase of the isoflavonoid content, in a diversification of isoflavonoid composition, and in an increase of estrogenic activity. The measured estrogenicity of the extracts on both ERs was lower than the theoretically calculated estrogenicity, indicating the presence of antagonists. This antagonistic activity was hypothesized to be associated with the presence of prenylated isoflavonoids.

TABLE OF CONTENTS

Chapter 1 General introduction 1

Chapter 2 A rapid screening method for prenylated flavonoids with 37 UHPLC-ESI-MS in licorice root extracts

Chapter 3 Identification of prenylated pterocarpans and other 59 isoflavonoids in Rhizopus spp. elicited soya bean seedlings by electrospray ionisation mass spectrometry

Chapter 4 Agonistic and antagonistic estrogens in licorice root 77 (Glycyrrhiza glabra )

Chapter 5 Increasing soya isoflavonoid content and diversity by 95 simultaneous malting and challenging by a fungus to modulate estrogenicity

Chapter 6 General discussion 117

Appendix 139

Summary 155

Samenvatting 159

Acknowledgements 165

Curriculum vitae 170

List of publications 171

Overvie w of completed training activities 172

Chapter 1

General introduction

Chapter 1

STRUCTURAL CLASSIFICATION AND BIOSYNTHESIS OF FLAVONOIDS

Flavonoids are a group of secondary metabolites that naturally occur throughout the plant [1] kingdom and are structurally characterised by a C 6-C3-C6 carbon framework . Flavonoids have gained increasing scientific interest due to their diverse functions in plant such as anti-microbial agents, photoreceptors, visual attractants, and feeding repellents. In addition, flavonoids are considered to provide health-promoting activity through a myriad of bioactivities including anti-oxidant, anti-inflammatory, anti-cancer and anti-allergic activities [2]. Flavonoids belong to the superfamily of phenyl benzopyrans, more specifically the term flavonoid refers to the 2-phenyl benzopyrans. Often isoflavonoids are also referred to as flavonoids, but actually this is a misnomer, as they are 3-phenyl benzopyrans. In this thesis, the term flavonoid will refer to the whole family of phenyl benzopyrans. The family of phenyl benzopyrans can be divided into 4 classes based on the position of the linkage of an aromatic ring to the benzopyran moiety (or chromene moiety, i.e. rings A and C): 2-phenyl benzopyrans, 3-phenyl benzopyrans (isoflavonoids), 4-phenyl benzopyrans (neoflavonoids) and miscellaneous flavonoids ( Figure 1 ). The miscellaneous flavonoids contain representatives in which the pyran C-ring (flavan core) is not retained. In these molecules, the C-ring is either open (e.g. chalcones), or ring-closed into a furan derivative (e.g. aurones). The first step in flavonoid biosynthesis is the conversion of phenylalanine via several steps into hydroxycinnamic acids, characterised by a C 6-C3 carbon framework. With the addition of 3 malonyl-CoA units, 2’,4’,6’,4-tetrahydroxychalcone can be formed from which all four phenyl benzopyran classes are formed [3]. Each of these four classes can be subdivided into subclasses based on the oxidation level of the C-ring and variations in the complexity of the carbon framework as a result of hydroxylation, alk(en)ylation (including methylation and prenylation), glycosylation and sulfonation. Flavonoids often occur as O-glycosides in which the sugar moiety is connected to the aromatic ring or the C- ring via a hydroxyl group with the formation of a glycosidic (acetal) bond. Occassionally, the glycosyl residue is linked to the aromatic ring by a C-C bond, an example of which is the isoflavone C-glycoside puerarin [4]. Glycosylation involves a broad range of sugars such as glucose and glucuronic acid, but also substituted sugars. For example, acetyl-glucosides and malonyl-glucosides are the predominant isoflavonoid conjugates found in soya. This thesis deals mainly with isoflavonoids from licorice roots ( Glycyrrhiza glabra ) and soya ( Glycine max ), two species belonging to the family of Leguminosae . Both sources are particularly rich in prenylated isoflavonoids with estrogenic activity. Therefore, isoflavonoids, prenylation, and estrogenic activity will be more extensively discussed below, together with background information on the two species and methods to enhance estrogenic activity.

2

Figure 1 . Flavonoid subdivision into 4 main classes based on the linkage of the aromatic B-ring to the benzopyran moiety.

Chapter 1

ISOFLAVONOIDS

The Leguminosae consist of more than 750 genera comprising 18,000 species, and is considered to be particularly rich in isoflavonoids [5]. Despite the fact that isoflavonoids are almost exclusively found in Leguminosae [5, 6], the class of isoflavonoids is characterised by large structural variations. These variations are due to different substitutions, different degrees of oxidation and the presence of extra heterocyclic rings. Hence, the isoflavonoids have been subdivided into 13 subclasses, based on oxidation level and the occurrence of additional heterocyclic rings [7]. Based on their proposed biosynthesis, α-methyldeoxybenzoins and 2-phenyl- benzofurans are included in the class of isoflavonoids. However, their deviating C-ring structures, a five-membered C-ring (2-phenyl-benzofurans) and open C-ring (α- methyldeoxybenzoins), suggest that these subclasses might also be categorised under the ‘miscellaneous flavonoids’. Six subclasses (excluding the 2-phenyl-benzofurans) have a typical 3-ring carbon frame with a 6-membered C-ring. In addition, 5 subclasses (pterocarpans, coumestans, rotenoids, coumaronochromones and coumaronochromenes) are characterised by an additional D-ring. It is thought that the isoflavones and pterocarpans are the largest two subclasses comprising ~40% and ~20%, respectively, of all molecular species of isoflavonoids [8]. The biosynthetic relationships of the isoflavonoid subclasses (including that of α-methyldeoxybenzoins and 2-phenyl-benzofurans) are illustrated in Figure 2 . The isoflavonoid subclasses share the same biosynthetic origin. They are derived from the chalcones chalconaringenin (2’,4’,6’,4-tetrahydroxychalcone) and isoliquiritigenin (2’,4’,4-trihydroxychalcone) that are converted into their corresponding flavanones. The subsequent migration of the aromatic ring from the 2-position to the 3-position of the flavanone, the so-called aryl rearrangement performed by isoflavone synthase, and water loss lead to the formation of the isoflavones genistein and daidzein, respectively ( Figure 3 ) [3, 7]. Daidzein and its precursor liquiritigenin differ from genistein and naringenin by the absence of the 5-hydroxyl group. Isoflavonoids lacking the 5-hydroxyl group are also called 5-deoxyisoflavonoids. From this perspective, the isoflavone biosynthesis consists of 2 parallel routes involving the formation of isoflavones and 5-deoxyisoflavones. All isoflavonoids are thought to originate from the main isoflavones daidzein and genistein, which are therefore considered key elements in the structural diversification of isoflavonoids [9].

4

Figure 2 . Biosynthetic relationships among several representative isoflavonoid subclasses, adapted from Dewick et al . [7]. The subclasses indicated in grey might also be classified under the Miscellaneous flavonoids.

Chapter 1

Soya beans are very rich in isoflavones, with a content ranging from 0.2 to 2 mg/g DW [10-13]. Another isoflavonoid-rich legume species is Glycyrrhiza glabra , or licorice. The isoflavonoid content of Glycyrrhiza glabra roots is similar or higher than that of soya beans [13]. The root content of the principal licorice isoflavonoid, glabridin, ranges from 0.5-5.7 mg/g DW [14-16]. Generally, also flavanones, chalcones, isoflavones, isoflavenes, and other isoflavans are abundantly present in the roots of Glycyrrhiza glabra [16].

Figure 3 . Biosynthetic conversion of (2S)-flavanones to isoflavones. 2HIS: 2-hydroxyisoflavanone synthase; 2HID: 2-hydroxyisoflavanone dehydratase; adapted from Veitch et al . [17].

Prenylation of isoflavonoids

Prenylation

Prenylation refers to substitution with a C 5-isoprenoid (or prenyl) group. In a more broad sense, prenylation also refers to the substitution with other groups, including C 10 - [18] isoprenoid (geranyl) or C 15 -isoprenoid (farnesyl) ( Figure 4 ) . Prenylation increases the lipophilicity of flavonoids, which results in an increased affinity to biological membranes and an improved interaction with target proteins [19]. Furthermore, it has been suggested that prenylation plays a crucial role in modulating the bioactivity of flavonoids [19, 20].

Representative examples of prenylated flavonoids are 8-prenyl naringenin, a C 5-prenyl substituted flavanone isolated from Humulus lupulus, and kurarinone, a C 10 -prenyl substituted flavanone isolated from Sophora flavescens . Both prenylated flavonoids exhibit strong bioactivities such as estrogenic activity [20-23].

6

General introduction

Figure 4. Representative overview of common prenyl-substituents in flavonoids. 1 7,8-(2,2-dimethylpyran) or pyran ring.

7

Chapter 1

Prenylated flavonoids are predominantly C-prenylated, whereas only a few studies reported O-prenylation [18]. The prenyl chain generally refers to the 3,3- dimethylallyl substituent (3,3-DMA), but also includes variations of this structure based on oxidation and hydroxylation. Hydroxylated prenyl groups are also called prenyl hydrates. A prenyl chain may undergo cyclisation with an ortho-phenolic hydroxyl group leading to 6- membered pyran derivatives. The common 2,2-dimethylchromeno substituent is often termed a pyran ring. Cyclisation of a prenyl chain may also result in the 5-membered furan derivative. C-prenylation at positions 6 and/or 8, as well as 3’ and / or 5’, appear to be [1] generally preferred . In this thesis, the term ‘prenyl’ will generally refer to C 5-isoprenoid group unless stated otherwise. Whereas most prenylated 2-phenyl benzopyrans appear to be constitutively synthesised compounds in plant tissue, most prenylated isoflavonoids are considered inducible metabolites [1, 18]. Prenylated isoflavonoids are often produced as part of the plant’s defensive strategy (see Germination under stressed conditions ). The majority of the prenylated isoflavonoids (~90 %) are restricted to the Leguminosae subfamily Papilionoideae [19, 20].

Prenyltransferases Prenyltransferases (PTs) are enzymes that catalyse the transfer of the dimethylallyl moiety from dimethylallyl pyrophosphate (DMAPP) to an acceptor molecule. PTs are distributed throughout all living organisms. Aromatic PTs (APTs) are enzymes that catalyse the substitution of an aromatic ring with a prenyl group. The APTs can be divided into three subgroups based on their respective substrates, being: quinones, flavonoids, and phenolic acids, of which the distribution of the flavonoid APTs is limited to plants [20]. Flavonoid APTs can be subdivided into soluble and membrane-bound enzymes, referring to whether the APT occurs in the cytosol, or attached to the membrane of e.g. the endoplasmatic reticulum and plastids [24]. Flavonoid APTs have been isolated from plants belonging to the families of Leguminosae , Moraceae , Cannabaceae, Euphorbiaceae , Umbelliferae , Guttiferae , Rutaceae and Compositae [20]. Within the Leguminosae , APTs have been characterised in Glycine max , Sophora flavescens , Phaseolus vulgaris and Lupinus albus , and all are membrane-bound [25-31]. The isoprenoid biosynthesis includes the mevalonate pathway localised in the cytosol and the methyl erythriol phosphate pathway (MEP) localised in the plastids. Regardless of the location of the APT (soluble or membrane- bound) or the type of aromatic molecule, it is assumed that the origin of the prenyl- moiety originates from the MEP pathway [20].

8

General introduction

IDENTIFICATION OF FLAVONOIDS BY MASS SPECTROMETRY

Mass spectrometry is based on the generation of charged molecular species. These ions are expressed in mass-to-charge ( m/z ) ratio and are plotted against their (relative) abundance, resulting in a mass spectrum. Ionisation of compounds can be established by a variety of different techniques. Ionisation techniques are generally classified as ‘hard’ or ‘soft’ ionisation [32, 33]. Flavonoids are generally studied with soft ionisation techniques such as electrospray ionisation (ESI) and atmospheric pressure chemical ionisation (APCI), but ESI is generally preferred over APCI due to its higher sensitivity [34, 35]. Soft ionisation involves either the protonation or deprotonation of the analyte. Protonation of an analyte is referred to as positive ion (PI) mode and results in a positively charged compound, generally indicated by [M+H] +. In the so-called negative ion (NI) mode, the analyte is deprotonated resulting in [M-H] -. NI mode is considered more sensitive and selective in flavonoid analysis [35-37]. Another method that provides structural information is the application of collision-induced dissociation (CID), resulting in the fragmentation of the charged analyte. The fragment-ions that are produced upon fragmentation of flavonoids can provide important structural information. Fragmentation of flavonoids is characterised by the retro -Diels Alder (RDA) reaction in which the C-ring of the flavonoid is cleaved. Cleavage of the C-ring results in A-ring and B-ring fragment ions that provide information on the number and type of substituents of these rings [36 , 37, 38]. The nomenclature used for C-ring fragment ions is based on the systematic nomenclature of carbohydrate fragmentation [38, 39]. A-ring and B-ring fragments are indicated by either i,j A+ and i,j B+ for PI mode fragmentation, or by i,j A- and i,j B- for NI mode fragmentation. The superscripts i,j represent the bonds that are cleaved in the C-ring, as indicated in Figure 5 . In this figure, the most common fragmentation pathways resulting from C-ring cleavages are presented for different subclasses of 2-phenyl benzopyrans, isoflavonoids and miscellaneous flavonoids. In addition to RDA fragments, neutral losses are also observed upon fragmentation. Neutral losses involve the loss of small molecules including H 2O (18), CO (28), CO 2 (44), but also sugar moieties such as glucosyl (162). Furthermore, the loss of 15 from the parent ion in both PI and NI mode corresponds to the loss of a methyl radical, indicating the presence of one or more O-methyl or methoxyl groups [40]. Both RDA fragments as well as neutral losses provide structural information of flavonoids [37].

9

Chapter 1

Figure 5 . Fragment ions (in either ionisation mode, ±) resulting from common C-ring cleavages of 2-phenyl benzopyrans and isoflavonoids. The left structure represents the fragmentation of chalcones. The middle and right panel represent the same structures (2-phenyl benzopyrans and isoflavonoids), but different fragmentation routes. The most common cleavages for 2-phenyl benzopyrans and 3-phenyl benzopyrans (isoflavonoids) are 0/2, 0/3, 1/2, 1/3 and 1/4.

Although the substitution pattern of flavonoids may influence the specific fragmentation pathway in NI and PI mode, some common fragmentation pathways have been described for several 2-phenyl benzopyrans and isoflavonoid subclasses. In Tables 1 and 2 an overview is given of fragment ions that are formed upon C-ring cleavage of different 2-phenyl benzopyrans and isoflavonoid subclasses in both PI and NI mode. The most common cleavage in both PI and NI mode is 1/3, which is observed in almost every studied subclass. Nevertheless, the fragmentation of flavonoids appears to follow more specific degradation pathways, characteristic for each subclass. These unique fragment ions, or combinations thereof, have been used for identification purposes [41]. The 2-phenyl benzopyran subclasses such as the flavones and flavonols have been most studied regarding their fragmentation pathways. Studies investigating the fragmentation of isoflavonoids have so far been limited to isoflavones, isoflavanones, isoflavans and pterocarpans. The fragmentation of these subclasses in PI mode typically results in 1,3 A+, 1,3 B+ and 2,3 B+ fragment ions. Their fragmentation in NI mode appears less straightforward, with several fragment ions for isoflavones and isoflavanones. The C-ring cleavage of isoflavones is often very limited or not observed in NI mode [42, 43]. Furthermore, a number of studies investigated the fragmentation of pterocarpans in NI mode, and coumestans in both PI and NI modes. To our knowledge, these studies reported no C-ring fragmentation [44-47].

10

General introduction

In general, it is not possible to discriminate 2-phenyl benzopyrans and isoflavonoids solely based on MS n fragmentation. Furthermore, the fragmentation of different 2-phenyl benzopyran and isoflavonoid subclasses in NI mode appears less distinctive compared to PI mode. In addition, the limited diversity of RDA fragments in PI mode might suggest the presence of isoflavonoids.

Table 1 . Fragment ions in PI mode resulting from C-ring cleavage reported for subclasses within 2-phenyl benzopyrans, isoflavonoids and miscellaneous flavonoids. 0/2 0/4 1/2 1/3 1/4 0 2 Remarks 2-Phenyl benzopyrans Flavones B B A,B Flavonols A,B (B) A B 1,4 A + 2H, 1,3 B - 2H Flavanones A,B A 1,3 B - 2H,

M+H-Bring Flavanols A A B Dihydroflavonols A B A A Isoflavonoids Isoflavones A,B Isoflavanones A,B 2,3 B 2,3 Isoflavans A,B B, M+H -Bring Pterocarpans A 2,3 B Miscellaneous flavonoids Chalcones A,B A,B Fragments in parenthesis indicate that these were fragments of minor abundance (<10%). This table has been compiled from data from the following references: [32, 35, 37, 38, 41, 43, 46, 48-53].

Table 2 . Fragment ions in NI mode resulting from C-ring cleavage reported for subclasses within 2- phenyl benzopyrans and isoflavonoids. 0/3 0/4 1/2 1/3 1/4 Remarks 2-phenyl benzopyrans Flavones A,B A* A A,B 1,4 A + 2H Flavonols A A,B A,B ** B* Flavanones A A* A A,B A Isoflavonoids Isoflavones B,(A) B A** A** Isoflavanones A,B Pterocarpans No RDA fragments were Coumestans observed Fragments in parenthesis indicate that these were fragments of minor abundance (<10%). * Ambiguous. ** Only observed for methoxylated derivatives. This table has been compiled from data from the following references: [32, 37, 41-45, 47, 54-58].

11

Chapter 1

INTERACTION OF FLAVONOIDS WITH THE HUMAN ESTROGEN RECEPTOR

Plant-derived compounds are known to induce biological responses by mimicking or modulating endogenous estrogens [59]. These so-called phytoestrogens usually exhibit their activity by binding to the estrogen receptor (ER), due to their structural similarity with the human sex hormone 17β-estradiol (E 2). Phytoestrogens possess a phenol ring, which together with another hydroxyl group at the other end of the molecule, is one of the most important prerequisites for binding to the ER [60, 61]. The spacing between the hydroxyl groups on both ends of the phytoestrogen is critical, and should resemble the [61] distance between the 3-hydroxyl and 17β-hydroxyl of E 2 . Many phytoestrogens belong to the isoflavonoid subclasses of isoflavones and coumestans, although also other phenolic compounds, such as lignans and stilbenes, are known to display estrogenic activity [62, 63]. The ER belongs to the superfamily of nuclear receptors. Its main function is the regulation of gene expression. The ER is activated by estrogens (C 18 -steroids) that control growth, differentiation and functions of many target organs. Two ER subtypes have been described in humans: the ERα and ERβ. The ERα is mainly expressed in the sex organs such as the endometrium, mammary gland and uterus. The ERβ is mainly distributed in the bone, cardiovascular system, central nervous system and the brain [59]. The most potent [64] hormone in the human body is 17β-estradiol . Estrogens like E 2 act as ligands binding to the ligand binding domain (LBD) of the receptor, upon which a dimeric configuration of the receptor is enabled, which facilitates binding to the estrogen response elements (ERE). Binding of the dimeric receptor complex to the ERE can initiate or inhibit DNA transcription, depending on whether the ER-ligand is an agonist or antagonist [60]. An agonist induces a conformational change that results in dimerisation of ERs initiating DNA transcription. E 2 is a well-known agonist. Antagonists, on the other hand, induce a conformational change that hampers dimerization, or the recruitment of so-called co- factors, resulting in inhibition of DNA transcription [60]. Currently, there is a broad range of in vitro bioassays available that allow the determination of estrogenic activity of a compound. There are different types of bioassays varying from ligand binding assays to receptor-dependent gene expression assays and cell proliferation assays. Whereas the ligand binding assay can only determine binding to the ER, receptor-dependent transactivation assays and proliferation assays can distinguish between agonists and antagonists [65]. Although in vitro bioassays help in determining the estrogenic potential of compounds, their physiological relevance needs to be confirmed in vivo . The binding affinity of ligands to a receptor is generally expressed as effective concentration at half maximal binding, or EC 50 value. Figure 6 illustrates typical binding curves of E 2 to the ERα and ER β, in which the EC 50 is indicated.

12

General introduction

Figure 6 . A typical binding curve of E 2 to estrogen receptors α and β.

Most phytoestrogens have a preferential binding to the ERβ. In general, their binding affinity is approx. 10 2 to 10 5 times weaker than that of endogenous estrogens, [66-68] resulting in an EC 50 in the µM range, depending on the assay type . Apart from binding to the ER, the estrogenic activity of phytoestrogens might also be mediated via other mechanisms such as the induction of sex hormone-binding globulin (SHBG) [69]. Furthermore, isoflavonoids have been reported to inhibit the activity of enzymes in the estrogen and androgen biosynthesis, i.e. aromatase, 5α-reductase and 17β-hydroxysteroid oxidoreductase [70-72]. Dietary intake of phytoestrogens might be beneficial in prevention or the treatment of hormone-dependent deficiencies including osteoporosis, cardiovascular diseases, chemo-prevention, (post-)menopausal symptoms [10, 12, 63, 67, 73-76]. As rich sources of phytoestrogens, legumes are the main contributor of the dietary intake of estrogenic compounds. The observation that the consumption of legumes was linked to health beneficial effects led to the development of health ingredients [77], which are nowadays used in the development of food products (functional foods) or dietary supplements (nutraceuticals). Both licorice and soya are legumes belonging to the family of the Leguminosae .

13

Chapter 1

LICORICE

Licorice is one of mankind’s oldest known medicines. Licorice refers to the dried underground parts (roots and stolons) of Glycyrrhiza ssp. and is also known as liquorice, radix Glycyrrhizae (Latin), gancao (‘sweet grass’ in Chinese), kanzou (Japanese) and gamcho ( 감초, Korean). For centuries, licorice has been applied in traditional herbal medicine and its origins can be traced back to Babylonia (Codex Hammurabi, ca. 1700 BC), Assyria (ca. 700 BC) and ancient China ( Shénnóng Běn Cǎo Jíng , Shennong's Materia Medica ) [78]. References of the medicinal application of licorice in East-Asia have been documented in the Shénnóng Běn Cǎo Jíng , the first Chinese dispensatory, where it is used for its life-enhancing properties, improving health and detoxification [79]. The Glycyrrhiza genus belongs to the legume family ( Leguminosae ) and contains 5 different species: G. glabra , G. uralensis , G. korshinskyi , G. inflata , and G. aspera [80]. G. glabra is considered to consist of 3 main varieties. G. glabra L. var. typica is commonly referred to as Spanish or Italian licorice. G. glabra L. var. violacea is known as Persian or Turkish licorice. G. glabra L. var. glandulifera is mainly known as Russian licorice [78, 81]. Confusion often occurs as the term licorice is applied for all species and varieties within the genus of Glycyrrhiza [82]. Several species ( G. glabra and G. uralensis ) have been reported to be distributed in the same geographical regions in Central Asia and Western China, and are able to form hybrids [14, 83]. In general, materials originating from G. glabra are used in the food and pharmaceutical industry. Nowadays, licorice is mainly applied in the tobacco, confectionary and pharmaceutical industries as a flavouring agent [16, 84]. One of the principal components of licorice root is glycyrrhizin ( Figure 7 ), a triterpenoid glycoside, which is considered 50-170 times sweeter than sucrose and primarily responsible for the licorice flavour and sweetness [16, 85]. Glycyrrhizin has long been considered as the bioactive constituent of licorice. However, it has been shown that many biological activities of licorice, including estrogenic, anti-cancer, anti-microbial, skin whitening and metabolic syndrome preventive, can be ascribed to its isoflavonoid constituents [15, 86-99]. In the past decades, extensive investigations have resulted in the identification of over 300 different flavonoids in Glycyrrhiza ssp., of which half were isolated from the underground parts [16].

14

General introduction

Licorice isoflavonoids

G. glabra is predominantly rich in isoflavans, isoflavenes and isoflavones. In addition, other flavonoids are abundantly present such as chalcones (miscellaneous flavonoids) and flavanones (2-phenyl benzopyrans). The chalcones and flavanones in licorice roots are represented by isoliquiritigenin and liquiritigenin, respectively, which occur as free aglycone form and as glycoside-conjugate. Furthermore, licorice is known to be a rich source of prenylated isoflavonoids. So far, studies have reported the isolation of ~75 prenylated isoflavonoids from the roots of G. glabra [16, 100-108]. Glabridin is the main prenylated isoflavonoid of licorice ( Figure 7 ), and is considered a species-specific marker compound for G. glabra [14, 82]. Glabridin is a prenylated isoflavan with a pyran-substitution on the A-ring. Its content in the dried roots varies from 0.08-0.35% (w/w) [109]. In addition to glabridin, several different glabridin-derivatives have been isolated from licorice, including hydroxylated, methoxylated and other prenylated derivatives [16]. In the past two decades, licorice has been investigated for its estrogenic activity. The estrogenic activity of the licorice extracts has been reported in different estrogenicity assays, implicating the presence of phytoestrogens [110-113]. The main estrogenic compounds that have been isolated from licorice root are glabridin and the isoflavene glabrene ( Figure 7 ) [114-118]. Glabrene is considered the most estrogenic compound with an

EC 50 of 1 µM, followed by glabridin with an EC 50 of 5 µM. Further investigations have shown that glabridin-derivatives were less estrogenic: hispaglabridin A and B, both B-ring prenyl-substituted glabridin-derivatives, were also estrogenic, but with an EC 50 of ~10 times higher than glabridin. Methylation of the hydroxyl groups on the B-ring of glabridin was shown to reduce the estrogenic properties of glabridin. The estrogenicity of glabrene and glabridin were determined using a mammalian cell based proliferation assay. However, their specific estrogenic potency on the ERα and ERβ, as well as possible antagonistic activities on both ER-subtypes, have not yet been established.

Figure 7 . Molecular structures of glycyrrhizin, the main licorice triterpenoid, glabridin and glabrene, the main prenylated isoflavonoids of licorice. GlcA, glucuronic acid.

15

Chapter 1

SOYA

Soya often refers to all species within the genus Glycine and belongs to the family of Leguminosae . The genus Glycine is divided into two subgenera, Glycine and Soja (Moench) F. J. Herm. The cultivated soya bean, Glycine max (L.) Merr., is derived from its wild ancestor Glycine soja . In this thesis the term soya refers to Glycine max unless stated otherwise. The soya bean is a complex matrix of proteins, oils, carbohydrates and bioactive compounds such as isoflavones, saponins, and phytosterols [119]. In Asia, soya forms an integral and important part of the diet. Soya beans are often processed resulting in a broad variety of food products. Common soya products are tofu (originating from China), tempeh (originating from Indonesia) and natto (originating from Japan). Soya beans are one of the richest sources of isoflavones. The isoflavone composition of soya is characterised by 3 main isoflavones, genistein, daidzein and glycitein, and their respective glucosides, acetyl-glucosides and malonyl-glucosides [120]. The isoflavone content can be affected by variety (cultivar), cultivation year, geographical location, temperature, storage conditions, and germination period [120-122]. The isoflavone content is highly variable and ranges between 0.2 and 2.0 mg/g dry weight, with occasional extremes up to 9.5 mg/g dry weight [123, 124]. With its high isoflavone content, soya is considered one of the richest dietary sources of phytoestrogens, and is associated with health-beneficial effects towards hormone-related conditions such as cancer- prevention, menopausal complaints and osteoporosis [10, 12, 63, 67, 73-76].

METHODS TO INCREASE ESTROGENIC BIOACTIVITY OF SOYA

Hydrolysis of isoflavone glycosides / Increasing bioavailability

The isoflavones in soya beans and soya foods are predominantly present in their glycosidic forms. Although isoflavone glycosides are more water soluble than the aglycones, in their conjugated form isoflavones are poorly absorbed in the intestinal tract [125]. Hydrolysis of the isoflavone glycoside is, therefore, essential in the bioavailability of isoflavones. Upon ingestion, isoflavone glycosides are hydrolysed by β-glycosidase of the brush border cells in the small intestine and of the intestinal microflora such as Lactobacillus spp., Bacteroides spp. and Bifidobacterium spp. [126]. Due to their stability, isoflavone glycosides remain intact during soya bean processing and, therefore, predominate in many soya foods [127]. Fermented soya foods such as miso, natto and tempeh, on the other hand, contain relatively high levels of isoflavone aglycones, due to bioconversion by microbial cultures used in fermentation [128-131]. Rhizopus spp., used in the production of tempeh, is for example mainly responsible for the conversion of isoflavone glycosides into their aglycones [132-134]. In relation to the enhanced bioavailability of isoflavone aglycones, the

16

General introduction

consumption of soya foods enriched with isoflavone aglycones is associated with reduced blood cholesterol and improves markers of cardiovascular risk [135]. Identification of different bacterial cultures capable of the hydrolysis of isoflavone glycosides stimulated the development of fermented functional foods that are enriched in isolavone aglycones. A common example is the enrichment of soymilk with isoflavone aglycones by fermentation with Bifidobacteria [136-138]. Several strains of the probiotic Bifidobacterium were capable of converting up to 90% of daidzin into its aglycone daidzein.

Enrichment of equol

For years, daidzein and genistein were considered the only estrogenic compounds that were responsible for the health benefits related to soya consumption. Studies on the microbial metabolism of soya isoflavones led to the discovery that daidzein can be converted into equol by intestinal microflora [139]. ER binding studies showed that equol is a more potent estrogen than daidzein, with a preferential binding affinity to the ERβ [140]. The intestinal conversion of equol was found to occur only in 25-30% of Western adults [141-143]. The prevalence of these so-called equol-producers is approximately twice as high among adults from Japan, Korea or China [144-147]. It has been hypothesised that the health benefits from soya consumption may be more pronounced in equol producers than in equol non-producers [139]. The additional health benefits for equol-producers led to the large-scale production of equol by chemical synthesis. However, this resulted in a racemic mixture of S-(-)equol and R-(+)equol as equol has a chiral carbon atom. S-(-)equol is the natural form that is produced by intestinal conversion, and is considered the most active form of both enantiomers [148, 149]. A few years ago, a specific bacterial strain ( Lactococcus garvieae ) was isolated from dairy products. This bacterium is capable of converting daidzein into S-(-) equol from daidzein-rich soya hypocotyls under controlled conditions [150]. This provided an alternative way to large-scale production of the natural S-(-)equol, and led to the development of new health ingredients and functional foods based on the enrichment of equol [151-153]. These equol-enriched health ingredients offer both equol-producers and non-producers the potential health benefits of equol.

17

Chapter 1

Figure 8 . The microbial conversion of daidzein to (S)-equol.

Germination

Germination or sprouting is another common process that is applied to different legumes, such as mung beans and soya beans, to enhance digestibility and the nutrient’s profile [154- 156]. Soya bean sprouts are a common vegetable mainly found in the diet of Asian counties, particularly in Korea ( kongnamul ; 콩나물), where soya bean sprouts are a common ingredient in soups, salads and side-dishes. Studies have shown that germination improves the nutritional quality by decreasing anti-nutritional factors such as trypsin inhibitors, phytates and undesirable (beany) flavour and odour caused by lipoxygenase [157-160], and increasing the levels of phytonutrients such as vitamins, phytosterols, saponins and tocopherols [160-164]. Germination is also known to induce the biosynthesis of bioactive phenolics and other phytonutrients as illustrated in e.g. broccoli and rye seedlings [155, 165, 166]. Their relatively high content of bioactive phytonutrients makes seedlings an interesting source for phytonutrient-rich extracts serving as a basis for developing new functional food ingredients. A summary of studies that have investigated the effects of soaking and germination on the isoflavone content of soya beans is shown in Table 3 . Soaking or imbibition of soya beans refers to the uptake of water during which the moisture content increases from 5-10% to 55-65% (w/w) [167]. Studies on the effect of soaking on the isoflavone content show contradictory results. In general, if soaking affected the isoflavone content, it resulted in either a moderate increase or decrease [159, 168-170]. Soaking generally leads to a decrease in isoflavone glycoside forms, which are hydrolysed by β-glucosidase [165, 169, 170]. Some studies suggested that the observed decrease in

18

General introduction

glycoside forms could also be due to leaching, i.e. the extraction of water-soluble isoflavone glycosides by the soaking water [169, 170]. The changes in isoflavone profile and content are more pronounced upon germination. Short-term germination (12-48 h) generally leads to a 9-57% increase of the total isoflavone content. This increase is usually followed by a decrease after prolonged germination ( Table 3 ). The changes in isoflavone profile are characterised by a decrease in glycoside forms (with the exception of malonyl-glucoside) and an increase in free aglycones. The malonyl-glucoside forms appear to dominate throughout in non-soaked, soaked and germinated beans, and often show an absolute increase. The absolute and relative content of aglycones appears to continuously increase during short and long-term germination. In general, short-term germination of soya beans is an effective method to increase the absolute isoflavone content by 10 to 60%. Moreover, both short and long- term germination lead to an increase of isoflavone aglycone content, thereby increasing the bioavailability of the soya isoflavones.

Table 3 . Overview of the effects of soaking, short-term and long-term germination on isoflavone content and profile of soya beans. The changes in isoflavone content (%) are compared to the non-soaked beans (100%). Soaking Short -term Long -term germination (12-48 h) germination (>48 h) Range of 4-20 h, 15 -50 °C 12 h to 8 days; 18 -30°C; 100% relative humidity conditions Bean/sprout Increas e moisture Development of main root and side -roots after 24 h; characteristics content from ca. Loss in dry weight ranging from 10% up to 40% after1-6 days 10% to max. 60% Isoflavone Slight increases to Increases ranging from Stabilisat ion or content slight decreases 9 to 57% decreases up to 40% Isoflavone Relative increase of Rel. incre ase of aglycone and decrease Inc rease of aglycones profile aglycone content of glycoside forms except malonyl-form; and decrease of other due to hydrolysis of Malonyl-forms predominate and forms glucoside forms continue to increase This table has been compiled from data of the following references: [119, 158-160, 165, 168-176].

Germination under stressed conditions

The exposure of plants and seedlings to pathogens can result in the mobilisation of defensive chemicals. This class of inducible secondary metabolites is called phytoalexins and are defined as low molecular weight, anti-microbial compounds that are both synthesized de novo by, and accumulated in, plants after exposure to micro-organisms [177]. Phytoalexin biosynthesis and accumulation are not only induced by micro-organisms (biotic elicitation), but also by other factors (abiotic elicitation) such as heavy metal salts, pesticides and UV irradiation, and by components of microbial origin, such as cell wall polysaccharides, proteins and fatty acid [178].

19

Chapter 1

One of the most studied legume species regarding phytoalexins is Glycine max . The main phytoalexins induced in soya are coumestrol (coumestans) and glyceollins (pterocarpans). The classification of coumestrol as a phytoalexin is rather ambiguous as coumestrol has also been reported in non-germinated soya beans [179, 180]. Glyceollins are a mixture of 6 different forms of which glyceollin I, II and III are predominantly produced in Glycine max (Figure 9 ) [181]. Glyceollin I and II are characterised by a pyran ring-substitution of the A-ring, whereas glyceollin III is substituted with an isopropenyl dihydrofuran on the

A-ring. Only trace amounts of glyceollin IV have once been isolated from CuCl 2-induced cotyledons of Glycine max [182]. Whereas glyceollins V (canescacarpin) and VI (clandestacarpin) have only been isolated from other members of the genus Glycine such as G. canescens , G. clandestina and G. tomentella , a recent study reported the isolation of trace amounts of glyceollin V (2 mg/kg DW) from the leaves of field grown plants of G. max [183]. The isoflavone daidzein is the main precursor of the glyceollin biosynthesis. Daidzein is converted into glycinol, which is subsequently substituted with a prenyl chain (originating from the MEP pathway) on either the 4 or 2 position, resulting in glyceollidin I and II, respectively [25]. Cyclisation of the prenyl chain of glyceollidin I leads to glyceollins I and VI, whereas cyclisation of glyceollidin II results in glyceollins II, III and glyceofuran. The formation of glyceollin IV involves methylation of the 3-hydroxyl group of glyceollidin II instead of cyclisation ( Figure 9 ). Both glyceollins and glycinol possess a broad range of anti-bacterial and antifungal activities against pathogenic micro-organisms, including the well-known soya pathogen Phytophthora megasperma f. sp. glycinea [184-186]. In the past decade, studies have shown that glyceollins also possess health-promoting activities such as anti-oxidant [187, 188], anti-cancer [189-191], anti-diabetic and anti-obesity activity [192]. However, particular interest goes out to the estrogenic activity of glyceollins. Due to their estrogenic activity, glycollins have been investigated in a number of in vitro and in vivo studies. The estrogenic activity of the glyceollin mixture was first reported a decade ago by means of a ligand binding assay [193]. Since its affinity for the ER was established, a number of studies demonstrated the in vitro and in vivo estrogenicity of the glyceollin mixture. Mixtures of glyceollins have been suggested to contain potent antagonists, which was demonstrated by the inhibitory effects of ER-mediated gene expression and proliferation of breast and ovarian tissue [189-191, 194, 195]. Moreover, it has been reported that glyceollins also show ER-mediated inhibition in the prostate cancer proliferation assay, indicating antagonistic activity towards prostate cancer [196]. Recently, glyceollin I has been identified as the active component of the glyceollin I-III mixture with an antagonistic activity that is somewhat different from the known ERα antagonists 4- hydroxytamoxifen and ICI 182.780 [194]. Similar to equol, glyceollin I can be chemically synthesised leading to a racemic mixture of (-)-glyceollin I and (+)-glyceollin I. Contrary to the natural form, (-)-glyceollin I, that possesses antagonistic activity, (+)-glyceollin I

20

General introduction

behaves as an agonist on the ERα [195]. Recently, the estrogenic activity of the non- prenylated glycinol has been reported [197]. Glycinol showed potent agonistic activity towards both ER-subtypes comparable to another potent phytoestrogen found in soya, coumestrol. The type of elicitation is of influence on the isoflavonoid content as well as the isoflavonoid profile. The differences between abiotic and biotic elicitation are summarised in Table 4 . In general, abiotic elicitation results in glyceollin levels ranging from trace amounts to 1528 µg/g FW. On average, biotic elicitation appears to induce higher levels of glyceollins ranging from 65 to 4327 µg/g FW. This may indicate that biotic elicitation is a more effective way of inducing glyceollins compared to non-biotic elicitation. Within the class of biotic elicitors, fungal elicitors appear to be the most effective inducers of high levels of glyceollins. In glyceollin research, the response of soya to the pathogenic fungus Phytophthora megasperma f. sp. glycinea is studied most, although the application of food grade fungi such as Rhizopus spp. and Aspergillus spp. have become popular more recently in view the development of novel food applications such as fermented soya bean yoghurt [198].

Table 4 . Induction and accumulation of glyceollins and other phytoalexins in soya bean seedlings or selected tissues categorised by the type and subtype of elicitation. Type of eliciation Total glyceollins I, II and Other compounds (µg/g References III (µg/g FW) FW) Biotic elicitation Bac teria 1 270 Glycinol: 60 -100 [181 , 185 , 199 , 200 ] Glyceollidins: 10 Glyceofuran: <10 Fungus - Phytophthora 65 -4327 n.d. [201 -204 ] megasperma f. sp. glycinea Fungus – other fungi 2 159 -3000 Coumestrol: 20 [188 , 189 , 192 , 205 -209 ] (1000-6800 3) Fungal cell wall 500 Glycinol: 34 [185 , 210 ] polysaccharides Abiotic elicitation UV radiation 4.3 Glycinol: 150 [185 , 211 ] Heav y metal salts 4 11 -42 Glyceollin IV [181 , 182 , 185 , 211 -213 ] Herbicides 5 79 -338 Glyceofuran: 16 [214 , 215 ] Miscellaneous chemicals 6 Traces - 1528 [181 , 211 , 216 ] 1 Include Erwinia carotovota , Pseudomonas fluorescens and Curtobacterium spp. 2 Include Fusarium spp., Rhizopus spp., Aspergillus spp, white rice yeast and Mucor ramosissimus . 3 1000-6800 µg/g DW reported for black soya bean cultivar [207]. 4 Heavy metals salts include CuCl 2, AgNO 3 and HgCl 2. 5 Diphenyl ether herbicides acifluorfen and lactofen. 6 Include hydroperoxides, sulfhydryl reagent (e.g. iodoacetate) and TX-100 detergent.

21

Figure 9 . Biosynthesis of glyceollins, involving the pterocarpan biosynthesis and MEP pathway. Glyceollin V* refers to the structure indicated as glyceollin V in Yuk et al . [217]

General introduction

Furthermore, a number of studies investigated the influence of tissue and type of elicitation on ratio of glyceollins I, II and III. In addition, other phytoalexins have been reported to accumulate, such as glycinol, glyceollidins I/II, glyceollin IV and glyceofuran [185, 199]. A summary on the effects of both tissue and elicitation types on the ratio of glyceollins I, II and III is presented in Table 5 . In the late 1980s, the effects of tissue and elicitor on the glyceollin I-III ratio were reviewed by Keen et al . [181]. It was concluded that not the elicitor but the soya tissue mostly determined the glyceollin ratio. Although some variation was observed, tissue-specific ratios could be observed. In leaf tissue, the ratio was approximately 2:1:3 to 1:7:11, indicating the predominance of glyceollin III. In the other tissues the ratio was characterised by more or less equal amounts of glyceollin II and III, and a 2 to 50 fold higher amount of glyceollin I relative to glyceollin II.

Table 5 . The ratio of glyceollins I, II and III in different soya tissues from Glycine max induced with different types of elicitors. Biotic Abiotic Tissue Fungi Cell wall poly - Misc. Heavy Misc. Ref . saccharides metal salts Seedling 11:1:1 2:1:2 [202 , 212 ] to 51:1:2 Hypocotyl 3:1:1 [202 , 218 ] to 15:2:1 Cotyledon 6:2:1 2:1:3 2:1:2 3:1:2 1 [181 , 186 , 202 , to to to 206, 219] 1:2:2 4:1:3 6:1:1 2 Leaves 1:1:3 1:2:3 3 1:3:5 1 [181 , 199 , 202 , to to to 215] 2:1:3 1:5:11 3 1:7:11 4 Roots 6:1:2 5 [181 ] Callus 17:1:2 [181 ] Pods 5:3:1 [212 ] Misc. – miscellaneous. 1 1 mM sodium iodoacetate. 2 0.2% sodium dodecyl sulfate. 3 Pseudomonas syringae pv pisi (bacterium). 4 Acifluorfen (herbicide). 5 Meloidogyne incognita (nematode).

The accumulating evidence that phytoalexins can promote health-beneficial effect in humans makes the induction process a potentially interesting method for enhancing the bioactivity of soya. With respect to functional food ingredients, phytoalexin-enriched soya seedlings could provide an alternative source of phytoestrogens in the development of new health-beneficial food ingredients.

23

Chapter 1

AIM AND OUTLINE OF THIS THESIS

In the field of functional foods and nutraceuticals, there is a constant need for the development of new and innovative health ingredients. Although phytoestrogen-rich extracts are one of the main health ingredients in this area, most of these ingredients are derived from a limited number of sources. The increasing knowledge on phytoestrogens and their structure-activity relationships provide an important basis for the diversification of phytoestrogen-rich extracts. Although prenylation has been suggested to play an important role in the modification of the structure-activity relationship of phytoestrogens, still relatively little is known regarding its effect on the estrogenic properties of isoflavonoids. In this thesis we aim (1) to provide a better structural characterisation of isoflavonoids, in particular the prenylated isoflavonoids occurring in soya and licorice, (2) to increase the estrogenic activity of soya beans by the application of malting in the presence of a food-grade fungus, (3) to correlate the agonistic/antagonistic estrogenicity of extracts, and fractions thereof, with the presence of prenylated isoflavonoids.

Chapter 1 presents an overview of prenyl-substituted flavonoid (sub-)classes in relation to their analysis by mass spectrometry and their estrogenic activity. Special attention is paid to methods that enhance the estrogenic potential of soya. In Chapter 2 , the development of an LC-MS screening method is described that allows the screening of prenylated flavonoids in complex mixtures such as plant extracts. An ethyl acetate extract from licorice root extract was used as a source of flavonoids. In Chapter 3, the prenyl screening developed is used as a tool to analyse for novel prenylated isoflavonoids that were induced in Rhizopus -challenged soya seedlings grown on lab-scale. Furthermore, the MS fragmentation behaviour in NI mode of particularly pterocarpans and coumestans is described. In Chapter 4, the licorice root extract was fractionated by centrifugal partitioning chromatography. The estrogenic potential of the fractions was determined towards both ER α and ER β, and some of the fractions were analysed for antagonistic activity. The induction of soya seedlings with Rhizopus microsporus performed on a larger scale with malting technology used in the brewing industry is described in Chapter 5 . The isoflavonoid content and composition, as well as the estrogenic potential towards both ER subtypes, were monitored over a 10 day period. The results of these chapters are discussed in Chapter 6 , including the implications for further development of novel phytoestrogen-rich extracts.

24

General introduction

REFERENCES

[1] R. K. Ibrahim, A forty-year journey in plant research: Original contributions to flavonoid biochemistry. Canadian Journal of Botany 2005 , 83 , 433-450. [2] F. Ververidis, E. Trantas, C. Douglas, G. Vollmer, G. Kretzschmar, N. Panopoulos, Biotechnology of flavonoids and other phenylpropanoid-derived natural products. Part I: Chemical diversity, impacts on plant biology and human health. Biotechnology Journal 2007 , 2, 1214-1234. [3] N. C. Veitch, Isoflavonoids of the Leguminosae. Natural Product Reports 2007 , 24 , 417-464. [4] J. K. Prasain, C. C. Wang, S. Barnes, Mass spectrometric methods for the determination of flavonoids in biological samples. Free Radical Biology and Medicine 2004 , 37 , 1324-1350. [5] M. Wink, G. I. A. Mohamed, Evolution of chemical defense traits in the Leguminosae: Mapping of distribution patterns of secondary metabolites on a molecular phylogeny inferred from nucleotide sequences of the rbcL gene. Biochemical Systematics and Ecology 2003 , 31 , 897-917. [6] R. Hegnauer, R. J. Grayer-Barkmeijer, Relevance of seed polysaccharides and flavonoids for the classification of the leguminosae: A chemotaxonomic approach. Phytochemistry 1993 , 34 , 3-16. [7] P. M. Dewick, in The flavonoids - Advances in research since 1986 (Ed.: J. B. Harborne), Chapman & Hall, London, 1986 , pp. 117-238. [8] T. Iwashina, The structure and distribution of the flavonoids in plants. Journal of Plant Research 2000 , 113 , 287-299. [9] S. Tahara, R. K. Ibrahim, Prenylated isoflavonoids - An update. Phytochemistry 1995 , 38 , 1073-1094. [10] W. Mazur, H. Adlercreutz, Naturally occurring oestrogens in food. Pure and Applied Chemistry 1998 , 70 , 1759-1776. [11] P. B. Kaufman, J. A. Duke, H. Brielmann, J. Boik, J. E. Hoyt, A comparative survey of leguminous plants as sources of the isoflavones, genistein and daidzein: Implications for human nutrition and health. Journal of Alternative and Complementary Medicine 1997 , 3, 7-12. [12] W. M. Mazur, J. A. Duke, K. Wähälä, S. Rasku, H. Adlercreutz, Isoflavonoids and lignans in legumes: Nutritional and health aspects in humans. Journal of Nutritional Biochemistry 1998 , 9, 193-200. [13] L. U. Thompson, B. A. Boucher, Z. Liu, M. Cotterchio, N. Kreiger, Phytoestrogen content of foods consumed in Canada, including isoflavones, lignans, and coumestan. Nutrition and Cancer 2006 , 54 , 184-201. [14] H. Hayashi, S. Hattori, K. Inoue, K. Sarsenbaev, M. Ito, G. Honda, Field survey of Glycyrrhiza plants in Central Asia (1). Characterization of G. uralensis , G. glabra and the putative intermediate collected in Kazakhstan. Biological and Pharmaceutical Bulletin 2003 , 26 , 867-871. [15] J. Vaya, P. A. Belinky, M. Aviram, Antioxidant constituents from licorice roots: Isolation, structure elucidation and antioxidative capacity toward LDL oxidation. Free Radical Biology and Medicine 1997 , 23 , 302-313. [16] T. Nomura, T. Fukai, Phenolic constituents of licorice (Glycyrrhiza species), Progress in the chemistry of organic natural products, Vol. 73 , Springer, 1998 . [17] N. C. Veitch, Isoflavonoids of the leguminosae. Natural Product Reports 2009 , 26 , 776-802. [18] D. Barron, R. K. Ibrahim, Isoprenylated flavonoids - A survey. Phytochemistry 1996 , 43 , 921-982. [19] B. Botta, A. Vitali, P. Menendez, D. Misiti, G. D. Monache, Prenylated flavonoids: Pharmacology and biotechnology. Current Medicinal Chemistry 2005 , 12 , 713-739. [20] K. Yazaki, K. Sasaki, Y. Tsurumaru, Prenylation of aromatic compounds, a key diversification of plant secondary metabolites. Phytochemistry 2009 , 70 , 1739-1745. [21] F. Roelens, N. Heldring, W. Dhooge, M. Bengtsson, F. Comhaire, J. Å. Gustafsson, E. Treuter, D. De Keukeleire, Subtle side-chain modifications of the hop phytoestrogen 8-prenylnaringenin result in distinct agonist/antagonist activity profiles for estrogen receptors α and β. Journal of Medicinal Chemistry 2006 , 49 , 7357-7365. [22] B. J. Doyle, G. B. Mahady, Phytomedicines for menopause. Drugs of the Future 2007 , 32 , 897-905. [23] P. Zanoli, M. Zavatti, Pharmacognostic and pharmacological profile of Humulus lupulus L. Journal of Ethnopharmacology 2008 , 116 , 383-396.

25

Chapter 1

[24] L. Heide, Prenyl transfer to aromatic substrates: genetics and enzymology. Current Opinion in Chemical Biology 2009 , 13 , 171-179. [25] T. Akashi, K. Sasaki, T. Aoki, S. I. Ayabe, K. Yazaki, Molecular cloning and characterization of a cDNA for pterocarpan 4-dimethylallyltransferase catalyzing the key prenylation step in the biosynthesis of glyceollin, a soybean phytoalexin. Plant Physiology 2009 , 149 , 683-693. [26] D. R. Biggs, R. Welle, F. R. Visser, H. Grisebach, Dimethylallylpyrophosphate:3,9-dihydroxypterocarpan 10-dimethylallyl transferase from Phaseolus vulgaris Identification of the reaction product and properties of the enzyme. FEBS Letters 1987 , 220 , 223-226. [27] K. Sasaki, K. Mito, K. Ohara, H. Yamamoto, K. Yazaki, Cloning and characterization of naringenin 8- prenyltransferase, a flavonoid-specific prenyltransferase of Sophora flavescens . Plant Physiology 2008 , 146 , 1075-1084. [28] U. Zähringer, E. Schaller, H. Grisebach, Induction of phytoalexin synthesis in soybean. Structure and reactions of naturally-occurring and enzymatically prepared prenylated pterocarpans from elicitor- treated cotyledons and cell-cultures of soybean. Zeitschrift Fur Naturforschung C-a 1981 , 36 , 234-241. [29] U. Zähringer, J. Ebel, L. J. Mulheirn, R. L. Lyne, H. Grisebach, Induction of phytoalexin synthesis in soybean - Dimethylallylpyrophosphate:trihydroxypterocarpan dimethylallyl transferase from elicitor- induced cotyledons. FEBS Letters 1979 , 101 , 90-92. [30] R. Welle, H. Grisebach, Properties and solubilization of the prenyltransferase of isoflavonoid phytoalexin biosynthesis in soybean. Phytochemistry 1991 , 30 , 479-484. [31] P. Laflamme, H. Khouri, P. Gulick, R. Ibrahim, Enzymatic prenylation of isoflavones in white lupin. Phytochemistry 1993 , 34 , 147-151. [32] W. M. A. Niessen, J. van der Greef, Liquid chromatography, mass spectrometry: principles and applications , Dekker, New York [etc.], 1992 . [33] Ø. M. Andersen, K. R. Markham, Flavonoids : chemistry, biochemistry, and applications , CRC, Taylor & Francis, Boca Raton, FL [etc.], 2006 . [34] E. De Rijke, H. Zappey, F. Ariese, C. Gooijer, U. A. T. Brinkman, Liquid chromatography with atmospheric pressure chemical ionization and electrospray ionization mass spectrometry of flavonoids with triple-quadrupole and ion-trap instruments. Journal of Chromatography A 2003 , 984 , 45-58. [35] R. Maul, N. H. Schebb, S. E. Kulling, Application of LC and GC hyphenated with mass spectrometry as tool for characterization of unknown derivatives of isoflavonoids. Analytical and Bioanalytical Chemistry 2008 , 391 , 239-250. [36] N. Fabre, I. Rustan, E. de Hoffmann, J. Quetin-Leclercq, Determination of flavone, flavonol, and flavanone aglycones by negative ion liquid chromatography electrospray ion trap mass spectrometry. Journal of the American Society for Mass Spectrometry 2001 , 12 , 707-715. [37] F. Cuyckens, M. Claeys, Mass spectrometry in the structural analysis of flavonoids. Journal of Mass Spectrometry 2004 , 39 , 1-15. [38] Y. L. Ma, Q. M. Li, H. van den Heuvel, M. Claeys, Characterization of flavone and flavonol aglycones by collision-induced dissociation tandem mass spectrometry. Rapid Communications in Mass Spectrometry 1997 , 11 , 1357-1364. [39] B. Domon, C. E. Costello, A systematic nomenclature for carbohydrate fragmentations in FAB-MS/MS spectra of glycoconjugates. Glycoconjugate Journal 1988 , 5, 397-409. [40] U. Justesen, Collision-induced fragmentation of deprotonated methoxylated flavonoids, obtained by electrospray ionization mass spectrometry. Journal of Mass Spectrometry 2001 , 36 , 169-178. [41] E. de Rijke, P. Out, W. M. A. Niessen, F. Ariese, C. Gooijer, U. A. T. Brinkman, Analytical separation and detection methods for flavonoids. Journal of Chromatography A 2006 , 1112 , 31-63. [42] Y. Wang, L. Yang, Y. Q. He, C. H. Wang, E. W. Welbeck, S. W. A. Bligh, Z. T. Wang, Characterization of fifty-one flavonoids in a Chinese herbal prescription Longdan Xiegan Decoction by high-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry and photodiode array detection. Rapid Communications in Mass Spectrometry 2008 , 22 , 1767-1778. [43] F. Kuhn, M. Oehme, F. Romero, E. Abou-Mansour, R. Tabacchi, Differentiation of isomeric flavone/isoflavone aglycones by MS 2 ion trap mass spectrometry and a double neutral loss of CO. Rapid Communications in Mass Spectrometry 2003 , 17 , 1941-1949.

26

General introduction

[44] M. Yang, W. Wang, J. Sun, Y. Zhao, Y. Liu, H. Liang, D. A. Guo, Characterization of phenolic compounds in the crude extract of Hedysarum multijugum by high-performance liquid chromatography with electrospray ionization tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2007 , 21 , 3833-3841. [45] R. Liu, M. Ye, H. Guo, K. Bi, D. A. Guo, Liquid chromatography/electrospray ionization mass spectrometry for the characterization of twenty-three flavonoids in the extract of Dalbergia odorifera . Rapid Communications in Mass Spectrometry 2005 , 19 , 1557-1565. [46] L. W. Qi, X. D. Wen, J. Cao, C. Y. Li, P. Li, L. Yi, Y. X. Wang, X. L. Cheng, X. X. Ge, Rapid and sensitive screening and characterization of phenolic acids, phthalides, saponins and isoflavonoids in Danggui Buxue Tang by rapid resolution liquid chromatography/diode-array detection coupled with time-of- flight mass spectrometry. Rapid Communications in Mass Spectrometry 2008 , 22 , 2493-2509. [47] D. Yang, S. Cai, H. Liu, X. Guo, C. Li, M. Shang, X. Wang, Y. Zhao, On-line identification of the constituents of Buyang Huanwu decoction in pig serum using combined HPLC-DAD-MS techniques. Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences 2006 , 831 , 288-302. [48] J. P. Antignac, R. Cariou, B. Le Bizec, J. P. Cravedi, F. Andre, Identification of phytoestrogens in bovine milk using liquid chromatography/electrospray tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2003 , 17 , 1256-1264. [49] G. C. Justino, C. M. Borges, M. Helena Florêncio, Electrospray ionization tandem mass spectrometry fragmentation of protonated flavone and flavonol aglycones: A re-examination. Rapid Communications in Mass Spectrometry 2009 , 23 , 237-248. [50] D. Tsimogiannis, M. Samiotaki, G. Panayotou, V. Oreopoulou, Characterization of flavonoid subgroups and hydroxy substitution by HPLC-MS/MS. Molecules 2007, 12 , 593-606. [51] B. Abad-García, S. Garmón-Lobato, L. A. Berrueta, B. Gallo, F. Vicente, A fragmentation study of dihydroquercetin using triple quadrupole mass spectrometry and its application for identification of dihydroflavonols in Citrus juices. Rapid Communications in Mass Spectrometry 2009 , 23 , 2785-2792. [52] B. Abad-García, L. A. Berrueta, S. Garmón-Lobato, B. Gallo, F. Vicente, A general analytical strategy for the characterization of phenolic compounds in fruit juices by high-performance liquid chromatography with diode array detection coupled to electrospray ionization and triple quadrupole mass spectrometry. Journal of Chromatography A 2009 , 1216 , 5398-5415. [53] R. March, J. Brodbelt, Analysis of flavonoids: Tandem mass spectrometry, computational methods, and NMR. Journal of Mass Spectrometry 2008 , 43 , 1581-1617. [54] B. Portet, N. Fabre, R. Rozenberg, J. L. Habib-Jiwan, C. Moulis, J. Quetin-Leclercq, Analysis of minor flavonoids in Piper hostmannianum var. berbicense using liquid chromatography coupled with atmospheric pressure chemical ionization mass spectrometry. Journal of Chromatography A 2008 , 1210 , 45-54. [55] W. Wu, C. Yan, L. Li, Z. Liu, S. Liu, Studies on the flavones using liquid chromatography-electrospray ionization tandem mass spectrometry. Journal of Chromatography A 2004 , 1047 , 213-220. [56] J. Kang, L. A. Hick, W. E. Price, A fragmentation study of isoflavones in negative electrospray ionization by MSn ion trap mass spectrometry and triple quadrupole mass spectrometry. Rapid Communications in Mass Spectrometry 2007 , 21 , 857-868. [57] F. Cuyckens, M. Claeys, Determination of the glycosylation site in flavonoid mono-O-glycosides by collision-induced dissociation of electrospray-generated deprotonated and sodiated molecules. Journal of Mass Spectrometry 2005 , 40 , 364-372. [58] G. Rak, P. Fodor, L. Abrankó, Three-step HPLC-ESI-MS/MS procedure for screening and identifying non-target flavonoid derivatives. International Journal of Mass Spectrometry 2010 , 290 , 32-38. [59] D. M. Harris, E. Besselink, S. M. Henning, V. L. W. Go, D. Heber, Phytoestrogens induce differential estrogen receptor alpha- or beta-mediated responses in transfected breast cancer cells. Experimental Biology and Medicine 2005 , 230 , 558-568. [60] A. C. W. Pike, Lessons learnt from structural studies of the oestrogen receptor. Best Practice and Research: Clinical Endocrinology and Metabolism 2006 , 20 , 1-14. [61] H. Fang, W. Tong, R. Perkins, A. M. Soto, N. V. Prechtl, D. M. Sheehan, Quantitative comparisons of in vitro assays for estrogenic activities. Environmental Health Perspectives 2000 , 108 , 723-729.

27

Chapter 1

[62] T. Cornwell, W. Cohick, I. Raskin, Dietary phytoestrogens and health. Phytochemistry 2004 , 65 , 995- 1016. [63] M. S. Kurzer, X. Xu, Dietary phytoestrogens. Annual Review of Nutrition 1997 , 17 , 353-381. [64] N. Heldring, A. Pike, S. Andersson, J. Matthews, G. Cheng, J. Hartman, M. Tujague, A. Ström, E. Treuter, M. Warner, J. Å. Gustafsson, Estrogen receptors: How do they signal and what are their targets. Physiological Reviews 2007 , 87 , 905-931. [65] T. F. H. Bovee, M. G. Pikkemaat, Bioactivity-based screening of antibiotics and hormones. Journal of Chromatography A 2009 , 1216 , 8035-8050. [66] K. R. Price, G. R. Fenwick, Naturally occurring oestrogens in foods - A review. Food Additives and Contaminants 1985 , 2, 73-106. [67] K. D. R. Setchell, Phytoestrogens: The biochemistry, physiology, and implications for human health of soy isoflavones. American Journal of Clinical Nutrition 1998 , 68 . [68] M. G. Glazier, M. A. Bowman, A review of the evidence for the use of phytoestrogens as a replacement for traditional estrogen replacement therapy. Archives of internal medicine 2001 , 161 , 1161-1172. [69] P. Cos, T. De Bruyne, S. Apers, D. Vanden Berghe, L. Pieters, A. J. Vlietinck, Phytoestrogens: Recent developments. Planta Medica 2003 , 69 , 589-599. [70] H. Adlercreutz, C. Bannwart, K. Wahala, T. Makela, G. Brunow, T. Hase, P. J. Arosemena, J. T. Kellis Jr, L. E. Vickery, Inhibition of human aromatase by mammalian lignans and isoflavonoid phytoestrogens. Journal of Steroid Biochemistry and Molecular Biology 1993 , 44 , 147-153. [71] B. A. J. Evans, K. Griffiths, M. S. Morton, Inhibition of 5 α-reductase in genital skin fibroblasts and prostate tissue by dietary lignans and isoflavonoids. Journal of Endocrinology 1995 , 147 , 295-302. [72] S. Mäkelä, M. Poutanen, M. L. Kostian, N. Lehtimäki, L. Strauss, R. Santti, R. Vihko, Inhibition of 17 β- hydroxysteroid oxidoreductase by flavonoids in breast and prostate cancer cells. Proceedings of the Society for Experimental Biology and Medicine 1998 , 217 , 310-316. [73] H. T. Adlercreutz, Phytoestrogens. State of the art. Environmental Toxicology and Pharmacology 1999 , 7, 201-207. [74] M. J. Messina, Legumes and soybeans: Overview of their nutritional profiles and health effects. American Journal of Clinical Nutrition 1999 , 70 , 439S-450S. [75] T. Usui, Pharmaceutical prospects of phytoestrogens. Endocrine Journal 2006 , 53 , 7-20. [76] D. M. Tham, C. D. Gardner, W. L. Haskell, Potential health benefits of dietary phytoestrogens: A review of the clinical, epidemiological, and mechanistic evidence. Journal of Clinical Endocrinology and Metabolism 1998 , 83 , 2223-2235. [77] S. M. Boué, T. E. Cleveland, C. Carter-Wientjes, B. Y. Shih, D. Bhatnagar, J. M. McLachlan, M. E. Burow, Phytoalexin-Enriched Functional Foods. Journal of Agricultural and Food Chemistry 2009 , 57 , 2614- 2622. [78] G. R. Fenwick, J. Lutomski, C. Nieman, Liquorice, Glycyrrhiza glabra L. - composition, uses and analysis. Food Chemistry 1990 , 38 , 119-143. [79] T. Nomura, T. Fukai, T. Akiyama, Chemistry of phenolic compounds of licorice ( Glycyrrhiza species) and their estrogenic and cytotoxic activities. Pure and Applied Chemistry 2002 , 74 , 1199-1206. [80] A. S. Ammosov, V. I. Litvinenko, Phenolic compounds of the genera Glycyrrhiza L. and Meristotropis Fisch. et Mey. (review). Pharmaceutical Chemistry Journal 2007 , 41 , 372-395. [81] M. Nassiri Asl, H. Hosseinzadeh, Review of pharmacological effects of Glycyrrhiza sp. and its bioactive compounds. Phytotherapy Research 2008 , 22 , 709-724. [82] K. Kondo, M. Shiba, R. Nakamura, T. Morota, Y. Shoyama, Constituent properties of licorices derived from Glycyrrhiza uralensis, G. glabra, or G. inflata identified by genetic information. Biological and Pharmaceutical Bulletin 2007 , 30 , 1271-1277. [83] H. Hayashi, S. L. Zhang, T. Nakaizumi, K. Shimura, M. Yamaguchi, K. Inoue, K. Sarsenbaev, M. Ito, G. Honda, Field survey of Glycyrrhiza plants in central asia (2). Characterization of phenolics and their variation in the leaves of Glycyrrhiza plants collected in Kazakhstan. Chemical and Pharmaceutical Bulletin 2003 , 51 , 1147-1152. [84] A. T. Leung, S. Foster, Encyclopedia of Common Natural Ingredients Used in Food, Drugs, and Cosmetics, Vol. 50 , Second edition, Springer New York, 1996 .

28

General introduction

[85] M. Blumenthal, The Complete German Commission E Monographs: Therapeutic Guide to Herbal Medicines , American Botanical Council, Austin, Texas, 1998 . [86] P. A. Belinky, M. Aviram, S. Mahmood, J. Vaya, Structural aspects of the inhibitory effect of glabridin on LDL oxidation. Free Radical Biology and Medicine 1998 , 24 , 1419-1429. [87] Z. Y. Wang, D. W. Nixon, Licorice and cancer. Nutrition and Cancer 2001 , 39 , 1-11. [88] M. M. Rafi, B. C. Vastano, N. Zhu, C. T. Ho, G. Ghai, R. T. Rosen, M. A. Gallo, R. S. DiPaola, Novel polyphenol molecule isolated from licorice root ( Glycyrrhiza glabra ) induces apoptosis, G2/M cell cycle arrest, and Bcl-2 phosphorylation in tumor cell lines. Journal of Agricultural and Food Chemistry 2002 , 50 , 677-684. [89] M. Watanabe, S. Hayakawa, M. Isemura, S. Kumazawa, T. Nakayama, C. Mori, T. Kawakami, Identification of licocoumarone as an apoptosis-inducing component in licorice. Biological and Pharmaceutical Bulletin 2002 , 25 , 1388-1390. [90] L. A. Mitscher, Y. H. Park, D. Clark, J. L. Beal, Antimicrobial agents from higher plants. Antimicrobial isoflavanoids and related substances from Glycyrrhiza glabra L. var. Typica . Journal of Natural Products 1980 , 43 , 259-269. [91] T. Fukai, A. Marumo, K. Kaitou, T. Kanda, S. Terada, T. Nomura, Antimicrobial activity of licorice flavonoids against methicillin-resistant Staphylococcus aureus . Fitoterapia 2002 , 73 , 536-539. [92] T. Hatano, Y. Shintani, Y. Aga, S. Shiota, T. Tsuchiya, T. Yoshida, Phenolic constituents of licorice. VIII. Structures of glicophenone and glicoisoflavanone, and effects of licorice phenolics on methicillin- resistant Staphylococcus aureus . Chemical and Pharmaceutical Bulletin 2000 , 48 , 1286-1292. [93] B. Fu, H. Li, X. Wang, F. S. C. Lee, S. Cui, Isolation and identification of flavonoids in licorice and a study of their inhibitory effects on tyrosinase. Journal of Agricultural and Food Chemistry 2005 , 53 , 7408- 7414. [94] O. Nerya, J. Vaya, R. Musa, S. Izrael, R. Ben-Arie, S. Tamir, Glabrene and isoliquiritigenin as tyrosinase inhibitors from licorice roots. Journal of Agricultural and Food Chemistry 2003 , 51 , 1201-1207. [95] F. Aoki, K. Nakagawa, M. Kitano, H. Ikematsu, K. Nakamura, S. Yokota, Y. Tominaga, N. Arai, T. Mae, Clinical safety of Licorice Flavonoid Oil (LFO) and pharmacokinetics of glabridin in healthy humans. Journal of the American College of Nutrition 2007 , 26 , 209-218. [96] F. Aoki, S. Honda, H. Kishida, M. Kitano, N. Arai, H. Tanaka, S. Yokota, K. Nakagawa, T. Asakura, Y. Nakai, T. Mae, Suppression by licorice flavonoids of abdominal fat accumulation and body weight gain in high-fat diet-induced obese C57BL/6J mice. Bioscience, Biotechnology and Biochemistry 2007 , 71 , 206-214. [97] H. Kamisoyama, K. Honda, Y. Tominaga, S. Yokota, S. Hasegawa, Investigation of the anti-obesity action of licorice flavonoid oil in diet-induced obese rats. Bioscience, Biotechnology and Biochemistry 2008 , 72 , 3225-3231. [98] Y. Tominaga, T. Mae, M. Kitano, Y. Sakamoto, H. Ikematsu, K. Nakagawa, Licorice flavonoid oil effects body weight loss by reduction of body fat mass in overweight subjects. Journal of Health Science 2006 , 52 , 672-683. [99] M. Kuroda, Y. Mimaki, S. Honda, H. Tanaka, S. Yokota, T. Mae, Phenolics from Glycyrrhiza glabra roots and their PPAR-γ ligand-binding activity. Bioorganic and Medicinal Chemistry 2010 , 18 , 962-970. [100] Y. Asada, W. Li, T. Yoshikawa, The first prenylated biaurone, licoagrone from hairy root cultures of Glycyrrhiza glabra . Phytochemistry 1999 , 50 , 1015-1019. [101] M. Baba, S. Sumi, N. Iwasaki, H. Kai, M. Sakano, Y. Okada, T. Okuyama, Studies of the Egyptian traditional folk medicines. IV. New isoflavonoid isolated from Egyptian licorice. Heterocycles 2008 , 75 , 3085-3089. [102] J. Bielenberg, Isoflavonoids from licorice root as mediators of anti-inflammatory and anti-allergic effects. Zeitschrift fur Phytotherapie 2001 , 22 , 289-293. [103] Y. W. Chin, H. A. Jung, Y. Liu, B. N. Su, J. A. Castoro, W. J. Keller, M. A. Pereira, A. D. Kinghorn, Anti- oxidant constituents of the roots and stolons of licorice ( Glycyrrhiza glabra ). Journal of Agricultural and Food Chemistry 2007 , 55 , 4691-4697. [104] T. Fukai, B. S. Cai, K. Maruno, Y. Miyakawa, M. Konishi, T. Nomura, An isoprenylated flavanone from Glycyrrhiza glabra and rec-assay of licorice phenols. Phytochemistry 1998 , 49 , 2005-2013.

29

Chapter 1

[105] T. Kinoshita, Y. Tamura, K. Mizutani, The isolation and structure elucidation of minor isoflavonoids from licorice of Glycyrrhiza glabra origin. Chemical and Pharmaceutical Bulletin 2005 , 53 , 847-849. [106] W. Li, Y. Asada, T. Yoshikawa, Flavonoid constituents from Glycyrrhiza glabra hairy root cultures. Phytochemistry 2000 , 55 , 447-456. [107] M. Shibano, Y. Matsumoto, E. Nakao, A. Henmi, G. Kusano, T. Shibata, Y. Hatakeyama, H. Hayashi, T. Kinoshita, Studies on index compounds for HPLC analysis of Glycyrrhiza glabra . Natural Medicines 1998 , 52 , 279-283. [108] E. M. Williamson, in Potter's Cyclopedia of Herbal Medicine , Saffron Walden, UK, 2003 , pp. 269-271. [109] H. Hayashi, S. Hattori, K. Inoue, O. Khodzhimatov, O. Ashurmetov, M. Ito, G. Honda, Field survey of Glycyrrhiza plants in central asia (3). Chemical characterization of G. glabra collected in Uzbekistan. Chemical and Pharmaceutical Bulletin 2003 , 51 , 1338-1340. [110] S. Dong, A. Inoue, Y. Zhu, M. Tanji, R. Kiyama, Activation of rapid signaling pathways and the subsequent transcriptional regulation for the proliferation of breast cancer MCF-7 cells by the treatment with an extract of Glycyrrhiza glabra root. Food and Chemical Toxicology 2007, 45 , 2470- 2478. [111] H. Kohno, K. Kouda, R. Tokunaga, Y. Sonoda, Detection of estrogenic activity in herbal teas by in vitro reporter assays. European Food Research and Technology 2007 , 225 , 913-920. [112] J. Liu, J. E. Burdette, H. Xu, C. Gu, R. B. Van Breemen, K. P. L. Bhat, N. Booth, A. I. Constantinou, J. M. Pezzuto, H. H. S. Fong, N. R. Farnsworth, J. L. Bolton, Evaluation of estrogenic activity of plant extracts for the potential treatment of menopausal symptoms. Journal of Agricultural and Food Chemistry 2001 , 49 , 2472-2479. [113] D. T. Zava, C. M. Dollbaum, M. Blen, Estrogen and Progestin Bioactivity of Foods, Herbs, and Spices. Experimental Biology and Medicine 1998 , 217 , 369-378. [114] S. Tamir, M. Eizenberg, D. Somjen, N. Stern, R. Shelach, A. Kaye, J. Vaya, Estrogenic and antiproliferative properties of glabridin from licorice in human breast cancer cells. Cancer Research 2000 , 60 , 5704-5709. [115] S. Tamir, M. Eizenberg, D. Somjen, S. Izrael, J. Vaya, Estrogen-like activity of glabrene and other constituents isolated from licorice root. Journal of Steroid Biochemistry and Molecular Biology 2001 , 78 , 291-298. [116] E. M. Choi, The licorice root derived isoflavan glabridin increases the function of osteoblastic MC3T3- E1 cells. Biochemical Pharmacology 2005 , 70 , 363-368. [117] D. Somjen, S. Katzburg, J. Vaya, A. M. Kaye, D. Hendel, G. H. Posner, S. Tamir, Estrogenic activity of glabridin and glabrene from licorice roots on human osteoblasts and prepubertal rat skeletal tissues. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 241-246. [118] D. Somjen, E. Knoll, J. Vaya, N. Stern, S. Tamir, Estrogen-like activity of licorice root constituents: Glabridin and glabrene, in vascular tissues in vitro and in vivo. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 147-155. [119] L. M. Paucar-Menacho, M. A. Berhow, J. M. G. Mandarino, E. G. de Mejia, Y. K. Chang, Optimisation of germination time and temperature on the concentration of bioactive compounds in Brazilian soybean cultivar BRS 133 using response surface methodology. Food Chemistry 2010 , 119 , 636-642. [120] H. J. Wang, P. A. Murphy, Isoflavone content in commercial soybean foods. Journal of Agricultural and Food Chemistry 1994 , 42 , 1666-1673. [121] S. Phommalth, Y. S. Jeong, Y. H. Kim, K. H. Dhakal, Y. H. Hwang, Effects of light treatment on isoflavone content of germinated soybean seeds. Journal of Agricultural and Food Chemistry 2008 , 56 , 10123- 10128. [122] J. A. Hoeck, W. R. Fehr, P. A. Murphy, G. A. Welke, Influence of genotype and environment on isoflavone contents of soybean. Crop Science 2000 , 40 , 48-51. [123] P. Seguin, R. Bodo, A. M. Al-Tawaha, Soybean isoflavones: factors affecting concentrations in seeds. , Research Signpost, Trivandrum, India, 2007 . [124] A. C. Eldridge, W. F. Kwolek, Soybean isoflavones: Effect of environment and variety on composition. Journal of Agricultural and Food Chemistry 1983 , 31 , 394-396. [125] K. D. R. Setchell, N. M. Brown, L. Zimmer-Nechemias, W. T. Brashear, B. E. Wolfe, A. S. Kirschner, J. E. Heubi, Evidence for lack of absorption of soy isoflavone glycosides in humans, supporting the crucial

30

General introduction

role of intestinal metabolism for bioavailability. American Journal of Clinical Nutrition 2002 , 76 , 447- 453. [126] T. E. Steer, I. T. Johnson, J. M. Gee, G. R. Gibson, Metabolism of the soyabean isoflavone glycoside genistin in vitro by human gut bacteria and the effect of prebiotics. British Journal of Nutrition 2003 , 90 , 635-642. [127] L. Coward, M. Smith, M. Kirk, S. Barnes, Chemical modification of isoflavones in soyfoods during cooking and processing. American Journal of Clinical Nutrition 1998 , 68 . [128] L. Coward, N. C. Barnes, K. D. R. Setchell, S. Barnes, Genistein, daidzein, and their beta-glycoside conjugates - Antitumor isoflavones in soybean foods from American and Asian diets. Journal of Agricultural and Food Chemistry 1993 , 41 , 1961-1967. [129] J. Ebata, Y. Fukuda, K. Hirai, K. Murata, β-Glucosidase involved in the antioxidant formation in tempe, fermented soybeans. J Agric Chem Soc Jpn 1972 , 46 , 323-329. [130] P. A. Murphy, Phytoestrogen content of processed soybean products. Food Technology 1982 , 36 , 60- 64. [131] H. J. Wang, P. A. Murphy, Mass Balance Study of Isoflavones during Soybean Processing. Journal of Agricultural and Food Chemistry 1996 , 44 , 2377-2383. [132] K. C. Cheng, J. T. Lin, J. Y. Wu, W. H. Liu, Isoflavone conversion of black soybean by immobilized rhizopus spp. Food Biotechnology 2010 , 24 , 312-331. [133] M. J. R. Nout, J. L. Kiers, Tempe fermentation, innovation and functionality: Update into the third millenium. Journal of Applied Microbiology 2005 , 98 , 789-805. [134] N. Nakajima, N. Nozaki, K. Ishihara, A. Ishikawa, H. Tsuji, Analysis of isoflavone content in tempeh, a fermented soybean, and preparation of a new isoflavone-enriched tempeh. Journal of Bioscience and Bioengineering 2005 , 100 , 685-687. [135] C. Clerici, K. D. R. Setchell, P. M. Battezzati, M. Pirro, V. Giuliano, S. Asciutti, D. Castellani, E. Nardi, G. Sabatino, S. Orlandi, M. Baldoni, O. Morelli, E. Mannarino, A. Morelli, Pasta Naturally Enriched with Isoflavone Aglycons from Soy Germ Reduces Serum Lipids and Improves Markers of Cardiovascular Risk. The Journal of Nutrition 2007 , 137 , 2270-2278. [136] S. Raimondi, L. Roncaglia, M. De Lucia, A. Amaretti, A. Leonardi, U. M. Pagnoni, M. Rossi, Bioconversion of soy isoflavones daidzin and daidzein by Bifidobacterium strains. Applied Microbiology and Biotechnology 2009 , 81 , 943-950. [137] D. Tsangalis, J. F. Ashton, A. E. J. McGill, N. P. Shah, Enzymic transformation of isoflavone phytoestrogens in soymilk by β-glucosidase-producing Bifidobacteria . Journal of Food Science 2002 , 67 , 3104-3113. [138] O. N. Donkor, N. P. Shah, Production of β-glucosidase and hydrolysis of isoflavone phytoestrogens by Lactobacillus acidophilus , Bifidobacterium lactis , and Lactobacillus casei in soymilk. Journal of Food Science 2008 , 73 , M15-M20 [139] K. D. Setchell, C. Clerici, Equol: history, chemistry, and formation. The Journal of Nutrition 2010 , 140 , 1355S-1362S [140] S. O. Mueller, S. Simon, K. Chae, M. Metzler, K. S. Korach, Phytoestrogens and their human metabolites show distinct agonistic and antagonistic properties on estrogen receptor α (ER α) and ER β in human cells. Toxicological Sciences 2004 , 80 , 14-25. [141] I. R. Rowland, H. Wiseman, T. A. Sanders, H. Adlercreutz, E. A. Bowey, Interindividual variation in metabolism of soy isoflavones and lignans: influence of habitual diet on equol production by the gut microflora. Nutrition and Cancer 2000 , 36 , 27-32. [142] C. Atkinson, C. L. Frankenfeld, J. W. Lampe, Gut bacterial metabolism of the soy isoflavone daidzein: exploring the relevance to human health. Experimental Biology and Medicine 2005 , 230 , 155-170. [143] J. W. Lampe, S. C. Karr, A. M. Hutchins, J. L. Slavin, Urinary Equol Excretion with a Soy Challenge: Influence of Habitual Diet (44241). Proceedings of the Society for Experimental Biology & Medicine 1998 , 217 , 335-339. [144] K. B. Song, C. Atkinson, C. L. Frankenfeld, T. Jokela, K. Wähälä, W. K. Thomas, J. W. Lampe, Prevalence of daidzein-metabolizing phenotypes differs between caucasian and korean american women and girls. The Journal of Nutrition 2006 , 136 , 1347-1351.

31

Chapter 1

[145] H. Akaza, N. Miyanaga, N. Takashima, S. Naito, Y. Hirao, T. Tsukamoto, T. Fujioka, M. Mori, W.-J. Kim, J. M. Song, A. J. Pantuck, Comparisons of Percent Equol Producers between Prostate Cancer Patients and Controls: Case-controlled Studies of Isoflavones in Japanese, Korean and American Residents. Japanese Journal of Clinical Oncology 2004 , 34 , 86-89. [146] Y. Arai, M. Uehara, Y. Sato, M. Kimira, A. Eboshida, H. Adlercreutz, S. Watanabe, Comparison of isoflavones among dietary intake, plasma concentration and urinary excretion for accurate estimation of phytoestrogen intake. Journal of Epidemiology 2000 , 10 , 127-135. [147] S. Watanabe, M. Yamaguchi, T. Sobue, Takahashi, Tosei, T. Miura, Y. Arai, W. Mazur, K. Wähälä, H. Adlercreutz, Pharmacokinetics of soybean isoflavones in plasma, urine and feces of men after ingestion of 60 g baked soybean powder (Kinako). The Journal of Nutrition 1998 , 128 , 1710-1715. [148] R. S. Muthyala, Y. H. Ju, S. Sheng, L. D. Williams, D. R. Doerge, B. S. Katzenellenbogen, W. G. Helferich, J. A. Katzenellenbogen, Equol, a natural estrogenic metabolite from soy isoflavones: convenient preparation and resolution of R- and S-equols and their differing binding and biological activity through estrogen receptors alpha and beta. Bioorganic & Medicinal Chemistry 2004 , 12 , 1559-1567. [149] K. D. R. Setchell, C. Clerici, E. D. Lephart, S. J. Cole, C. Heenan, D. Castellani, B. E. Wolfe, L. Nechemias- Zimmer, N. M. Brown, T. D. Lund, R. J. Handa, J. E. Heubi, S-equol, a potent ligand for estrogen receptor beta, is the exclusive enantiomeric form of the soy isoflavone metabolite produced by human intestinal bacterial flora. American Journal of Clinical Nutrition 2005 , 81 , 1072-1079. [150] M. G. Fortina, G. Ricci, R. Foschino, C. Picozzi, P. Dolci, G. Zeppa, L. Cocolin, P. L. Manachini , Journal of Applied Microbiology, 2007 , 103 , 445. [151] K. D. R. Setchell, X. Zhao, S. E. Shoaf, K. Ragland, The pharmacokinetics of S-(-)equol administered as SE5-OH tablets to healthy postmenopausal women. Journal of Nutrition 2009 , 139 , 2037-2043. [152] R. A. Matulka, I. Matsuura, T. Uesugi, T. Ueno, G. Burdock, Developmental and Reproductive Effects of SE5-OH: An Equol-Rich Soy-Based Ingredient. Journal of Toxicology 2009 , 1-13. [153] S. Yee, G. A. Burdock, Y. Kurata, Y. Enomoto, K. Narumi, S. Hamada, T. Itoh, Y. Shimomura, T. Ueno, Acute and subchronic toxicity and genotoxicity of SE5-OH, an equol-rich product produced by Lactococcus garvieae . Food and Chemical Toxicology 2008 , 46 , 2713-2720. [154] H. M. Bau, C. Villaume, J. P. Nicolas, L. Méjean, Effect of germination on chemical composition, biochemical constituents and antinutritional factors of soya bean ( Glycine max ) seeds. Journal of the Science of Food and Agriculture 1997 , 73 , 1-9. [155] K. Katina, K. H. Liukkonen, A. Kaukovirta-Norja, H. Adlercreutz, S. M. Heinonen, A. M. Lampi, J. M. Pihlava, K. Poutanen, Fermentation-induced changes in the nutritional value of native or germinated rye. Journal of Cereal Science 2007 , 46 , 348-355. [156] A. A. Khalil, Nutritional improvement of an Egyptian breed of mung bean by probiotic Lactobacilli . African Journal of Biotechnology 2006 , 5, 206-212. [157] K. Liu, Soybeans : chemistry, technology and utilization , Chapman and Hall, New York, 1997 . [158] U. K. Choi, M. H. Kim, N. H. Lee, Y. S. Jeong, O. J. Kwon, Y. C. Kim, Y. H. Hwang, The characteristics of cheonggukjang , a fermented soybean product, by the degree of germination of raw soybeans. Food Science and Biotechnology 2007 , 16 , 734-739. [159] D. Zhu, N. S. Hettiarachchy, R. Horax, P. Chen, Isoflavone contents in germinated soybean seeds. Plant Foods for Human Nutrition 2005 , 60 , 147-151. [160] H. Shi, P. K. Nam, Y. Ma, Comprehensive profiling of isoflavones, phytosterols, tocopherols, minerals, crude protein, lipid, and sugar during soybean ( Glycine max ) germination. Journal of Agricultural and Food Chemistry 2010 , 58 , 4970-4976. [161] S. B. Ryu, L. Zheng, X. Wang, Changes in phospholipase D expression in soybeans during seed development and germination. Journal of the American Oil Chemists' Society 1996 , 73 , 1171-1176. [162] A. Sattar, Neelofar, M. A. Akhtar, Irradiation and germination effects on phytate, protein and amino acids of soybean. Plant Foods for Human Nutrition 1990 , 40 , 185-194. [163] A. K. McGrain, J. C. Chen, K. A. Wilson, A. L. Tan-Wilson, Degradation of trypsin inhibitors during soybean germination. Phytochemistry 1989 , 28 , 1013-1017. [164] T. C. Jyothi, T. C. Sindhu Kanya, A. G. Appu Rao, Influence of germination on saponins in soybean and recovery of soy sapogenol I. Journal of Food Biochemistry 2007 , 31 , 1-13.

32

General introduction

[165] P. Y. Lin, H. M. Lai, Bioactive compounds in legumes and their germinated products. Journal of Agricultural and Food Chemistry 2006 , 54 , 3807-3814. [166] J. W. Fahey, Y. Zhang, P. Talalay, Broccoli sprouts: An exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proceedings of the National Academy of Sciences of the United States of America 1997 , 94 , 10367-10372. [167] D. H. Wardhani, J. A. Vázquez, S. S. Pandiella, Kinetics of daidzin and genistin transformations and water absorption during soybean soaking at different temperatures. Food Chemistry 2008 , 111 , 13-19. [168] P. H. Jeong, D. H. Shin, Y. S. Kim, Effects of germination and osmopriming treatment on enhancement of isoflavone contents in various soybean cultivars and Cheonggukjang (fermented unsalted soybean paste). Journal of Food Science 2008 , 73 . [169] S. P. Góes-Favoni, M. C. Carrão-Panizzi, A. Beleia, Changes of isoflavone in soybean cotyledons soaked in different volumes of water. Food Chemistry 2010 , 119 , 1605-1612. [170] T. Toda, A. Sakamoto, T. Takayanagi, K. Yokotsuka, Changes in Isoflavone Compositions of Soybean during Soaking in Water. Food Science and Technology Research 2001 , 7, 171-175. [171] L. A. Rosenberg, R. W. Rinne, Changes in seed constituents during germination and seedling growth of precociously matured soybean seeds ( Glycine max ). Annals of Botany 1987 , 60 , 705-712. [172] W. J. Kim, H. Y. Lee, M. H. Won, S. H. Yoo, Germination effect of soybean on its contents of isoflavones and oligosaccharides. Food Science and Biotechnology 2005 , 14 , 498-502. [173] J. S. Han, H. D. Hong, S. R. Kim, Changes in isoflavone content and mass balance during soybean processing. Food Science and Biotechnology 2007 , 16 , 426-433. [174] U. K. Choi, Y. S. Jeong, M. H. Kim, N. H. Lee, Y. H. Hwang, Quality characteristics of meju according to germination time of raw soybean ( Glycine max : Hwanggeumkong ). Food Science and Biotechnology 2007 , 16 , 386-391. [175] J. S. Kim, J. G. Kim, W. J. Kim, Changes in isoflavone and oligosaccharides of soybeans during germination. Korean Journal of Food Science and Technology 2004 , 36 , 294-298. [176] M. D. Chiarello, J. L. Le Guerroué, C. M. S. Chagas, O. L. Franco, E. Bianchini, M. J. João, Influence of heat treatment and grain germination on the isoflavone profile of soy milk. Journal of Food Biochemistry 2006 , 30 , 234-247. [177] F. D. Dakora, D. A. Phillips, Diverse functions of isoflavonoids in legumes transcend anti-microbial definitions of phytoalexins. Physiological and Molecular Plant Pathology 1996 , 49 , 1-20. [178] J. A. Bailey, J. W. Mansfield, Phytoalexins , Blackie, Glasgow, 1982 . [179] G. L. Lookhart, P. L. Finney, K. F. Finney, Coumestrol in soybeans and fractions at various germination times. Cereal Chemistry 1979 , 56 , 495-496. [180] G. L. Lookhart, B. L. Jones, K. F. Finney, Determination of coumestrol in soybeans by high-performance liquid chromatography and thin-layer chromatography. Cereal Chemistry 1978 , 55 , 967-972. [181] N. T. Keen, R. L. Lyne, T. Hymowitz, Phytoalexin Production as a chemosystematic parameter within the genus Glycine . Biochemical Systematics and Ecology 1986 , 14 , 481-486. [182] R. L. Lyne, L. J. Mulheirn, Minor pterocarpinoids of soybean. Tetrahedron Letters 1978 , 19 , 3127-3128. [183] H. J. Yuk, J. H. Lee, M. J. Curtis-Long, J. W. Lee, Y. S. Kim, H. W. Ryu, C. G. Park, T.-S. Jeong, K. H. Park, The most abundant polyphenol of soy leaves, coumestrol, displays potent [alpha]-glucosidase inhibitory activity. Food Chemistry 2011 , 126 , 1057-1063. [184] M. K. Bhattacharyya, E. W. B. Ward, Differential sensitivity of Phytophthora megasperma f.sp. glycinea isolates to glyceollin isomers. Physiological Plant Pathology 1985 , 27 , 299-310. [185] L. I. Weinstein, M. G. Hahn, P. Albersheim, Host-pathogen interactions XVIII: Isolation and biological- activity of glycinol, a pterocarpan phytoalexin synthesized by soybeans. Plant Physiology 1981 , 68 , 358-363. [186] W. S. Garcez, D. Martins, F. R. Garcez, M. R. Marques, A. A. Pereira, L. A. Oliveira, J. N. Rondon, A. D. Peruca, Effect of spores of saprophytic fungi on phytoalexin accumulation in seeds of frog-eye leaf spot and stem canker-resistant and -susceptible soybean ( Glycine max L.) cultivars. Journal of Agricultural and Food Chemistry 2000 , 48 , 3662-3665. [187] H. J. Kim, H.-J. Suh, J. H. Kim, S. Park, Y. C. Joo, J.-S. Kim, Antioxidant activity of glyceollins derived from soybean elicited with Aspergillus sojae . Journal of Agricultural and Food Chemistry 2010 , 58 , 11633-11638.

33

Chapter 1

[188] S. M. Boué, F. F. Shih, B. Y. Shih, K. W. Daigle, C. H. Carter-Wientjes, T. E. Cleveland, Effect of biotic elicitors on enrichment of antioxidant properties and induced isoflavones in soybean. Journal of Food Science 2008 , 73 , H43-H49. [189] M. E. Burow, S. M. Boué, B. M. Collins-Burow, L. I. Melnik, B. N. Duong, C. H. Carter-Wientjes, S. Li, T. E. Wiese, T. E. Cleveland, J. A. McLachlan, Phytochemical glyceollins, isolated from soy, mediate antihormonal effects through estrogen receptor α and β. Journal of Clinical Endocrinology and Metabolism 2001 , 86 , 1750-1758. [190] V. A. Salvo, S. M. Boué, J. P. Fonseca, S. Elliott, C. Corbitt, B. M. Collins-Burow, T. J. Curiel, S. K. Srivastav, B. Y. Shih, C. Carter-Wientjes, C. E. Wood, P. W. Erhardt, B. S. Beckman, J. A. McLachlan, T. E. Cleveland, M. E. Burow, Antiestrogenic glyceollins suppress human breast and ovarian carcinoma tumorigenesis. Clinical Cancer Research 2006 , 12 , 7159-7164. [191] C. E. Wood, T. B. Clarkson, S. E. Appt, A. A. Franke, S. M. Boue, M. E. Burow, T. McCoy, J. M. Cline, Effects of soybean glyceollins and estradiol in postmenopausal female monkeys. Nutrition and Cancer 2006 , 56 , 74-81. [192] S. Park, I. S. Ahn, J. H. Kim, M. R. Lee, J. S. Kim, H. J. Kim, Glyceollins, one of the phytoalexins derived from soybeans under fungal stress, enhance insulin sensitivity and exert insulinotropic actions. Journal of Agricultural and Food Chemistry 2010 , 58 , 1551-1557. [193] G. N. Nikov, N. E. Hopkins, S. Boue, W. L. Alworth, Interactions of dietary estrogens with human estrogen receptors and the effect on estrogen receptor-estrogen response element complex formation. Environmental Health Perspectives 2000 , 108 , 867-872. [194] M. C. Zimmermann, S. L. Tilghman, S. M. Boué, V. A. Salvo, S. Elliott, K. Y. Williams, E. V. Skripnikova, H. Ashe, F. Payton-Stewart, L. Vanhoy-Rhodes, J. P. Fonseca, C. Corbitt, B. M. Collins-Burow, M. H. Howell, M. Lacey, B. Y. Shih, C. Carter-Wientjes, T. E. Cleveland, J. A. McLachlan, T. E. Wiese, B. S. Beckman, M. E. Burow, Glyceollin I, a novel antiestrogenic phytoalexin isolated from activated soy. Journal of Pharmacology and Experimental Therapeutics 2010 , 332 , 35-45. [195] F. Payton-Stewart, R. S. Khupse, S. M. Boué, S. Elliott, M. C. Zimmermann, E. V. Skripnikova, H. Ashe, S. L. Tilghman, B. S. Beckman, T. E. Cleveland, J. A. McLachlan, D. Bhatnagar, T. E. Wiese, P. Erhardt, M. E. Burow, Glyceollin I enantiomers distinctly regulate ER-mediated gene expression. Steroids 2010 , 75 , 870-878. [196] F. Payton-Stewart, N. W. Schoene, Y. S. Kim, M. E. Burow, T. E. Cleveland, S. M. Boue, T. T. Wang, Molecular effects of soy phytoalexin glyceollins in human prostate cancer cells LNCaP. Molecular Carcinogenesis 2009 , 48 , 862-871. [197] S. M. Boué, S. L. Tilghman, S. Elliott, M. C. Zimmerman, K. Y. Williams, F. Payton-Stewart, A. P. Miraflor, M. H. Howell, B. Y. Shih, C. H. Carter-Wientjes, C. Segar, B. S. Beckman, T. E. Wiese, T. E. Cleveland, J. A. McLachlan, M. E. Burow, Identification of the potent phytoestrogen glycinol in elicited soybean ( Glycine max ). Endocrinology 2009 , 150 , 2446-2453. [198] S. Feng, C. L. Saw, Y. K. Lee, D. Huang, Novel process of fermenting black soybean [ Glycine max (L.) Merrill] yogurt with dramatically reduced flatulence-causing oligosaccharides but enriched soy phytoalexins. Journal of Agricultural and Food Chemistry 2008 , 56 , 10078-10084. [199] J. L. Ingham, N. T. Keen, L. J. Mulheirn, R. L. Lyne, Inducibly-formed isoflavonoids from leaves of soybean. Phytochemistry 1981 , 20 , 795-798. [200] B. Ramos-Solano, E. Algar, A. García-Villaraco, J. García-CristóBal, J. Antonio Lucas García, F. J. Gutierrez-Mañero, Biotic elicitation of isoflavone metabolism with plant growth promoting rhizobacteria in early stages of development in Glycine max var. osumi . Journal of Agricultural and Food Chemistry 2010 , 58 , 1484-1492. [201] M. G. Hahn, A. Bonhoff, H. Grisebach, Quantitative localization of the phytoalexin glyceollin I in relation to fungal hyphae in soybean roots infected with Phytophthora megasperma f. sp. glycinea . Plant Physiology 1985 , 77 , 591-601. [202] M. K. Bhattacharyya, E. W. B. Ward, Resistence, susceptibility and accumulation of glyceollins I-III in soybean organs inoculated with Phytophthora megasperma f. sp. glycinea . Physiological and Molecular Plant Pathology 1986 , 29 , 227-237.

34

General introduction

[203] E. W. B. Ward, R. I. Buzzell, Influence of light, temperature and wounding on the expression of soybean genes for resistence to Phytophthora megasperma f. sp. glycinea . Physiological Plant Pathology 1983 , 23 , 401-409. [204] C. B. Bloch, P. J. G. M. De Wit, J. Kuć, Elicitation of phytoalexins by arachidonic and eicosapentaenoic acids: a host survey. Physiological Plant Pathology 1984 , 25 , 199-208. [205] V. V. Lozovaya, A. V. Lygin, O. V. Zernova, S. Li, G. L. Hartman, J. M. Widholm, Isoflavonoid accumulation in soybean hairy roots upon treatment with Fusarium solani . Plant Physiology and Biochemistry 2004 , 42 , 671-679. [206] S. M. Boué, C. H. Carter, K. C. Ehrlich, T. E. Cleveland, Induction of the soybean phytoalexins coumestrol and glyceollin by Aspergillus . Journal of Agricultural and Food Chemistry 2000 , 48 , 2167- 2172. [207] S. Feng, L. S. Chin, K. L. Yuan, D. Huang, Fungal-stressed germination of black soybeans leads to generation of oxooctadecadienoic acids in addition to glyceollins. Journal of Agricultural and Food Chemistry 2007 , 55 , 8589-8595. [208] H. J. Kim, H. J. Suh, J. H. Kim, S. C. Kang, S. Park, C. H. Lee, J. S. Kim, Estrogenic activity of glyceollins isolated from soybean elicited with Aspergillus sojae . Journal of Medicinal Food 2010 , 13 , 382-390. [209] V. V. Lozovaya, A. V. Lygin, S. Li, G. L. Hartman, J. M. Widholm, Biochemical response of soybean roots to Fusarium solani f. sp. glycines infection. Crop Science 2004 , 44 , 819-826. [210] D. S. Park, S. Landini, M. Y. Graham, T. L. Graham, Induced distal defence potentiation against Phytophthora sojae in soybean. Physiological and Molecular Plant Pathology 2002 , 60 , 293-310. [211] T. L. Graham, M. Y. Graham, Glyceollin elicitors induce major but distinctly different shifts in isoflavonoid metabolism in proximal and distal soybean cell-populations. Molecular Plant-Microbe Interactions 1991 , 4, 60-68. [212] S. W. Banks, P. M. Dewick, Biosynthesis of glyceollins I, II and III in soybean. Phytochemistry 1983 , 22 , 2729-2733. [213] C. Werner, H. R. Hohl, The effects of Ca 2+ and Mg 2+ on accumulation and secretion of isoflavonoids by soybean roots. Plant Science 1990 , 72 , 181-191. [214] S. Landini, M. Y. Graham, T. L. Graham, Lactofen induces isoflavone accumulation and glyceollin elicitation competency in soybean. Phytochemistry 2003 , 62 , 865-874. [215] T. Kömives, J. E. Casida, Acifluorfen increases the leaf content of phytoalexins and stress metabolites in several crops. Journal of Agricultural and Food Chemistry 1983 , 31 , 751-755. [216] N. Degousee, C. Triantaphylides, J. L. Montillet, Involvement of oxidative processes in the signaling mechanisms leading to the activation of glyceollin synthesis in soybean ( Glycine max ). Plant Physiology 1994 , 104 , 945-952. [217] H. J. Yuk, J. H. Lee, M. J. Curtis-Long, J. W. Lee, Y. S. Kim, H. W. Ryu, C. G. Park, T. S. Jeong, K. H. Park, The most abundant polyphenol of soy leaves, coumestrol, displays potent α-glucosidase inhibitory activity. Food Chemistry 2011 , 126 , 1057-1063. [218] P. Moesta, H. Grisebach, Investigation of the mechanism of glyceollin accumulation in soybean infected by Phytophthora megasperma f. sp. glycinea . Archives of Biochemistry and Biophysics 1981 , 212 , 462-467. [219] R. L. Lyne, L. J. Mulheirn, D. P. Leworthy, New pterocarpinoid phytoalexins of soybean. Journal of the Chemical Society, Chemical Communications 1976 , 497-498.

35

Chapter 2

A rapid screening method for prenylated flavonoids with UHPLC-ESI-MS in licorice root extracts

Based on : Rudy Simons, Jean-Paul Vincken, Edwin J. Bakx, Marian A. Verbruggen, Harry Gruppen, A rapid screening method for prenylated flavonoids with UHPLC-ESI-MS in licorice root extracts, Rapid Communications in Mass Spectrometry 2009 , 23, 3083-3093 Reprinted with permission

Chapter 2

ABSTRACT

Due to their substitution with an isoprenoid group, prenylated flavonoids have an increased affinity for biological membranes and target proteins, enhancing their potential bioactivity. Although many prenylated flavonoids have been described, there are no methods that specifically screen for their presence in complex mixtures, prior to purification. We describe a method based on ultra-high performance liquid chromatography (UHPLC) with electrospray ionisation mass spectrometry (ESI-MS) that allows rapid screening for prenylated flavonoids in multi-component plant extracts. Identification of the prenylated flavonoids is based on screening for neutral losses of 42 and 56 in the positive-ion mode MS 2 and MS 3 spectra within the MS chromatograms. In addition, this method discriminates between prenyl chain and ring-closed prenyl (pyran ring), based on the ratio of the relative abundances of the ions that lose 42 and 56 (42:56). The application of this screening method on a 70% aq. ethanol, ethanol and ethyl acetate extract of the roots of Glycyrrhiza glabra indicated the presence of 70 mono and di- prenylated flavonoids. In addition, of each prenylated flavonoid the type of prenylation, chain or pyran ring, was determined.

Keywords : Prenylation, isoflavonoids, licorice, mass spectrometry, screening

38

LC-MS based screening method for prenylated flavonoids in licorice

INTRODUCTION

Flavonoids are a class of naturally occurring compounds ubiquitous in plants, with a broad variety of bioactivities [1]. Flavonoids can be decorated with one or more C5 isoprenoid substituents, and these are referred to as prenylated flavonoids [2-4]. Prenyl groups can be attached to a flavonoid in the form of a prenyl chain (3-methyl-2-butenyl or dimethylallyl) or in the ring-closed form of a pyran ring (2,2-dimethylpyran). In general, most flavonoids are C-prenylated, whereas O-prenylation is less common [5]. Prenylation increases the lipophilicity of flavonoids leading to an increased affinity to biological membranes and to an improved interaction with target proteins. This results in enhanced biological activities such as anti-microbial and estrogenic activities [5, 6]. Examples of prenylated flavonoids with estrogenic activities are 8-prenylnaringenin and xanthohumol from hop ( Humulus lupulus ) [7, 8]. Prenylated compounds seem to offer interesting possibilities for developing new food supplements, e.g. for the alleviation of osteoporosis and menopausal complaints, which are both under hormonal control. In the past two decades, many prenylated flavonoids were reported, mainly in plants belonging to the Leguminosae and Moraceae [2]. Most of these flavonoids have been structurally elucidated by means of NMR spectroscopy and mass spectrometry upon their isolation and purification. A screening method to directly detect prenylated flavonoids in multicomponent samples is unavailable until now, but might offer opportunities in facilitating the discovery of potential bioactivities. Flavonoids have been extensively investigated with electrospray ionisation (ESI) and atmospheric pressure chemical ionisation (APCI), both in negative-ion (NI) and in positive-ion (PI) mode, each with their own advantages. In general, NI mode is considered more sensitive and selective for the analysis of flavonoids [9]. An essential part of the identification of flavonoids by MS is the fragmentation of the deprotonated molecules in MS n, in which the flavonoid is degraded into fragment ions. So far, there are no clear indications for the degradation of the prenyl group of prenylated flavonoids in the MS 2 spectra in NI mode. This supports the hypothesis that due to the nucleophilic nature of the prenyl-group degradation is highly unfavourable [10]. Licorice refers to the roots from Glycyrrhiza glabra and has been traditionally used in herbal medicine for over 4000 years [11, 12]. In the past decades, extensive chemical studies resulted in the identification of many flavonoids in the roots of Glycyrrhiza glabra , including about 75 prenylated flavonoids [13-22]. The aim of the present study was to develop an UHPLC-MS method that allows rapid screening of prenylated flavonoids in multi-component plant extracts. Licorice roots were used as a source because of the presence of many different prenylated and non- prenylated flavonoids.

39

Chapter 2

EXPERIMENTAL

Materials

The roots of Glycyrrhiza glabra , collected in Afghanistan, were provided by Frutarom US (North Bergen, NJ, USA). Glabridin and glycyrrhizinic acid were purchased from Wako Chemicals GmbH (Neuss, Germany). UHPLC/MS grade acetonitrile (ACN) was purchased from Biosolve BV (Valkenswaard, The Netherlands). Water was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, USA). All other chemicals were purchased from Merck (Darmstadt, Germany).

Sample preparation

The roots were milled with a Retsch Ultra Centrifugal Mill ZM 200 (Haan, Germany) to yield a root powder with a particle size of 1 mm. The root powder (0.01 g powdered root/ ml solvent) was extracted with either 70% (v/v) aq. ethanol (EtOH), EtOH or ethyl acetate (EA). The compounds were extracted by a two-step sequential extraction with each solvent for 30 min in a sonication bath at 30°C. The extracts were centrifuged at 2500 g for 15 min. The dried extracts were obtained after evaporation (EtOH and EA) or after subsequent lyophilization (70% (v/v) aq. EtOH). The extracts was resolubilised in methanol (MeOH) and stored at -20°C. All samples were thawed and centrifuged before analysis.

RP-UHPLC analysis

Samples were analysed on a Thermo Accela UHPLC system (Thermo Scientific, San Jose, CA, USA) equipped with pump, autosampler and PDA detector. Samples (1 µl) were injected on an Acquity UPLC BEH C18 column (2.1 x 150 mm, 1.7 µm particle size) with an Acquity UPLC BEH C18 Vanguard pre-column (2.1 x 5 mm, 1.7 µm particle size; Waters, Milford, MA, USA). Water acidified with 0.1% (v/v) acetic acid, eluent A, and acetonitrile (ACN) acidified with 0.1% (v/v) acetic acid, eluent B, were used as eluents. The flow rate was 300 µl/min, and the PDA detector was set to measure at a range of 205-400 nm. The following elution profile was used: 0-18 min, linear gradient from 10%-100% (v/v) B; 18-22 min, isocratic on 100% B; 22-23 min, linear gradient from 100%-10% B 23-25 min, isocratic on 10% B.

Electrospray ionization mass spectrometry (ESI-MS)

Mass spectrometric data were obtained by analysing samples on a Thermo Scientific LTQ- XL (San Jose, CA, USA) equipped with an ESI-MS probe coupled to the reversed-phase (RP)-

40

LC-MS based screening method for prenylated flavonoids in licorice

UHPLC, 150 µl/min of the flow from the RP-UHPLC was directed to the MS. Helium was used as sheath gas and nitrogen as auxiliary gas. Data were collected over an m/z -range of 150-1500. Data dependent MS n analysis was performed with a normalised collision energy of 35%. The MS 3 fragmentation was always performed on the most intense daughter ion in the MS 2 spectrum. Most settings were optimised using “tune plus” (Xcalibur 2.07, Thermo Scientific) via automatic tuning. The system was tuned with glabridin in both positive and negative ionisation mode. In the NI mode, the ion transfer tube temperature was 350°C and the source voltage 4.5 kV. In the PI mode, the ion transfer tube temperature was 350°C and the source voltage 5.0 kV. Data acquisition and reprocessing were done with Xcalibur 2.07 (Thermo Scientific). Mass spectral data interpretation and peak determination were performed with Mass Frontier 4.0 (Highchem, Bratislava, Slovakia).

RESULTS AND DISCUSSION

Chromatographic analysis of root extracts from Glycyrrhiza glabra

The extraction of the licorice roots resulted in extraction yields of 23.7, 6.3 and 4.7 g/100g for 70% aq. EtOH (v/v), EtOH and EA, respectively. Each of the UHPLC-UV-profiles of the three licorice root extracts indicated a complex mixture of phenolic compounds ( Figure 1). Preliminary chemical analysis of the roots confirmed the presence of glabridin, which is a marker compound for G. glabra roots [23]. The 70% aq. EtOH (v/v) extract contained compounds eluting in the first 10 min of the chromatogram. Extracting with EtOH resulted in an extract with relatively less polar compounds and relative more apolar compounds, eluting in the second half of the chromatogram. With EA, mainly apolar compounds were extracted. A total of 18 peaks could be tentatively assigned in the UHPLC profiles by means of UV spectra, the m/z values of the parent ion, and Collision Induced Dissociation (CID) fragmentation of these parent ions in PI and NI mode ( Figure 2 and Table 1).

Peaks 1-9 Peaks 1-6 represent the flavonoid glycosides in the 70% (v/v) aq. EtOH and EtOH extracts, i.e. the three glycosidic forms of both liquiritigenin and isoliquiritigenin. The fragmentation of liquiritigenin, isoliquiritigenin and their corresponding glycosides have been described previously in PI and NI modes [24, 25]. Our observations are in line with these results ( Table 1).

41

Table 1 . Compounds tentatively assigned in roots Glycyrrhiza glabra by UHPLC-ESI-MS.

- 2 + 2 No Rt. UV max (nm) Identification Class Molecular [M -H] MS product ions [M+H] MS product ions (min) formula (relative intensity ) (relative intensity )

1 6.01 216, 271, 316 Liquiritin apioside Flavanone C26 H30 O13 549 429( 84 ), 417( 4), 297( 13 ), 255( 100 ) 551 419( 35 ), 257( 100 )

2 6.15 214, 277, 312 Liquiritin apioside Flavanone C26 H30 O13 549 429( 8), 417( 14 ), 297( 13 ), 255( 100 ) 551 419( 100 ), 257( 84 )

3 6.34 215, 275, 310 Liquiritin Flavanone C21 H22 O9 417 255( 100 ) 419 257( 100 )

4 7.32 240, 361 Isoliquiritin apioside Chalcone C26 H30 O13 549 429( 10 ), 417( 15 ), 297( 15 ), 255( 100 ) 551 419( 88 ), 257( 100 )

5 7.45 240, 369 Isoliquiritin apioside Chalcone C26 H30 O13 549 429( 80 ), 417( 4), 297( 12 ), 255( 100 ) 551 419( 100 ), 257( 40 )

6 7.80 240, 370 Isoliquiritin Chalcone C21 H22 O9 417 255( 100 ) 419 257( 100 )

7 8.17 215, 360 Licochalcone B Chalcone C16 H14 O5 285 270( 100 ), 253( 8), 225( 1), 209( 1), 287 245( 100 ), 193( 10), 167( 7), 191( 7), 150( 2) 147( 7), 121( 6)

8 10.00 219 Glycyrrhizinic acid Triterpenoid C42 H62 O16 822 803( 15 ), 759( 7), 645( 10 ), 351( 100 ) 823 647( 30 ), 471( 16 ), 453( 100 )

9 10.70 216, 248, 297 Formononetin Isoflavone C16 H12 O4 267 252( 100 ), 223( 1), 208( 1), 191( 1) 269 254( 100 ), 237( 39 ), 213( 33 ), 107( 10 )

10 12.26 217, 284, 324 Glabrene Isoflav -3-ene C20 H18 O4 321 306( 100 ), 303( 27 ), 293( 18 ), 277( 27 ), 322 Not determined 175( 18 ), 145( 14 )

11 13.50 222, 250, 302 Glabrone Isoflavone C20 H16 O5 335 320( 18 ), 307( 10 ), 292( 11 ), 291( 10 0), 337 295( 41 ), 283( 100 ), 213( 17 ) 239( 17 ), 137( 22 )

12 13.53 230, 281 Glabridin Isoflavan C20 H20 O4 323 213( 37 ), 201( 71 ), 147( 28 ), 135( 100 ), 325 269( 14 ), 215( 8), 203( 21 ), 121( 35 ) 189( 100 ), 123( 35 )

13 14.09 220, 283 Glabrol Flavanone C25 H28 O4 392 203( 100 ), 187( 24 ), 159( 5) 394 337( 100 ), 205( 15 ), 203( 10 )

14 14.59 223, 271, 323 3’ -hydroxy -4’ -O- Isoflavan C21 H22 O5 353 338( 31 ), 201( 100 ), 175( 28 ), 165( 45 ) 355 215( 4), 189( 100 ), 153( 73 ), methyl-glabridin 147( 5)

15 15.35 208, 226, 281 4’ -O-methyl - Isofla van C21 H22 O4 337 322( 47 ), 213( 11 ), 201( 100 ), 175( 54 ), 339 215( 4), 189(100), 147( 3), glabridin 149( 10 ), 123( 11 ) 137( 52 )

16 15.83 206, 228, 280 Hispaglabridin A Isoflavan C25 H28 O4 391 215( 36 ), 203( 100 ), 201( 49 ), 189( 40 ), 393 337( 100 ), 189( 79 ), 191( 79 ) 177( 63 )

17 15. 93 215,265, 3-hydroxy -glabrol Flavanone C25 H28 O5 407 203( 100 ), 221( 3), 159( 2) 409 353( 100 ), 335( 29 ), 229( 6), 280,383 205( 21 ), 177( 6)

18 17.06 228, 280 Hispaglabridin B Isoflavan C25 H26 O4 389 201( 100 ), 187( 5), 175( 13 ) 391 189( 100 ), 147( 10 )

LC-MS based screening method for prenylated flavonoids in licorice

The fragmentation pattern of both liquiritigenin (flavanone) and isoliquiritigenin (chalcone) in ESI were very similar, making it difficult to discriminate between chalcone and flavanone. This is caused by thermal isomerisation by which the C-ring of the flavanone is opened to form its corresponding chalcone [26]. Based on the UV spectra described by Fu et al . [24], the peaks were assigned either liquiritigenin glycoside or isoliquiritigenin glycoside. Peaks 7, 8 and 9 were assigned as licochalcone B, glycyrrhizinic acid and formononetin, respectively. Peaks 7 and 9 were confirmed by UV spectra and MS fragmentation patterns in NI mode, comparable to those found by Wang et al . [27]. Peak 8 was assigned as the main triterpenoid glycoside in the roots of Glycyrrhiza glabra by means of the authentic standard.

Figure 1 . RP-UHPLC-UV profile of Glycyrrhiza glabra roots extracted with 70% aq. EtOH (A), EtOH (B) and EA (C). Peak numbers refer to compounds in Table 1 .

43

Chapter 2

Figure 2 . Structures of compounds assigned in a Glycyrrhiza glabra root extract.

44

LC-MS based screening method for prenylated flavonoids in licorice

Peaks 10-18 Peaks 10 -18 , eluting in the second half of the chromatogram, were assigned as prenylated flavonoids. The predominant prenylated flavonoid in the chromatogram was glabridin (12 ). There were 4 other prenylated isoflavans assigned with the same skeleton as glabridin: 3’-hydroxy-4’-O-methyl-glabridin ( 14 ), 4’-O-methyl-glabridin ( 15 ), hispaglabridin A ( 16 ) and hispaglabridin B ( 18 ). The fragmentation patterns of these 5 isoflavans are shown in Figure 3. The C-ring is cleaved in several places resulting in different retro-Diels Alder (RDA) fragments. The nomenclature for the RDA fragmentation was adapted from Ma et al . [28] and is shown in Figure 4 . The RDA fragments dominate the MS 2 spectra of the isoflavans ( 12 , 14 -16 and 18 ). In addition, a number of small neutral losses were observed, such as 28 (CO) and 44 (CO 2). This indicated that the C-ring of the isoflavans studied is less stable, and more prone to degradation, than that of, for example, glabrene (isoflav-3-ene) and glabrone (isoflavone) (see Appendix I and II, respectively).

Figure 3 . Proposed fragmentation pathway of glabridin ( 12 ), 3’-hydroxy-4’-O-methyl-glabridin ( 14 ), 4’-O-methyl- glabridin ( 15 ), hispaglabridin A ( 16 ) and B ( 18 ) in NI mode. The numbers inside the parenthesis refer to the compounds in Table 1 .

45

Chapter 2

The fragmentation patterns of 12 , 14 -16 and 18 were basically the same in NI mode. Fragments 1,3 B- and 2,3 A- were the most abundant ions in all spectra. The other RDA fragments were present in all MS 2 spectra of these isoflavans, with the exception of fragment ions 1,4 B- and 2,3 B-. These ions were only formed in the MS 2 of 12 and 16 . This might be due to the conversion of the para hydroxyl group of the B-ring into a carbonyl, which is only possible if the para hydroxyl group is unsubstituted. Both fragmentation patterns of glabrol ( 13 ) and 3-hydroxy-glabrol ( 17 ) in NI mode were characterised by a predominant 1,3 A- ion and a much smaller 1,3 B- ion (as given 1,3 - 1,3 in see Appendix III ). Furthermore, the A ion underwent a further CO 2 loss yielding a [ A - [9] - CO 2] fragment ion, a mechanism that has been described for flavonoids in NI mode . The fragmentation of glabrene ( 10 ) is characterised by the losses of a methyl • - - radical CH 3 , a CO and a CO 2. Furthermore, the fragment ions [B ring -H] and [M-H-Bring ] were observed, as well as four RDA fragments (see Appendix I ). The main fragment ions in the MS 2 spectrum of the isoflavone glabrone ( 11 ) were • due to small neutral losses such as a methyl radical (CH 3 ) and CO 2. The most prominent 1,4 - RDA fragment observed was [ B - H2O] (see Appendix II ). This is the first report on the fragmentation of prenylated flavonoids 10 -18 in ESI- MS. The fragmentation patterns indicate that prenylated flavonoids follow the same degradation pattern, like the formation of RDA fragments and small neutral losses like CO and CO 2. Furthermore, it appeared that the prenyl-groups are not directly involved in the MS 2 fragmentation in negative mode.

Figure 4 . Nomenclature adopted for the different retrocyclization cleavages on a flavanone and an isoflavan.

46

LC-MS based screening method for prenylated flavonoids in licorice

Neutral losses 42 and 56 in PI mode MS n spectra of prenylated flavonoids

The MS 2 spectra of prenylated flavonoids 13 , 16 and 17 in PI mode were characterised by the dominant fragment ion [M+H-56] +. Furthermore, the fragmentation of the isoflavans 12 , 14 , 15 and 18 all had similar degradation patterns in PI mode characterised by the presence of RDA fragments (see Appendix IV ). The dominant RDA fragment ion 1,3 A+ is substituted with a pyran ring. Upon fragmentation of 1,3 A+, a predominant loss of 42 was observed in the MS 3 spectra. The MS 2 spectrum of glabrone ( 11 ) in PI mode was characterised by neutral losses of 42, 54 and 56, next to RDA fragment ions 1,3 A+ and 1,3 B+. Table 2 summarises the prenylated flavonoids, assigned by us in the licorice root extract, for which neutral losses 42 and 56 were observed. The [M+H-56] + ion is the most abundant ion in the MS 2 spectra of compounds 13 , 16 and 17 , which have in common that they are substituted with at least one prenyl chain. Furthermore, in the MS 3 spectra of 12 , 14 -16 and 18 a dominant neutral loss of 42 was observed. These 5 prenylated flavonoids have in common that they are substituted with at least one pyran ring.

Table 2 . The presence of neutral losses 42 [C 3H6] and 56 [C 4H8] originating from the degradation of the prenyl group in PI mode. MS 2 MS 3 No 1 Name 42 56 42 56 10 Glabrene n.d. n.d. n.d. n.d. 11 Glabrone ++ + n.d. n.d. 12 Glabridin - ± ++ ± 13 Glabrol - ++ + + 14 3’ -hydroxy -4’ -O-methyl -Glabridin - - ++ ± 15 4’ -O-methyl -Glabridin - - ++ ± 16 Hispaglabridin A - ++ + ± 17 3-hydroxy -Glabrol - ++ - + 18 Hispaglabridin B - - ++ ± a Number refers to the compounds in Table 1 . - Fragment ion not determined, less than 0.1%. ± The relative abundance of this fragment ion is less than 5 %. + The relative abundance of this fragment ion is between 5% and 100%. ++ The relative abundance of this fragment ion is 100% (dominating fragment in the MS n spectrum). n.d. not determined 1 Glabrene was tentatively assigned in NI mode. The [M-H] - parent had an m/z value of 321. In PI mode, however, the [M+H] + of the peak at the same retention time had an m/z value of 322, while 323 was expected. Additional experiments with high resolution MS showed that two compounds were co-eluting with an m/z value of 321 and 323. This resulted in an average m/z value of 322 in the UHPLC-MS chromatogram. It was not possible to determine the fragmentation of glabrene.

47

Chapter 2

From the results shown in Table 2, it appeared that the degradation of a prenyl chain in PI mode in MS 2 preferably resulted in a neutral loss of 56, sometimes accompanied by a minor loss of 42. In contrast, the degradation of the pyran ring appeared to be characterised by a predominant neutral loss of 42 in combination with a minor loss of 56. The loss of 56 is suggested to be attributed to the loss of C 4H8, whereas the loss of 42 corresponds with the loss of C 3H6. This characteristic difference in fragmentation of both prenyl groups is supported by observations of Stevens et al . [10] and Da Costa et al . [29] in PI mode APCI. Based on our observations and the literature, a proposed fragmentation pathway for both prenyl chain and pyran ring is shown in Figure 5.

Figure 5 . Proposed degradation pathways of a prenyl chain and a pyran ring attached to the A-ring of a flavonoid in PI mode. The same degradation pathways apply for a B-ring prenylation. Bold arrow indicates preferred fragmentation pathway, dotted arrow indicates less likely pathway. The loss of 56 u is always observed in both prenyl chain and pyran ring degradation. The loss of 42 u is always observed in the degradation of a pyran ring, but not necessarily in that of the prenyl chain.

48

LC-MS based screening method for prenylated flavonoids in licorice

The fragmentation of prenyl groups of prenylated polyphenolic compounds has been reviewed by Takayama et al . [26]. This review focused on the fragmentation patterns in electron ionisation (EI), PI mode FAB [9] and chemical ionisation (CI) and summarised the neutral losses encountered upon fragmentation of the prenyl group ( Table 3). This overview shows that the neutral losses of 42 and 56 are commonly observed upon fragmentation of prenylated flavonoids. In a more recent study using ESI-MS [4] a neutral loss of 56 was also observed. These observations are consistent with our findings on the fragmentation of prenylated flavonoids obtained with ESI-MS.

Table 3. Characteristic fragments for the fragmentation of a prenyl-group reported for different ionisation techniques in positive mode. EI 1 (+)FAB 1 (+)APCI 2, 3 (+)ESI 4 Neutral Loss Chain Pyran Chain Pyran Chain Pyran Chain Loss

42 C3H6

• 43 C3H7

54 C4H6
• 55 C4H7

56 C4H8




• 57 C4H9

68 C5H8

• 69 C5H9

71 C5H11
EI – electron ionisation; APCI – atmospheric pressure chemical ionisation; ESI – electrospray ionisation 1 Takayama (1992); 2 Stevens (1997); 3 Da Costa (2000); 4 Zhang (2008)

Table 4. Number of individual peaks detected in 70% aq. EtOH, EtOH, and EA extracts and their classification based on the screening method. The ‘total’ column summarises the total number of unique peaks in all three extracts. Group 70% EtOH EtOH EA Total Total number of peaks 66 82 60 118 Screening for 56 and 42 45 58 48 88 - m/z < 305 (non -prenylated) 5 5 1 8 - Sugar loss detected in MS 2 6 6 3 10 Total prenylated 34 47 44 70 To tal non -prenylated 32 35 16 48 Mono -prenylated 26 34 33 52 - Prenyl chain 20 26 26 41 - Pyran ring 6 8 7 11 Di -prenylated 9 14 10 18 - Prenyl chain + Prenyl chain 5 10 6 12 - Prenyl Chain + Pyran Ring 4 4 4 6

49

Chapter 2

Screening method for prenylated flavonoids based on neutral losses 42 and 56

The screening method for prenylated flavonoids is based on the detection of neutral losses. In addition to the detection of prenyl substituents, it is possible to discriminate between prenyl chain and pyran ring, based on the relative abundances of the ions that lose 56 and 42, respectively. From these relative abundances for the compounds shown in Table 2 it appears that if the ratio (42:56) is <1.0, the prenyl group is a prenyl chain. If the ratio is >1.0, the prenyl group is a pyran ring. Neutral losses 42 and 56 have also been observed in the fragmentation of non- prenylated flavonoid aglycones. The fragmentation of licochalcone B ( 7) was characterised by a predominant neutral loss of 42, and in the MS 2 of formononetin ( 9) a neutral loss of 56 was observed ( Table 1). This indicated that the occurrence of these neutral losses in the MS n spectra of flavonoids in PI mode do not exclusively point at the presence of prenyl groups. Losses of 42 have been observed by Ma et al . [28] upon fragmentation of flavonoid aglycones in the PI mode, and were attributed to the loss of C 2H2O, originating from the C- ring of the flavonoid aglycones. The neutral loss of 56 has been observed by Kuhn et al . in the fragmentation patterns of different isoflavonoid aglycones [30]. In their study the authors concluded that a double loss of two CO molecules accounted for the loss of 56. To rule out the possibility that flavonoid aglycones were assigned as prenylated flavonoids, a criterion was set to selectively exclude flavonoid aglycones on the basis of their molecular weight. The prenylated flavonoids of the lowest molecular weight isolated from Glycyrrhiza glabra are glabrene and 2',4'-dihydroxy-6'',6''-dimethylpyrano- [2'',3'':7,8]isoflav-3-ene (both isoflav-3-enes) and have a molecular weight of 322 [20] (C 20 H18 O4) . According to an extensive review on prenylated flavonoids by Barron and Ibrahim, the smallest prenylated flavonoid isolated from a natural source is 7,8-(2,2- [5] dimethylchromeno) flavone, with a molecular weight of 304 (C 20 H16 O3) . The m/z of the parent ion of this molecule in PI mode would be 305. To exclude non-prenylated flavonoid aglycones from the screening method, an m/z threshold of 305 was set. In plants, flavonoids often appear as O-glycoside conjugates. The O-glycoside conjugates of non-prenylated flavonoid aglycones that lose 56 u upon fragmentation exceed the m/z threshold of 305, and they might be inappropriately scored as prenylated flavonoids: in MS 2 the glycoside moiety is cleaved off, after which the aglycone loses 56 in MS 3. In order to prevent that non-prenylated flavonoid O-glycosides are assigned as prenylated, the MS 2 spectra are checked for neutral losses corresponding with the loss of a sugar moiety. The predominant loss of 132 (pentose), 146 (deoxyhexose), 162 (hexose) or 176 (uronic acid), or a loss of more than 264 (2 pentoses) in the MS 2 spectrum is diagnostic for a flavonoid O-glycoside, and these compounds are, therefore, assigned as non-prenylated. The exclusion of flavonoid O-glycosides may consequently lead to the possibility that prenylated flavonoid O-glycosides are not detected. Five prenylated

50

LC-MS based screening method for prenylated flavonoids in licorice

flavonoid O-glycosides have been described occurring within the plant families of the Leguminosae , Rutaceae , Rhamnaceae and Pteridaceae [31]. However, most prenylated flavonoid O-glycosides have been isolated from species within the genera Epimedium and Vancouveria belonging to the family of Berberidaceae [5]. Our screening procedure, developed on the basis of fragmentation of the prenylated flavonoids from Glycyrrhiza glabra in the PI mode, is summarised in Figure 6 . The first selection criterion is the presence of a neutral loss of 56, alone or in combination with a neutral loss of 42 in the MS 2 spectrum. If a neutral loss is not present in MS 2, the MS 3 spectrum is analysed. The next step is to verify whether the m/z value of the parent ion [M+H] + is above 305 to rule out non-prenylated flavonoid aglycones. In case neutral losses 56, or both 56 and 42, are detected in the MS 3, the MS 2 of this peak is analysed for neutral losses corresponding to the loss of sugar moieties. If the MS 2 spectra indicates loss of glycosyl residues, the peak is considered non-prenylated. If these criteria are met, the compound is considered to be prenylated.

Figure 6. Interpretation guideline for identification of prenylated flavonoids.

51

Chapter 2

The second prenyl group of doubly prenylated flavonoids can be determined as well. The minimum molecular weight of a double prenylated flavonoids is 372, calculated by the addition of a prenyl chain (68) to the single prenylated flavone, 7,8-(2,2- dimethylchromeno) flavone (MW=304), described earlier. Therefore, if the m/z value of the parent ion [M+H] + is higher than 373, the MS 3 spectrum is checked. This can only be done in case the MS 3 was performed on the [M+H-42] + or the [M+H-56] + ion.

Verification of the screening method proposed

To verify the screening method proposed, the 70% aq. EtOH, EtOH and EA extracts of Glycyrrhiza glabra roots were screened for the presence of prenylated flavonoids in addition to the tentatively assigned compounds 10 -18 . Based on the analysis of the MS chromatograms of the three different extracts by Mass Frontier software, 118 individual peaks were detected (see Table 4 and Appendix V ). Out of these 118 peaks, 88 showed a neutral loss of 56, alone or in combination with a neutral loss of 42, in either the MS 2 or MS 3 spectrum. Of these 88 peaks, 8 were assigned as non-prenylated aglycones, because their m/z value did not exceed 305. Another 10 peaks were assigned as flavonoid glycosides, as their MS 2 fragmentation indicated a loss of one or more sugar moieties. In total, 70 unique peaks were assigned as prenylated flavonoids present in the 3 different extracts of Glycyrrhiza glabra . Of these 70 peaks, 52 peaks were assigned as mono- prenylated and 18 as di-prenylated. Based on the ratio (42:56), the majority of the 70 peaks have a prenyl chain (see Appendix V ). The EA extract appeared to selectively extract prenylated flavonoids, which may be related to the non-polar properties of this solvent. For this reason, the EA extract was chosen as an illustrative example in Table 5 and Figure 7 . Table 5 shows an overview of the 48 peaks in the MS chromatogram of the EA extract that exhibit a loss of 56, alone or in combination with 42 in MS n. The 43 peaks finally assigned as being prenylated, are shown in the chromatograms in Figure 7. Roughly half of the peaks in Table 5 assigned as prenylated, had the [M+H-56] + ion as their most abundant fragment in their MS 2 spectrum and, in addition, most of the peaks in Table 5 had a ratio (42:56) smaller than 1, indicating the presence of a prenyl chain. Peak S21 was the only peak with a pyran ring determined by the MS 2 spectrum. The other peaks that were assigned having a pyran ring, were based on their ratio (42:56) in their MS 3 spectrum. This observation indicated that a pyran ring is more stable to MS degradation than a prenyl chain, as the fragmentation of the prenyl chain mostly occurred in MS 2, whereas the pyran ring is mostly fragmented in MS 3.

52

LC-MS based screening method for prenylated flavonoids in licorice

Figure 7. Screening results for neutral losses 42 u and 56 u in the PI mode MS 2 and MS 3 spectra. RP-UHPLC-MS base peak chromatogram in positive mode (A) of the EA extract of Glycyrrhiza glabra roots. The MS chromatograms were screened for neutral losses in MS 2 of 42 u (B) and 56 u (C), and for neutral losses in MS 3 of 42 u (D) and 56 u (E). The numbers in the figure relate to the number given in Table 5 .

53

Table 5. Screening results for prenylated flavonoids in the MS chromatogram of the EA extract of Glycyrrhiza glabra in PI mode. (C - prenyl chain; P - pyran ring; np – non-prenylated). 2 3 MS MS No * Rt ** Identification [M+H] + (min) MS (r.a.) 42 loss Neutral (r.a.) 56 loss Neutral (42:56) Ratio -type Prenyl (r.a.) 42 loss Neutral (r.a.) 56 loss Neutral (42:56) Ratio -type Prenyl

S23 10.44 327 - - - - 3 1 3.0 P S26 10.79 321 24 25 1.0 C S27 10.89 Formononetin 269 - 33 0.0 np S29 11.27 341 <1 100 <0.1 C S31 11.38 339 <1 100 <0.1 C S32 11.51 323 2 100 <0.1 C S33 11.57 341 - 100 0.0 C S34 11.67 323 1 100 <0.1 C S35 11.79 355 41 100 0.4 C S36 11.90 337 1 7 0.1 C S37 12.10 355 <1 100 <0.1 C S38 12.20 325 3 100 <0.1 C S39 12.28 321 24 28 0.9 C S41 12.53 322 - 7 0.0 C S42 12.68 337 28 57 0.5 C S43 12.72 339 5 17 0.3 C S45 12.97 308 - - - - 3 4 0.8 C S47 13.25 355 - 100 0.0 C S48 13.31 341 - 100 0.0 C S49 13.52 409 - 21 0.0 C S50 13.64 Glabrone 337 100 6 16.7 P S51 13.77 Glabridin 325 - <1 0.0 - 100 1 100 P S53 13.91 321 18 64 0.3 C S54 14.08 391 - 100 0.0 C 1 100 <0.1 C S55 14.13 645 1 100 <0.1 C 21 10 2.1 P

S57 14.29 Glabrol 393 - 100 0.0 C 21 22 1.0 C S58 14.49 409 - 100 0.0 C 19 100 0.2 C S59 14.66 395 - 82 0.0 C S60 14.72 647 - 60 0.0 C S62 14.79 3’ -hydroxy -4’ -O-methyl -glabridin 355 - - - - 100 1 100 P S64 14.89 409 - 100 0.0 C 30 20 1.5 P S66 15.04 647 - 14 0.0 C S67 15.08 661 - 100 0.0 C 1 35 <0.1 C S69 15.27 391 - 18 0.0 C S72 15.42 645 <1 100 <0.1 C 35 18 1.9 P S74 15.52 4’ -O-methyl -glabridin 339 - - - - 100 1 100 P S76 15.73 631 <1 100 <0.1 C 34 6 5.7 P S77 15.88 391 - 100 <0.1 C 6 7 0.9 C S78 15.97 Hispaglabridin A 393 - 100 <0.1 C 21 3 7.0 P S80 16.08 3-hydroxy -glabrol 409 - 100 0.0 C 1 9 0.1 C S81 16.18 391 1 39 <0.1 C S82 16.34 409 - 42 0.0 C S83 16.44 643 2 63 <0 .1 np S84 16.53 393 - 100 0.0 C 10 8 1.3 P S85 16.72 659 - 26 0.0 np S86 16.81 469 - - - - 4 14 0.3 C S87 16.89 388 1 2 0.5 C 11 6 1.8 P S88 17.21 Hispaglabridin B 391 - 3 - - 100 1 100 P r.a. relative abundance < 1 - 0.5% tot 1.0% n.d. - not detected, <0.5% * the ‘S’ refers to a peak found with the screening method in the EA extract ** The retention time in MS has a delay of 0.20±0.02 min compared to the UV retention time

Chapter 2

Around 75 prenylated flavonoids have been isolated from the roots of Glycyrrhiza glabra [13-22]. The application of the screening method described here indicated the presence of 70 unique prenylated flavonoids in 3 different extracts of the roots of Glycyrrhiza glabra without any purification or fractionation steps. This suggested that many of the prenylated flavonoids present can be detected in a crude extract by the application of the described screening method. A next step would be the application of this screening method for the detection of prenylated flavonoids in other plant extracts and, furthermore, to provide a greater number of definitive assignments of the prenylated flavonoids detected in complex extracts using the current screening method.

CONCLUSIONS

A rapid screening method is described to screen for prenylated flavonoids. The method is based on the fragmentation of prenyl substituents of flavonoids in PI mode MS n, involving a neutral loss 56, alone or in combination with 42. In addition to the detection of prenyl substituents, it is possible to discriminate between prenyl chain and pyran ring, based on the relative abundances of the ions that lose 42 and 56 (42:56). A ratio (42:56) of <1.0 indicates the presence of a prenyl chain, a ratio >1.0 that of a pyran ring. The screening of three different extracts of roots from Glycyrrhiza glabra resulted in the detection of 70 peaks that are likely to be prenylated.

ACKNOWLEDGEMENTS

This work was supported by the Food & Nutrition Delta of the Ministry of Economic Affairs, the Netherlands.

REFERENCES

[1] R. K. Ibrahim, A forty-year journey in plant research: Original contributions to flavonoid biochemistry. Canadian Journal of Botany 2005 , 83 , 433-450. [2] B. Botta, A. Vitali, P. Menendez, D. Misiti, G. D. Monache, Prenylated flavonoids: Pharmacology and biotechnology. Current Medicinal Chemistry 2005 , 12 , 713-739. [3] P. Shen, S. P. Wong, J. Li, E. L. Yong, Simple and sensitive liquid chromatography-tandem mass spectrometry assay for simultaneous measurement of five Epimedium prenylflavonoids in rat sera. Journal of Chromatography B 2009 , 877 , 71-78.

56

LC-MS based screening method for prenylated flavonoids in licorice

[4] Y. Zhang, P. Zhang, Y. Cheng, Structural characterization of isoprenylated flavonoids from Kushen by electrospray ionization multistage tandem mass spectrometry. Journal of Mass Spectrometry 2008 , 43 , 1421-1431. [5] D. Barron, R. K. Ibrahim, Isoprenylated flavonoids - A survey. Phytochemistry 1996 , 43 , 921-982. [6] B. Botta, G. D. Monache, P. Menendez, A. Boffi, Novel prenyltransferase enzymes as a tool for flavonoid prenylation. Trends in Pharmacological Sciences 2005 , 26 , 606-608. [7] B. J. Doyle, G. B. Mahady, Phytomedicines for menopause. Drugs of the Future 2007 , 32 , 897-905. [8] P. Zanoli, M. Zavatti, Pharmacognostic and pharmacological profile of Humulus lupulus L. Journal of Ethnopharmacology 2008 , 116 , 383-396. [9] N. Fabre, I. Rustan, E. de Hoffmann, J. Quetin-Leclercq, Determination of flavone, flavonol, and flavanone aglycones by negative ion liquid chromatography electrospray ion trap mass spectrometry. Journal of the American Society for Mass Spectrometry 2001 , 12 , 707-715. [10] J. F. Stevens, M. Ivancic, V. L. Hsu, M. L. Deinzer, Prenylflavonoids from Humulus lupulus . Phytochemistry 1997 , 44 , 1575-1585. [11] M. Blumenthal, The Complete German Commission E Monographs: Therapeutic Guide to Herbal Medicines , American Botanical Council, Austin, Texas, 1998 . [12] T. Nomura, T. Fukai, T. Akiyama, Chemistry of phenolic compounds of licorice ( Glycyrrhiza species) and their estrogenic and cytotoxic activities. Pure and Applied Chemistry 2002 , 74 , 1199-1206. [13] Y. Asada, W. Li, T. Yoshikawa, The first prenylated biaurone, licoagrone from hairy root cultures of Glycyrrhiza glabra . Phytochemistry 1999 , 50 , 1015-1019. [14] M. Baba, S. Sumi, N. Iwasaki, H. Kai, M. Sakano, Y. Okada, T. Okuyama, Studies of the Egyptian traditional folk medicines. IV. New isoflavonoid isolated from Egyptian licorice. Heterocycles 2008 , 75 , 3085-3089. [15] J. Bielenberg, Isoflavonoids from licorice root as mediators of anti-inflammatory and anti-allergic effects. Zeitschrift fur Phytotherapie 2001 , 22 , 289-293. [16] Y. W. Chin, H. A. Jung, Y. Liu, B. N. Su, J. A. Castoro, W. J. Keller, M. A. Pereira, A. D. Kinghorn, Anti- oxidant constituents of the roots and stolons of licorice ( Glycyrrhiza glabra ). Journal of Agricultural and Food Chemistry 2007 , 55 , 4691-4697. [17] T. Fukai, B. S. Cai, K. Maruno, Y. Miyakawa, M. Konishi, T. Nomura, An isoprenylated flavanone from Glycyrrhiza glabra and rec-assay of licorice phenols. Phytochemistry 1998 , 49 , 2005-2013. [18] T. Kinoshita, Y. Tamura, K. Mizutani, The isolation and structure elucidation of minor isoflavonoids from licorice of Glycyrrhiza glabra origin. Chemical and Pharmaceutical Bulletin 2005 , 53 , 847-849. [19] W. Li, Y. Asada, T. Yoshikawa, Flavonoid constituents from Glycyrrhiza glabra hairy root cultures. Phytochemistry 2000 , 55 , 447-456. [20] T. Nomura, T. Fukai, Phenolic constituents of licorice ( Glycyrrhiza species), Progress in the chemistry of organic natural products , Vol. 73 , Springer, 1998 . [21] M. Shibano, Y. Matsumoto, E. Nakao, A. Henmi, G. Kusano, T. Shibata, Y. Hatakeyama, H. Hayashi, T. Kinoshita, Studies on index compounds for HPLC analysis of Glycyrrhiza glabra . Natural Medicines 1998 , 52 , 279-283. [22] E. M. Williamson, in Potter's Cyclopedia of Herbal Medicine , Saffron Walden, UK, 2003 , pp. 269-271. [23] K. Kondo, M. Shiba, R. Nakamura, T. Morota, Y. Shoyama, Constituent properties of licorices derived from Glycyrrhiza uralensis, G. glabra, or G. inflata identified by genetic information. Biological and Pharmaceutical Bulletin 2007 , 30 , 1271-1277. [24] B. Fu, H. Li, X. Wang, F. S. C. Lee, S. Cui, Isolation and identification of flavonoids in licorice and a study of their inhibitory effects on tyrosinase. Journal of Agricultural and Food Chemistry 2005 , 53 , 7408- 7414. [25] M. Ye, S. H. Liu, Z. Jiang, Y. Lee, R. Tilton, Y. C. Cheng, Liquid chromatography/mass spectrometry analysis of PHY906, a Chinese medicine formulation for cancer therapy. Rapid Communications in Mass Spectrometry 2007 , 21 , 3593-3607. [26] M. Takayama, T. Fukai, Y. Hano, T. Nomura, Mass spectrometry of prenylated flavonoids. Heterocycles 1992 , 33 , 405-434. [27] Y. Wang, L. Yang, Y. Q. He, C. H. Wang, E. W. Welbeck, S. W. A. Bligh, Z. T. Wang, Characterization of fifty-one flavonoids in a Chinese herbal prescription Longdan Xiegan Decoction by high-performance

57

Chapter 2

liquid chromatography coupled to electrospray ionization tandem mass spectrometry and photodiode array detection. Rapid Communications in Mass Spectrometry 2008 , 22 , 1767-1778. [28] Y. L. Ma, Q. M. Li, H. van den Heuvel, M. Claeys, Characterization of flavone and flavonol aglycones by collision-induced dissociation tandem mass spectrometry. Rapid Communications in Mass Spectrometry 1997 , 11 , 1357-1364. [29] C. T. Da Costa, J. J. Dalluge, M. J. Welch, B. Coxon, S. A. Margolis, D. Horton, Characterization of prenylated xanthones and flavanones by liquid chromatography/atmospheric pressure chemical ionization mass spectrometry. Journal of Mass Spectrometry 2000 , 35 , 540-549. [30] F. Kuhn, M. Oehme, F. Romero, E. Abou-Mansour, R. Tabacchi, Differentiation of isomeric flavone/isoflavone aglycones by MS 2 ion trap mass spectrometry and a double neutral loss of CO. Rapid Communications in Mass Spectrometry 2003 , 17 , 1941-1949. [31] N. C. Veitch, R. J. Grayer, Flavonoids and their glycosides, including anthocyanins. Natural Product Reports 2008 , 25 , 555-611.

58

Chapter 3

Identification of prenylated pterocarpans and other isoflavonoids in Rhizopus spp. elicited soya bean seedlings by electrospray ionisation mass spectrometry

Based on : Rudy Simons, Jean-Paul Vincken, Maxime C. Bohin, Tomas F.M. Kuijpers, Marian A. Verbruggen, Harry Gruppen, Identification of prenylated pterocarpans and other isoflavonoids in Rhizopus spp. elicited soya bean seedlings by electrospray ionisation mass spectrometry, Rapid Communications in Mass Spectrometry 2011 , 25, 55-65 Reprinted with permission

Chapter 3

ABSTRACT

Phytoalexins from soya are mainly characterised as prenylated pterocarpans, the glyceollins. Extracts of non-soaked and soaked soya beans, as well as that of soya seedlings, grown in the presence of Rhizopus microsporus var. oryzae , were screened for the presence of prenylated flavonoids with an LC-MS based screening method. The glyceollins I-III and glyceollidins I-II, belonging to the isoflavonoid subclass of the pterocarpans, were tentatively assigned. The formation of these prenylated pterocarpans was accompanied by that of other prenylated isoflavonoids of the subclasses of the isoflavones and the coumestans. It was estimated that approx. 40% of the total isoflavonoid content in Rhizopus -challenged soya bean seedlings were prenylated pterocarpans, whereas 7% comprised prenylated isoflavones and prenylated coumestans. The site of prenylation (A-ring or B-ring) of the prenylated isoflavones was tentatively annotated using positive-ion mode MS by comparing the 1,3 A+ RDA fragments of prenylated and non-prenylated isoflavones. Furthermore, the fragmentation pathways of the 5 pterocarpans in NI mode were proposed, which involved the cleavage of the C ring and/or D ring. The absence of the ring-closed prenyl (pyran or furan) gave exclusively - x,y H2O RDA fragments, whereas its presence predominantly the common RDA fragments.

Keywords : Prenylation, isoflavonoids, soya, phytoalexins, germination, pterocarpans

60

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

INTRODUCTION

The application of stress on plants is a well-known approach to induce changes in the chemical profile of secondary metabolites, including phytoalexins. This is a heterogeneous group of low molecular weight anti-biotic compounds that are de novo synthesized in response to stress, pathogens or chemical elicitors [1]. Phytoalexins increasingly gain scientific interest due to their bioactivity and potential to enhance the nutraceutical value of crops [2]. Soya ( Glycine max ) is an important source of isoflavonoids. Soya beans contain isoflavones, a subclass of isoflavonoids that are considered to be bio-active constituents responsible for some of the health-improving effects ascribed to soya [3]. The three main isoflavones in soya beans are daidzein, genistein and glycitein, and their respective glucosides, acetyl-glucosides and malonyl-glucosides [4]. Upon application of stress, such as fungal infection, soya seedlings produce coumestrol and glyceollins as the main phytoalexins [5]. Both coumestrol and glyceollins are derived from the precursor daidzein and belong to the isoflavonoid subclasses of the coumestans and pterocarpans, respectively. Another group of compounds known to be generated in black soya beans upon pathogen-stressed germination are keto-oxo-octadecadienoic acids (KODEs) [6]. KODEs belong to the class of oxylipins, a broad class of signalling molecules derived from the oxidation of polyunsaturated fatty acids, of which jasmonic acid is the most well- known representative. In the context of plant defence, oxylipins are reported to have antimicrobial effects, stimulate plant defence gene expression, and regulate plant cell death. Their exact contribution to each functionality is largely unknown [7-9]. The formation of glyceollins in soya seedlings involves the conversion of daidzein into glycinol, which is subsequently A-ring prenylated on the 4 or the 2 position leading to the glyceollidins I and II, respectively. Cyclisation of the prenyl chain results for glyceollidin I into glyceollin I, and for glyceollidin II into glyceollin II and III [5, 10-13]. Prenylation of isoflavonoids is a common modification in the formation of phytoalexins in Leguminosae [14]. Prenylation of flavonoids is thought to increase their lipophilicity, their affinity to biological membranes, and their interaction with target proteins [15]. In this study, the extract of soya beans germinated in the presence of the fungus Rhizopus microsporus var. oryzae was screened for the presence of prenylated flavonoids.

EXPERIMENTAL

Materials

Soya beans, Glycine max (L.) Merrill, were provided by Frutarom Belgium (Londerzeel, Belgium). Authentic standards of genistein, daidzein and coumestrol were purchased from Sigma Aldrich (St. Louis, MO, USA). UHPLC/MS grade acetonitrile (ACN) was purchased

61

Chapter 3

from Biosolve BV (Valkenswaard, The Netherlands). Water was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, USA). All other chemicals were purchased from Merck (Darmstadt, Germany). The fungus, Rhizopus microsporus var. oryzae (LU 581), was purchased from the Fungal Biodiversity Centre CBS (Utrecht, The Netherlands).

Soya bean germination and challenging with Rhizopus

The soya beans were surface-sterilised by soaking them in a 1% (w/v) hypochlorite solution (5 L kg -1 beans) under continuous stirring for 1 h at 20 °C. Subsequently, the soya beans were rinsed with sterile Milli-Q water and soaked for 4 h at 40 °C in sterile Milli-Q water. After soaking, the beans were germinated in 370 mL glass jars of which the bottom was covered with Whatman filter paper (Type 520 A), humidified with sterile Milli-Q water to prevent the beans from drying out. The lid of the jars was loosely closed, and parafilm was loosely wrapped around the lid to allow air passage. The beans were incubated for 4 days at 30 °C in the dark. After germination, the beans were inoculated with Rhizopus microsporus var. oryzae (LU 581). For the fungal inoculation, a sporangiospore suspension was prepared by scraping off the sporangia from pure slant cultures of this fungus, grown on malt extract agar (CM59; Oxoid, Basingstoke, UK) for 7 days at 30 °C. Subsequently, the sporangia were suspended in sterile Milli-Q water with 0.85% (w/v) NaCl (10 8 CFU mL -1). The beans were inoculated with the sporangiospore suspension (0.2 mL g -1) and incubated for 4 days at 30 °C in the dark.

Soya bean extraction Non-soaked soya beans, soaked soya beans and Rhizopus -challenged germinated soya beans were freeze-dried, after which they were milled with a Retsch Ultra Centrifugal Mill ZM 200 (Haan, Germany) using a 1 mm sieve. The meal was defatted by hexane extraction for 30 min in a sonication bath at 30 °C (0.04 g meal mL -1 hexane) followed by air-drying of the residue. The flavonoids were extracted with absolute ethanol (0.04 g meal mL -1 EtOH) by a two-step sequential extraction of the defatted powder, each for 30 min in a sonication bath at 30 °C. The extracts were centrifuged (2500 g, 15 min at 20°C). The dried extracts were obtained after evaporation and were subsequently, dried, re-solubilised in methanol (MeOH) to a concentration of 10 mg mL -1 and stored at -20 °C. All samples were thawed and centrifuged before analysis.

62

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

RP-UHPLC analysis Samples were analysed on a Thermo Accela UHPLC system (San Jose, CA, USA) equipped with pump, autosampler and PDA detector. Samples (1 µL) were injected on a Waters Acquity UPLC BEH shield RP18 column (2.1 x 150 mm, 1.7 µm particle size) with a Waters Acquity UPLC shield RP18 Vanguard pre-column (2.1 x 5 mm, 1.7 µm particle size; Waters, Milford, MA, USA). Water acidified with 0.1% (v/v) acetic acid, eluent A, and ACN acidified with 0.1% (v/v) acetic acid, eluent B, were used as eluents. The flow rate was 300 µL min -1, the column oven temperature was controlled at 25 °C, and the PDA detector was set to measure at a range of 200-400 nm. The following elution profile was used: 0-2 min, linear gradient from 10%-25% (v/v) B; 2-9 min, linear gradient from 25%-50% (v/v) B; 9-12 min, isocratic on 50% B; 12-22 min, linear gradient from 50%-100% (v/v) B; 22-25 min, isocratic on 100% B; 25-27 min, linear gradient from 100%-10% (v/v) B; 27-29 min, isocratic on 10% (v/v) B.

Electrospray ionization mass spectrometry (ESI-MS)

Mass spectrometric data were obtained by analysing samples on a Thermo Scientific LTQ- XL (San Jose, CA, USA) equipped with an ESI-MS probe coupled to the RP-UHPLC. Helium was used as sheath gas and nitrogen as auxiliary gas. Data were collected over an m/z - range of 150-1500. Data dependent MS n analysis was performed with a normalised collision energy of 35%. The MS n fragmentation was always performed on the most intense daughter ion in the MS n-1 spectrum. Most settings were optimised via automatic tuning using “Tune Plus” (Xcalibur 2.07, Thermo Scientific). To this end, the system was tuned with genistein in both positive ionisation (PI) and negative ionisation (NI) mode. For both modes, the ion transfer tube temperature was 350 °C and the source voltage 4.8 kV. Data acquisition and reprocessing were done with Xcalibur 2.07 (Thermo Scientific).

RESULTS AND DISCUSSION

Precursors of prenylated isoflavonoids

The isoflavone aglycones daidzein ( 12 ), genistein ( 14 ) and glycitein ( 11 ), their glucoside- (1, 3 and 2, respectively) and malonyl-forms ( 6, 9 and 5, respectively) were all assigned in the EtOH extracts of the non-soaked soya beans, soaked soya beans and Rhizopus - challenged soya seedlings using UHPLC-UV-MS ( Figures 1 and 2; Table 1 ). Furthermore, two peaks were annotated as acetyl-daidzin ( 7) and acetyl-genistin (10 ). The non-soaked beans contained predominantly glycosylated and malonylated forms of daidzein and genistein. These forms were partly converted into their respective aglycones upon soaking. Of the isoflavone aglycones present in the soaked beans, daidzein ( 12 ) and

63

Chapter 3

genistein ( 14 ) were the most abundant, whereas glycitein ( 11 ) was only present in small amounts. This is consistent with previous data [4]. The changes in isoflavonoid profile upon germination and subsequent application of R. microsporus were characterised by a further decrease of the glycoside-conjugated isoflavonoids. Daidzein, genistein and glycitein became the predominant isoflavonoid compounds in the UV-profile. This has also been reported for soya seedlings that were inoculated with Aspergillus sojae [16]. The decrease of the daidzin and malonyl-daidzin content caused by deglucosylation has been proposed to increase the pool of daidzein in the soya seedling, in addition to the de novo synthesis by flavonoid metabolism, which might facilitate glyceollin production [17]. Furthermore, the accumulation of three other isoflavonoids was induced by R. microsporus . At a retention time of 6.23 min a peak ( 4) appeared that was annotated as glycinol, followed by peaks 8 and 13 that were tentatively annotated as 2’- hydroxydaidzein (7,2’,4’-trihydroxyisoflavone) and 2’-hydroxygenistein (5,7,2’,4’- tetrahydroxyisoflavone), respectively. Glycinol (3,6a,9-trihydroxypterocarpan, 4) is a pterocarpan that is formed from daidzein ( 12 ), with the 2’-hydroxylation of daidzein leading to compound 8 as first step [10]. Based on MS fragmentation analysis in positive mode, compound 13 was tentatively assigned as genistein with an extra hydroxyl group on the B-ring. The tentative assignment was supported by the comparison of the MS 2 and MS 3 spectra of compound 13 with those reported for 2’-hydroxygenistein [18]. Nevertheless, it should be mentioned that the MS n fragmentation of 3’-hydroxygenistein (orobol) has never been described. Therefore, it cannot be excluded that its fragmentation is very similar to that of 2’- hydroxygenistein. As a consequence, the extra hydroxyl group on 13 might be on either the 2’- or the 3’-position. Both hydroxylated forms of genistein have not been reported in soya. Regarding the occurrence of both genistein forms in the family of Leguminosae, the presence of orobol has been reported in only 2 species, whereas 2’-hydroxygenistein was found in more than 30 species [19]. Therefore, 13 most likely corresponds to 2’- hydroxygenistein. The formation of 2’-hydroxygenistein might be related to the conversion of daidzein to 2’-hydroxydaidzein as the first step in the formation of glycinol, in which 2’- hydroxygenistein may be formed as a by-product. Whereas 2’-hydroxydaidzein is converted to different pterocarpans, 2’-hydroxygenistein appears to accumulate as it might not be a suitable precursor for pterocarpan biosynthesis. This is supported by a study in which the enzyme NADPH:2’-hydroxydaidzein oxidoreductase was characterised [20]. In pterocarpan biosynthesis, this enzyme performs the conversion of 2’- hydroxydaidzein into 2’-hydroxydihydrodaidzein and a low conversion of 2’- hydroxygenistein to the corresponding isoflavanone was observed. Besides glycinol, another non-prenylated isoflavonoid, coumestrol ( 17 ) was formed. Coumestrol was assigned on the basis of the reference compound. In addition to the annotated isoflavonoids, the peak eluting at 10.3 min was tentatively assigned as naringenin, a non-prenylated flavanone.

64

Table 1. Compounds tentatively assigned in the extracts of Glycine max by UHPLC-ESI-MS. - 2 + 2 No Rt. UV max (nm) Identification Class Molecular [M -H] MS product ions [M+H] MS product ions (min) formula (relative abundance ) (relative abundance )

1 4.86 225,248 Daidzin Isoflavone C21 H20 O9 415 253( 100 ) 417 255( 100 )

2 4.91 225,256 Glycitin Isoflavone C22 H22 O10 445 283( 100 ) 447 283( 100 )

3 5.99 225,260 Genistin Isoflavone C21 H20 O10 431 431( 100 ), 269( 11 ) 433 271( 100 ) * 4 6.23 205,226,281 Glycinol Pterocarpan C15 H12 O5 271 256( 34 ), 227( 28 ), 161( 100 ) 255 199( 100 ), 237( 52 )

5 6.27 209,225 Malonyl -glycitin Isoflavone C25 H24 O13 531 283( 100 ) 533 285( 100 )

6 6.33 225,255 Malonyl -daidzin Isoflavone C24 H22 O12 501 253( 100 ) 503 255( 100 )

7 7.04 208,225 Acetyl -daidzin Isoflavone C23 H22 O10 457 269( 100 ) 459 271( 100 )

8 7.63 209,225,260 2' -Hydroxydaidzein Isoflavone C15 H10 O5 269 269( 100 ) 271 215( 55 ), 201( 100 ), 137( 36 )

9 7.70 209,225,260 Malonyl -genistin Isoflavone C24 H22 O13 517 269( 100 ) 519 271( 100 )

10 7.89 210,225,256 Acetyl -genistin Isoflavone C27 H30 O15 473 269( 100 ) 475 271( 100 )

11 8.46 226,257 Glycitein Isoflavone C16 H12 O5 283 268( 100 ) 285 270( 100 ), 229( 21 ), 225( 16 )

12 8.79 226,300 Daidzein Isoflavone C15 H10 O4 253 253( 100 ) 255 227( 49 ), 199( 100 ), 137( 70 )

13 8.98 212,226,257 2' -Hydroxygenistein Isoflavone C15 H10 O6 285 217( 100 ) 287 245( 49 ), 217( 100 ), 153( 57 )

14 11.33 224,260 Genistein Isoflavone C15 H10 O5 269 269( 100 ) 271 153( 100 ), 215( 79 ) * 15 11.52 209,227 ,285 Glyceollidin I/II Pterocarpan C20 H20 O5 339 324( 45 ), 161( 100 ) 323 267( 100 ) * 16 12.01 211,227,289, Glyceollin III Pterocarp an C20 H18 O5 337 319( 100 ), 293( 8), 149( 18 ) 321 306( 76 ), 297( 79 ), 251( 100 ) 354

17 12.14 226,303,343 Coumestrol Coumestan C15 H8O5 267 267( 100 ) 269 241( 100 ), 225( 22 ), 197( 23 ) * 18 12.35 211,228,283, Glyceollin II Pterocarpan C20 H18 O5 337 319( 100 ), 293( 24 ), 149 (44 ) 321 306( 74 ), 297( 100 ), 251( 94 ) 307 * 19 12.48 210,230,282 Glyceollin I Pterocarpan C20 H18 O5 337 319( 100 ), 293( 34 ), 149( 95 ) 321 306( 71 ), 303( 100 ), 293( 33 )

20 16.98 212,227 9-Z,E/E,E -KODE Oxylipin C18 H30 O3 293 275( 100 ), 235( 17 ), 185( 2), 295 n.a. 171( 9)

21 17.22 211,228 13 -Z,E/E,E -KODE Oxylipin C18 H30 O3 293 275( 100 ), 249( 12 ), 195( 29 ), 295 n.a. 179( 11 )

22 18.65 211,227,274 13 -Z,E/E,E -KODE Oxylipin C18 H30 O3 293 275( 63 ), 249( 100 ), 179( 17 ), 295 277( 100 ) 113( 54 )

23 18.80 211,228,271 9-Z,E/E,E -KODE Oxylipin C18 H30 O3 293 275( 46 ), 249( 17 ), 197( 37 ), 295 277( 100 ) 185( 100 )

24 18.88 211,227,343 Unknown KODE Oxylipin C18 H30 O3 293 n.a. 295 277( 100 )

25 19.06 211,228,268 9-Z,E/E,E -KODE Oxylipin C18 H30 O3 293 275( 20 ), 249( 11 ), 197( 41 ), 295 277( 100 ) 185( 100 ) * + 1 n.a. – not available; The [M+H-H2O] dominated in the MS compared to the [M+H]

Chapter 3

Figure 1. RP-UHPLC-UV profile of EtOH extracts of non-soaked soya beans (A), soaked soya beans (B), and Rhizopus -challenged soya bean seedlings (C). Peak numbers refer to compounds in Table 1 .

66

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

Oxylipins

Upon soaking and challenging of the soaked beans, several peaks appeared in the UV- profile eluting after 15 min (peaks 20 -25 ; Figure 1 ; Table 1 ). These peaks were annotated as KODEs. These peaks were not observed in the non-soaked soya bean extract. Based on in-source ESI fragmentation in NI mode, it was possible to distinguish between 13-KODE and 9-KODE [6]. Peaks 20 , 23 and 25 were assigned as 9-KODEs, whereas peaks 21 and 22 were assigned as 13-KODEs. Based on the MS 2 spectrum of peak 24 , it was not possible to determine the type of KODE.

Figure 2. Structures of isoflavonoids in soya beans and challenged soya seedlings.

Prenylated pterocarpans

The incubation of germinated soya beans with R. microsporus led to the accumulation of glyceollins and glyceollidins ( Figure 1 ; Table 1 ). Peaks 16 , 18 and 19 with an m/z value in NI mode of 337 were tentatively assigned as glyceollins. The area of peaks 16 , 18 and 19 had a relative ratio of approximately 1:2:6. It has been reported that glyceollin I is the most abundant of the three in an extract of pathogen-challenged soya beans, followed by glyceollin II [21]. Based on this data and by comparison of their UV spectra with those reported in literature [22], peaks 16 , 18 and 19 were tentatively assigned as glyceollins III, II and I, respectively.

67

Chapter 3

The three glyceollin peaks were preceded by peak 15 with an m/z value of 339 in NI mode, which was tentatively assigned as glyceollidin I and II, assuming that both glyceollidin I (precursor of glyceollin I) and II (precursor of glyceollin II and III) are likely to be present in an extract containing the three glyceollins, and because no other peak with this m/z value was observed close to peak 15 . The mass spectra of peaks 4, 15 , 16 , 18 and + 19 measured in PI mode were all characterised by a predominant [M+H-H2O] ion and the less abundant [M+H] + parent ion. Loss of water of glyceollins I-III in such ESI-MS 1 spectra has been reported earlier [6, 21]. We also observed that glyceollidin I/II ( 15 ) and glycinol ( 4) lost a water molecule upon protonation in the PI mode. The hydroxyl groups on positions 3 ( 4, 15 ) and 9 ( 4, 15 , 16 , 18 and 19 ) are part of the conjugated system of the aromatic ring via hyper-conjugation, unlike the 6a-hydroxyl. Therefore, the 6a-hydroxyl might be released from the parent ion as a neutral water molecule upon protonation in PI mode. The confirmation of the assignments of the glyceollins, glyceollidins and glycinol was performed by the analysis of their characteristic MS n fragmentations, obtained in NI mode. The nomenclature for RDA fragmentations suggested for isoflavones and pterocarpans was partly adopted from Ma et al . and Fabre et al. (Figure 3 ) [23, 24]. Fragmentation of the pterocarpans involves C-ring and/or D-ring cleavage, each resulting in different retro-Diels Alder (RDA) fragments. RDA pathways are proposed for glyceollin I degradation in NI mode ( Figure 4 ). Upon fragmentation of the parent ion [M-H] -, several RDA fragments were formed by cleavage of the C-ring (2,4 A-, 2,4 B-), the D-ring ( 5,6 A-, 5,6 B-) or both C-and D-ring ( 2,3,7 A-, 2,3,7 B-). Although less common for glyceollins, a water loss of the - 2 parent ion [M-H-H2O] might initiate the formation of different RDA fragments in MS , i.e. 1,4 - 6,7 - 0,4 - 2,4 - -H2O A , -H2O A , -H2O B and -H2O B +2H. Furthermore, the loss of a methyl radical was observed in the MS 2 spectra of glycinol and glyceollidin I/II. The observation of a methyl radical upon fragmentation of flavonoids is indicative for the presence of a methoxy-group [24]. The absence of a methoxy group in both glycinol and glyceollidin I/II suggests that the methyl radical originates from either the prenyl chain (glyceollidin I/II) or the carbon 6 of the pterocarpan skeleton.

Figure 3. Nomenclature of the different retrocyclization cleavages, illustrated for an isoflavone (daidzein), adopted from Ma et al . and Fabre et al . [23, 24], and for a 6a-hydroxyl-pterocarpan (glycinol) and its parent ion after a water loss.

68

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

The fragmentation behaviour of glyceollin I can be extrapolated to the other pterocarpans, as can be seen from the formation of RDA fragments and their relative abundances observed in NI mode MS 2 spectra ( Table 2 ). The fragmentation pathways of - glyceollins I-III ( 16 , 18 and 19 ) were characterised by a predominant water loss [M-H-H2O] in combination with RDA fragment 2,4 B-. On the other hand, the MS 2 spectra of glycinol ( 4) 0,4 - 6,7 - and glyceollidin I/II ( 15 ) were dominated by RDA fragments -H2O B and -H2O A and no RDA fragments originating from the fragmentation of parent ion [M-H] - were observed. This indicated that the presence of the pyran or furan ring predominantly resulted in RDA - x,y fragments that originate from the parent ion [M-H] and to a lesser extent in -H2O RDA x,y fragments, whereas their absence only results in -H2O RDA fragments (see Appendix VI ). - Apparently, a ring-closed prenyl chain inhibited further fragmentation of the [M-H-H2O] fragment ion, the mechanism of which remains unclear.

Table 2. MS 2 product ions obtained from the [M-H] - parent ion of pterocarpans 4, 15 , 16 , 18 and 19 . Pterocarpan Glycinol Glyc eollidin I/II Glyceollin I Glyceollin II Glyceollin III [M -H] - 271 ( 1)a 339 ( 1) 337 ( 4) 337 ( 5) 337 ( 2) •- [M -H-CH 3] 256 ( 37 ) 324 ( 46 ) 322 ( 1) 322 ( 1) - - [M -H-H2O] 253 ( 4) 321 ( 5) 319 ( 100 ) 319 ( 100 ) 319 ( 100 ) [M -H-CO] - 243 ( 5) 311 ( 7) 309 ( 3) 309 ( 2) 309 ( 2) - [M -H-CO 2] 227 ( 29 ) 295 ( 20 ) 293 ( 37 ) 293 ( 36 ) 293 ( 9) 2,4 A- - - 187 ( 5) 187 ( 14 ) - 2,4 B- - - 149 ( 98 ) 149 ( 43 ) 149 ( 18 ) 2,3,7 A- - - 215 ( 3) 215 ( 3) 215 ( 1) 2,3,7 B- - - 121 ( 3) 121 ( 1) 121 ( 1) 5,6 A- - - 229 ( 3) - - 5,6 B- 109 ( 2) - 109 ( 1) - - 0,4 - -H2O B 161 ( 100 ) 161 ( 100 ) 161 ( 20 ) 161 ( 7) 161 ( 3) 1,4 - -H2O A 109 ( 2) 177 ( 5) 175 ( 4) 175 ( 3) - 6,7 - -H2O A 161 ( 100 ) 229 ( 8) 227 ( 1) 227 ( 5) 227 ( 1) 2,4 - -H2O B + 2H - 133 ( 2) - - - a m/z ( relative abundance )

Assignment of other prenylated isoflavonoids

An LC-MS-based method was employed to screen for the presence of novel prenylated isoflavonoids in soya beans [25]. This method is based on characteristic neutral losses (56 u alone, or in combination with 42 u), which are related to the specific fragmentation of the prenyl-moiety in MS 2 and MS 3 spectra measured in PI mode. The ratio of the MS n + fragments corresponding to neutral losses of 42 u and 56 u, [M+H-C3H6] and [M+H- + C4H8] , respectively, indicates whether the prenyl-moiety is a prenyl chain (42:56<1) or a pyran ring (42:56>1). Furthermore, compounds that have an m/z value of less than 305 are assigned as non-prenylated.

69

Chapter 3

Figure 4. Proposed fragmentation pathways of glyceollin I in NI mode. The neutral losses are given between parentheses.

70

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

Figure 5. Screening results for neutral losses 42 u and 56 u in the PI mode MS 2 spectra. RP-UHPLC-MS base peak chromatogram in positive mode (A) of the EtOH extract of germinated soya bean ( Glycine max ) in combination with Rhizopus microsporus var. oryzae . The MS chromatograms were screened for neutral losses in MS 2 of 42 u (B) and 56 u (C). The numbers in the figure correspond to those given in Table 3 .

Based on the analysis of the chromatograms of non-soaked, soaked and Rhizopus -challenged soya bean seedling, 25 peaks showed a clear neutral loss of 56 u, alone or in combination with 42 u upon fragmentation in MS 2 (Table 3 and Figure 5 ). Of these 25 peaks, several were eliminated due to the m/z < 305 criterion or the low relative + abundance of the [M+H-C4H8] fragment ion. Twelve peaks ( S7-15 , S21-22 and S25 ) were assigned as prenylated flavonoids. MS 3 spectra did not yield additional candidate peaks.

71

Chapter 3

Peaks S11 -S15 , S21, S22 and S25 could be tentatively assigned based on the predominant fragment ions in the MS 2 and MS 3 spectra. Peaks S13 and S15 both had an m/z value of 323 in PI mode. This is 68 u more than daidzein, indicating that S13 and S15 might be prenylated daidzein molecules. A comparison of the MS 3 spectra of peaks S13 and S15 with the MS 2 spectrum of daidzein ( 12 ) showed that the spectra were quite similar: Neutral losses calculated from the ions in the MS 2 spectrum of daidzein were also 3 + found in the MS spectra of the [M+H-C4H8] fragments of S13 and S15 (see Appendix VII ). 3 + The MS fragmentation of [M+H-C4H8] resulted for S13 in a predominant RDA fragment with an m/z value of 149. This is 12 u more than the 1,3 A+ fragment from daidzein ( m/z 137). The difference of 12 u between the MS 2 fragments of daidzein and the MS 3 fragments of both A-ring and B-ring prenylated daidzein can be accounted for the remaining CH-group after the degradation of the prenyl group. The 1,3 A+ fragment from 1,3 + S13 with an m/z value of 149 is in fact -C4H8 A , suggesting that S13 is prenylated on the + 3 A-ring. For peak S15 , the fragmentation of [M+H-C4H8] in MS resulted in the formation of the predominant RDA fragment 1,3 A+ with an m/z value of 137, leading to the tentative assignment of this peak as a B-ring prenylated daidzein. This approach was also applied to the other prenylated isoflavonoids leading to the tentative assignment of these peaks as A-ring prenylated 2’-hydroxydaidzein ( S11 ), B-ring prenylated glycitein ( S12 ), A-ring prenylated 2’-hydroxygenistein ( S14 ), A-ring prenylated genistein ( S21 ), and B-ring prenylated genistein ( S22 ). The fragmentation pathways for these novel prenylated flavonoids are proposed in Figure 6 . Our study does not allow us to conclude on the position of the prenyl chain on the A-ring (6- or 8-position) or on the B-ring (2’- or 3’- positions). NMR spectroscopic analysis would be required to establish this. Several of these prenylated isoflavones have been isolated from other plant species belonging to the Leguminosae; Neobavaisoflavone, a B-ring prenylated daidzein, was isolated from Psoralea ssp. and Pueraria ssp.; Isowighteone, a B-ring prenylated genistein, has been isolated from Cajanus ssp. and Piscidia ssp.; Lupiwighteone and wighteone, both A-ring prenylated genistein molecules, and 2,3-dehydrokievitone and luteone, both A-ring prenylated 2’-hydroxygenistein molecules, were isolated from several Lupinus species [26- 31]. This is the first report on the occurrence of these prenylated isoflavonoids in soya. Peak S25 was tentatively assigned as a prenylated coumestrol. The assignment was based on the comparison of the UV and MS spectra of S25 with those of the authentic standard of coumestrol, in a similar way as described for prenylated daidzein. Since the fragmentation of coumestrol did not result in cleavage of the C-ring, assignment of the prenyl group to specifically the A-ring or B-ring based on MS alone was not possible. However, the only coumestrol molecule with a prenyl chain isolated from Glycine max is phaseol, 4-prenyl-coumestrol [32]. Thereofore, S25 was tentatively assigned as phaseol. A rough estimation, based on expressing the peak areas as daidzein equivalents, indicated that approx 40% of the total isoflavonoids content in Rhizopus -challenged soya bean seedlings were prenylated pterocarpans and 7% were prenylated isoflavones and prenylated coumestan, whereas the remainder were isoflavones.

72

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

Table 3. Screening results for prenylated flavonoids in the MS chromatograms of the EtOH extracts of non- soaked, soaked and Rhizopus -challenged soya bean seedlings ( Glycine max ) in PI mode. (C – prenyl-chain; P – pyran ring; np – non-prenylated). MS 2 No † Rt ‡ Identification Extract* [M+H] + R.A. R.A. Ratio Prenyl (min) neutral neutral (42:56) type loss 42 loss 56 S1 6.28 Glycinol RS 255 n.a. S2 6.54 n.d. RS 443 - <1 - n.a. S3 8.52 Glycitein NS/SB/RS 285 n.a. S4 8.86 Daidzein NS/SB/RS 255 n.a. S5 9.04 2’ -hydroxygenistein RS 285 n.a. S6 11.39 Genistein NS/SB/RS 271 n.a. S7 11.57 Glyceollidin I/II RS 323 2 100 0.0 C S8 12.06 Glyceollin III RS 321 80 29 2.8 ** S9 12.39 Glyceollin II RS 321 100 34 2.9 P S10 12.56 Glyceollin I RS 321 28 24 1.2 P

S11 13.83 A-ring prenylated RS 339 - 100 0.0 C 2’-hydroxydaidzein S12 14.64 B-ring prenylated RS 353 - 100 0.0 C glycitein

S13 15.17 A-ring prenylated RS 323 - 100 0.0 C daidzein

S14 15.33 A-ring prenylated RS 355 - 100 0.0 C 2’-hydroxygenistein

S15 15.49 B-ring prenylated RS 323 - 100 0.0 C daidzein S16 15.92 n.d. RS 353 1 <1 - n.a. S17 16.67 n.d. NS/SB/RS 267 n.a. S18 16.91 n.d. RS 277 n.a. S19 17.14 n.d. RS 279 n.a. S20 17.57 n.d. RS 279 n.a .

S21 17.91 A-ring prenylated RS 339 - 100 0.0 C genistein

S22 18.08 B-ring prenylated RS 339 <1 100 0.0 C genistein S23 18.22 n.d. RS 279 n.a. S24 18.52 n.d. RS 279 n.a. S25 18.99 4-prenyl -coumestrol RS 337 - 100 0.0 C *NS – non-soaked beans; SB – soaked beans; RS – Rhizopus -challenged soya seedlings; R.A. – Relative abundance of ion within MS 2 spectrum. ** Whereas glyceollins I and II are substituted with a pyran ring, glyceollin III has a furan ring; apparently the ratio (42:56) is similar for pyran and furan rings † the ‘S’ refers to a peak found with the screening method in an extract. ‡ The retenon me in MS has a delay of 0.20 ± 0.02 min compared to the UV retenon me (Rt). < 1 – Relative abundance of ion within MS 2 spectrum is 0.5% to 1.0%. - not detected, relative abundance of ion within MS 2 spectrum < 0.5%. n.d. – not determined; n.a. – not applicable.

73

Chapter 3

CONCLUSIONS

Using LC-MS, three subclasses of prenylated isoflavonoids were found to accumulate in soya bean seedlings upon challenging by Rhizopus microsporus var. oryzae : Five prenylated pterocarpans, seven prenylated isoflavones and one prenylated coumestan. The prenylated isoflavonoids were tentatively assigned as A-ring and B-ring prenylated daidzein, A-ring prenylated 2’-hydroxydaidzein, B-ring prenylated glycitein, A-ring prenylated 2’-hydroxygenistein, A-ring and B-ring prenylated genistein and 4-prenyl- coumestrol. Furthermore, the fragmentation pathways of the 5 pterocarpans in NI mode were proposed, which involves the cleavage of the C ring and/or D ring. The presence of a ring-closed prenyl group affects the fragmentation pathway of the molecules in that the - x,y H2O RDA fragments are more stable than in the absence of this group.

Figure 6. Proposed fragmentation pathways of tentatively assigned prenylated isoflavones in PI mode. The numbers inside the parentheses refer to the compounds in Tables 1 and 3. The prenyl groups on the 6, 8, 2’ and 3’-position are indicated by 6P, 8P, 2’P and 3’P, respectively.

74

Prenylated isoflavonoids in Rhizopus spp. elicited soya bean seedlings

ACKNOWLEDGEMENTS

This work was financially supported by the Food & Nutrition Delta of the Ministry of Economic Affairs, The Netherlands. We would like to thank Dr. Ir. M.J.R. Nout and Dr. Ir. P.J. Roubos-van den Hil (Wageningen University, the Netherlands) for supplying the fungal spores and their advice and Dr. H.W.M. Hilhorst (Wageningen University, the Netherlands) for his assistance in setting up the germination experiments.

REFERENCES

[1] U. Zähringer, J. Ebel, H. Grisebach, Induction of phytoaxelin synthesis in soybean - elicitor-induced increase in enzume-activitiet of flavonoid biosynthesis and incorporation of mevalonate into glyceollin. Archives of Biochemistry and Biophysics 1978 , 188 , 450-455. [2] S. M. Boué, T. E. Cleveland, C. Carter-Wientjes, B. Y. Shih, D. Bhatnagar, J. M. McLachlan, M. E. Burow, Phytoalexin-Enriched Functional Foods. Journal of Agricultural and Food Chemistry 2009 , 57 , 2614- 2622. [3] H. T. Adlercreutz, Phytoestrogens. State of the art. Environmental Toxicology and Pharmacology 1999 , 7, 201-207. [4] H. J. Wang, P. A. Murphy, Isoflavone content in commercial soybean foods. Journal of Agricultural and Food Chemistry 1994 , 42 , 1666-1673. [5] J. L. Ingham, N. T. Keen, L. J. Mulheirn, R. L. Lyne, Inducibly-formed isoflavonoids from leaves of soybean. Phytochemistry 1981 , 20 , 795-798. [6] S. Feng, L. S. Chin, K. L. Yuan, D. Huang, Fungal-stressed germination of black soybeans leads to generation of oxooctadecadienoic acids in addition to glyceollins. Journal of Agricultural and Food Chemistry 2007 , 55 , 8589-8595. [7] J. Fliegmann, G. Schüler, W. Boland, J. Ebel, A. Mithöfer, The role of octadecanoids and functional mimics in soybean defense responses. Biological Chemistry 2003 , 384 , 437-446. [8] A. Mosblech, I. Feussner, I. Heilmann, Oxylipins: Structurally diverse metabolites from fatty acid oxidation. Plant Physiology and Biochemistry 2009 , 47 , 511-517. [9] I. Prost, S. Dhondt, G. Rothe, J. Vicente, M. J. Rodriguez, N. Kift, F. Carbonne, G. Griffiths, M. T. Esquerré-Tugayé, S. Rosahl, C. Castresana, M. Hamberg, J. Fournier, Evaluation of the antimicrobial activities of plant oxylipins supports their involvement in defense against pathogens. Plant Physiology 2005 , 139 , 1902-1913. [10] S. W. Banks, P. M. Dewick, Biosynthesis of glyceollins I, II and III in soybean. Phytochemistry 1983 , 22 , 2729-2733. [11] R. S. Burden, J. A. Bailey, Structure of the phytoalexin from soybean. Phytochemistry 1975 , 14 , 1389- 1390. [12] R. L. Lyne, L. J. Mulheirn, D. P. Leworthy, New pterocarpinoid phytoalexins of soybean. Journal of the Chemical Society, Chemical Communications 1976 , 497-498. [13] U. Zähringer, E. Schaller, H. Grisebach, Induction of phytoalexin synthesis in soybean. Structure and reactions of naturally-occurring and enzymatically prepared prenylated pterocarpans from elicitor- treated cotyledons and cell-cultures of soybean. Zeitschrift Fur Naturforschung C-a 1981 , 36 , 234-241. [14] R. A. Dixon, The phytoalexin response: Elicitation, signaling and control of host gene-expression. Biological Reviews of the Cambridge Philosophical Society 1986 , 61 , 239-291. [15] B. Botta, G. D. Monache, P. Menendez, A. Boffi, Novel prenyltransferase enzymes as a tool for flavonoid prenylation. Trends in Pharmacological Sciences 2005 , 26 , 606-608. [16] S. M. Boué, F. F. Shih, B. Y. Shih, K. W. Daigle, C. H. Carter-Wientjes, T. E. Cleveland, Effect of biotic elicitors on enrichment of antioxidant properties and induced isoflavones in soybean. Journal of Food Science 2008 , 73 , H43-H49.

75

Chapter 3

[17] T. L. Graham, J. E. Kim, M. Y. Graham, Role of constitutive isoflavone conjugates in the accumulation of glyceollin in soybean infected with Phytophthora megasperma Molecular Plant-Microbe Interactions 1990 , 3, 157-166. [18] P. Kachlicki, Å. Marczak, L. Kerhoas, J. Einhorn, M. Stobiecki, Profiling isoflavone conjugates in root extracts of lupine species with LC/ESI/MSn systems. Journal of Mass Spectrometry 2005 , 40 , 1088- 1103. [19] P. M. Dewick, in The flavonoids - Advances in research since 1986 (Ed.: J. B. Harborne), Chapman & Hall, London, 1986 , pp. 117-238. [20] D. Fischer, C. Ebenau-Jehle, H. Grisebach, Phytoalexin synthesis in soybean: Purification and characterization of NADPH:2'-hydroxydaidzein oxidoreductase from elicitor-challenged soybean cell cultures. Archives of Biochemistry and Biophysics 1990 , 276 , 390-395. [21] S. M. Boué, C. H. Carter, K. C. Ehrlich, T. E. Cleveland, Induction of the soybean phytoalexins coumestrol and glyceollin by Aspergillus . Journal of Agricultural and Food Chemistry 2000 , 48 , 2167- 2172. [22] W. F. Osswald, High-performance liquid chromatography of isoflavonoid phytoalexins from soybean (Glycine max, L. Harosoy 63). Journal of Chromatography A 1985 , 333 , 225-230. [23] N. Fabre, I. Rustan, E. de Hoffmann, J. Quetin-Leclercq, Determination of flavone, flavonol, and flavanone aglycones by negative ion liquid chromatography electrospray ion trap mass spectrometry. Journal of the American Society for Mass Spectrometry 2001 , 12 , 707-715. [24] Y. L. Ma, Q. M. Li, H. van den Heuvel, M. Claeys, Characterization of flavone and flavonol aglycones by collision-induced dissociation tandem mass spectrometry. Rapid Communications in Mass Spectrometry 1997 , 11 , 1357-1364. [25] R. Simons, J.-P. Vincken, E. J. Bakx, M. A. Verbruggen, H. Gruppen, A rapid screening method for prenylated flavonoids with ultra-high-performance liquid chromatography/electrospray ionisation mass spectrometry in licorice root extracts. Rapid Communications in Mass Spectrometry 2009 , 23 , 3083-3093. [26] M. Stobiecki, P. Wojtaszek, Applications of gas-chromatography mass-spectrometry to the identification of isoflavonoids in lupine root extracts. Journal of Chromatography 1990 , 508 , 391-398. [27] G. A. Lane, O. R. W. Sutherland, R. A. Skipp, Isoflavonoids as insect feeding deterrents and antifungal components from root of Lupinus angustifolius Journal of Chemical Ecology 1987 , 13 , 771-783. [28] G. A. Lane, R. H. Newman, Isoflavones from Lupinus angustifolius root. Phytochemistry 1987 , 26 , 295- 300. [29] J. L. Ingham, S. Tahara, S. Shibaki, J. Mizutani, Isoflavonoids from the root bark of Piscidia erythrina and a note on the structure of piscidone. Zeitschrift Fur Naturforschung 1989 , 44c , 905-913. [30] Y. Hashidoko, S. Tahara, J. Mizutani, Isoflavonoids of yellow lupin 1. New complex isoflavones in the root of yellow lupin ( Lupinus luteus L. cv barpine). Agricultural and Biological Chemistry 1986 , 50 , 1797-1807. [31] S. J. Dahiya, Reversed-phase high-performance liquid chromatography of cajanus cajan phytoalexins. Journal of Chromatography A 1987 , 409 , 355-359. [32] P. Caballero, C. M. Smith, F. R. Fronczek, N. H. Fischer, Isoflavones from an insect-resistant variety of soybean and the molecular structure of afrormosin. Journal of Natural Products 1986 , 49 , 1126-1129.

76

Chapter 4

Agonistic and antagonistic estrogens in licorice root ( Glycyrrhiza glabra )

Based on : Rudy Simons, Jean-Paul Vincken, Loes A.M. Mol, Susan The, Toine F.H. Bovee, Teus J.C. Luijendijk, Marian A. Verbruggen, Harry Gruppen, Agonistic and antagonistic estrogens in licorice root ( Glycyrrhiza glabra ), Analytical and Bioanalytical Chemistry , 2011 , in press

Chapter 4

ABSTRACT

The roots of licorice ( Glycyrrhiza glabra ) are a rich source of flavonoids, in particular prenylated flavonoids, such as the isoflavan glabridin and the isoflavene glabrene. Fractionation of an ethyl acetate extract from licorice root by centrifugal partitioning chromatography yielded 51 fractions, which were characterized by LC-MS and screened for activity in yeast estrogen bioassays. One third of the fractions displayed estrogenic activity towards either one or both estrogen receptor (ER α and ER β). Glabrene-rich fractions displayed an estrogenic response, predominantly to the ERα. Surprisingly, glabridin did not exert agonistic activity to both ER subtypes. Several fractions displayed higher responses than the maximum response obtained with the reference compound, the natural hormone 17 β-estradiol (E 2). The estrogenic activities of all fractions, including this so-called superinduction, were clearly ER-mediated, as the estrogenic response was inhibited by 20-60% by known ER antagonists, and no activity was found in yeast cells that did not express the ER α or ER β subtype. Prolonged exposure of the yeast to the estrogenic fractions that showed superinduction did, contrary to E 2, not result in a decrease of the fluorescent response. Therefore, the superinduction was most likely the result of stabilization of the ER, yEGFP, or a combination of both. Most fractions displaying superinduction were rich in flavonoids with single prenylation. Glabridin displayed ER α- selective antagonism, similar to the ER α-selective antagonist RU 58668. Whereas glabridin -6 was able to reduce the estrogenic response of E 2 by approximately 80% at 6 × 10 M, glabrene-rich fractions only exhibited agonistic responses, preferentially on ERα.

Keywords : Prenylation, isoflavonoids, estrogen, antagonism, licorice

78

Agonistic and antagonistic estrogens in licorice root

INTRODUCTION

Flavonoids are a broad class of phenolic compounds mainly found in plants with a wide range of bioactivities [1]. Prenylated flavonoids in particular are of interest with respect to bioactivity, as prenylation is considered to modulate the responses towards the estrogen receptor [1, 2]. Prenyl-substitution of the flavonoid subclasses flavones, flavanones and flavonols has been linked to an increased affinity to the estrogen receptor α ERα [3-5], e.g. prenylation of the 8-position of the flavanone naringenin resulted in a 200-1000 higher estrogenic activity [6]. Furthermore, the prenylation of isoflavonoids has been suggested to induce antagonistic activity when binding to ERα [5, 7, 8]. Licorice roots ( Glycyrrhiza glabra ) are a rich source of prenylated flavonoids. They might offer opportunities for the development of new food supplements related to e.g. the alleviation of osteoporosis and menopausal complaints [9]. Approximately 75 prenylated flavonoids have been identified, mainly belonging to isoflavans, isoflavenes and flavanones [10, 11]. Previous studies have shown that licorice root extracts have estrogenic activity towards the ERα and ERβ [12, 13]. The key estrogenic compounds isolated from G. glabra were identified as glabrene and glabridin, both prenylated isoflavonoids [14, 15]. The estrogen-like activities of both compounds have been established by means of competitive ligand binding assays, in vitro cell assays, and in vivo animal models [16, 17]. It has been demonstrated that glabrene and glabridin bind to the ER with EC 50 values of 5 ×10 -5 M and 5×10 -6 M, respectively. These values were obtained using an MCF-7 cell line that is known to express the ERα type mainly (no detectable ER β amounts on the protein level), indicating that both compounds were agonists [18, 19]. However, the specific estrogenic potencies of glabrene and glabridin towards ERα and ERβ, and their potential antagonistic activities, have not yet been investigated. Such information is vital for understanding their specific estrogenic activity in the human body. The aim of the present study was to determine the predominant estrogenic compounds of licorice roots that are active on both ER-subtypes, and investigate their agonistic and antagonistic potencies. To this end, fractions of a licorice root extract obtained by centrifugal partitioning chromatography were characterized by LC-MS and subsequently screened for (anti)estrogenic activity using yeast estrogen bioassays.

79

Chapter 4

EXPERIMENTAL

Materials

The roots of Glycyrrhiza glabra , collected in Afghanistan, were provided by Frutarom US (North Bergen, NJ, USA). Estradiol was purchased from Sigma Aldrich (St. Louis, MO, USA), glabridin from Wako Chemicals GmbH (Neuss, Germany). RU 58668 and R,R-diethyl-THC (R,R-THC) were purchased from Tocris Bioscience (Bristol, United Kingdom). AR grade n- hexane, acetone and absolute ethanol and ULC/MS grade acetonitrile (ACN) were purchased from Biosolve BV (Valkenswaard, The Netherlands). Water was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, USA). Dimethylsulfoxide (DMSO) and all other chemicals were purchased from Merck (Darmstadt, Germany).

Preparation of licorice extract

The roots were milled with a ZM 200 Retsch Ultra centrifugal mill (Haan, Germany) using a 1 mm sieve. The root powder was extracted with ethyl acetate (EA) in a ratio of 1 to 25 (w/w) for 2 h at 40 °C under continuous stirring. The extract was obtained by pressing the mixture with a Fischer Maschinenbau hydraulic press type HP 5M (Gemmrigheim, Germany) under 40 bar for 1 h. The dried extract was obtained after evaporation of the EA under reduced pressure at 40 °C.

CPC fractionation of licorice extract

Centrifugal partitioning chromatography (CPC) was performed using a thermostatted Kromaton FCPC machine (Angers, France) connected to an Armen AP 100 (Chromtech, Boronia, Victoria, Australia) plunger pump. The two-phase solvent system used consisted of n-hexane/acetone/water in a ratio of 5:9:1 (v/v/v). It was equilibrated under stirring at 22 °C for at least one hour. Small-scale fractionations as part of the method development were done with a 200 mL rotor in ascending mode (i.e. lower phase is stationary phase) at 22 °C, a rotation speed of 1000 rpm, and a flow rate of 10 mL/min. The volume of displaced stationary phase was approximately 83 mL. Eighty five mg dried extract was dissolved in a mixture of upper and lower phase, 4 mL of each phase. The fractionation process was monitored using a Jasco UV-2075 UV detector equipped with a 1 mm preparative cell at a wavelength of 330 nm (absorbance is expressed as relative response to the highest peak). For the actual fractionation of the licorice root extract, a 1000 mL rotor was used (22 °C; rotation speed 1100 rpm; flow rate 25 mL/min). The volume of displaced stationary phase in the 1000 mL rotor was approximately 625 mL. Seven hundred fifty mg dried

80

Agonistic and antagonistic estrogens in licorice root

extract was dissolved in 28 mL of a mixture of upper and lower phase (1:1). Seven subsequent runs were performed that resulted in 51 fractions per run; the fraction size was 50 mL. Based on the CPC UV profile, corresponding fractions were combined and evaporated in combination with lyophilization, in order to remove solvents. The combined fractions were resolubilized in absolute ethanol (EtOH) and stored at -20 °C. All samples were thawed and centrifuged before analysis. Fractions collected were analyzed by UHPLC-MS at a concentration of 1 mg/mL.

RP-UHPLC

Samples were analyzed using an Accela UHPLC system (Thermo Scientific, San Jose, CA, USA) equipped with pump, autosampler and PDA detector. Samples (1 µL) were injected onto an Acquity UPLC BEH C18 column (2.1 x 150 mm, 1.7 µm particle size) with an Acquity UPLC BEH C18 Vanguard pre-column (2.1 x 5 mm, 1.7 µm particle size; Waters, Milford, MA, USA). Eluents were water acidified with 0.1% (v/v) acetic acid (eluent A), and acetonitrile (ACN) acidified with 0.1% (v/v) acetic acid (eluent B). The flow rate was 300 µL/min and the PDA detector was set to measure at a range of 205-400 nm. The following elution profile was used: 0-18 min, linear gradient from 10%-100% (v/v) B; 18-22 min, isocratic on 100% B; 22-23 min, linear gradient from 100%-10% B 23-25 min, isocratic on 10% B.

Electrospray ionization mass spectrometry (ESI-MS)

Mass spectrometric data were obtained by analyzing samples on an ion trap LTQ-XL (Thermo Scientific) equipped with an ESI-MS probe coupled to the RP-UHPLC. Helium was used as sheath gas and nitrogen as auxiliary gas. Data were collected over an m/z -range of 150-1500. Data dependent MS n analysis was performed with a normalized collision energy of 35%. The MS n fragmentation was always performed on the most intense daughter ion in the MS n-1 spectrum. Most settings were optimized via automatic tuning using “Tune Plus” (Xcalibur 2.0.7, Thermo Scientific). To this end, the system was tuned with glabridin in both positive ionization (PI) and negative ionization (NI) mode. In both modes, the ion transfer tube temperature was 350 °C and the source voltage 4.8 kV. Data acquisition and reprocessing were done with Xcalibur 2.0.7. Mass spectral data interpretation and peak determination were performed with Mass Frontier 5.0 (Highchem, Bratislava, Slovakia).

Determination of estrogenic activity

The protocols for the yeast estrogen bioassays were adopted from Bovee et al. [20] with slight modifications. The genetically modified yeast strains have a strong constitutive

81

Chapter 4

expression vector stably integrated in the genome to express either the human estrogen receptor α (ERα) or the human estrogen receptor β (ERβ). The yeast genome also contains a reporter construct. This reporter construct contains an inducible yeast enhanced green fluorescent protein (yEGFP) regulated by the activation of a minimal promoter with estrogen responsive elements (EREs). Cultures of the yeast estrogen biosensor with either ERα or ERβ were grown overnight at 30 °C with shaking at 200 rpm. At the late log phase, the cultures of both estrogen receptors were diluted in the selective minimal medium (MM) supplemented with either leucine (ERα) or histidine (ERβ) to an OD value (630 nm) of 0.05±0.01 (ERα) and 0.15±0.05 (ERβ). For exposure, 200 μL aliquots of these diluted yeast cultures were combined with 2 μL of test compound or extract (in various concentrations) in a 96-well plate to test the agonistic properties of these compounds.

DMSO (blank) and control samples containing 17 β-estradiol (E 2) or genistein dissolved in DMSO were included in each experiment. Dilution series of each sample were prepared in DMSO and the final concentration of DMSO in the assay did not exceed 1% (v/v). Each sample concentration was assayed in triplicate. Exposure was performed for 24 h or 6 h for the ER α or ER β asssay, respectively, at 30 °C and orbital shaking at 200 rpm. Fluorescence and OD were measured at 0 and 24 h for the ERα and 0 and 6-8 h for the ER β in a Tecan Infinite F500 (Männedorf, Switzerland), using an excitation filter of 485 nm (bandwidth 20 nm) and an emission filter of 535 nm (bandwidth 35 nm). The fluorescence signals of the samples were corrected with the signal obtained with the diluted yeast suspension at t 0 h (background signal). In order to check the viability of the yeast in each well, the absorbance was measured at 630 nm. Each fraction was tested in a concentration series ranging from 0.1 - 100 µg/mL. EC 50 calculations were performed in Sigma Plot (8.02, SPSS Inc.). In a number of cases, a concentration of 10 µg/mL for the ERα and 3 µg/mL for the ERβ resulted in decreased yeast growth during incubation of more than 50% during incubation due to cytotoxicity. This cytotoxicity could be due to the anti- microbial properties of licorice root constituents as reported before [21, 22]. A dilution-series of estradiol and genistein were used as reference controls in this bioassay. The EC 50 values in the ERα bioassay were 0.86 nM and 1.73 µM for estradiol and genistein, respectively, and 0.12 nM and 9.1 nM in the ERβ bioassay, respectively. All EC 50 values were in line with those reported previously [20].

For the determination of ER antagonism, the yeast cells were exposed to the EC 70 (ERα) or the EC 90 (ERβ) of estradiol in combination with different dilutions of a test compound or fraction (measured in triplicates). As a positive control, the yeast cells were exposed to the EC 70 of estradiol in combination with the known ERα antagonist RU 58688 [23] . For ERβ antagonism, the yeast cells were exposed to the EC 90 of estradiol in combination with R,R-THC, a known antagonist on the ERβ [24]. In addition to the yeasts expressing the yEGFP reporter gene in combination with either ERα or ERβ, a third yeast strain was used that only contained the reporter gene but not the vector with the ER. This yeast strain was used as a negative control.

82

Agonistic and antagonistic estrogens in licorice root

RESULTS AND DISCUSSION

Estrogenic activity of CPC fractions for ER ααα and ER βββ.

CPC fractionation of the licorice root extract resulted in 51 fractions ( Figure 1 ). After the 51 st fraction no UV response was observed anymore. Fractions F1 to 5, F6 to 21 and F22 to 51 comprised ~25, ~40 and ~35% DW of the total extraction yield, respectively (see Appendix VIII ). Most fractions showed some estrogenicity on both ERs, indicating the presence of phytoestrogens ( Table 1 ). A compound is considered a phytoestrogen, when it activates the ER at 4 [25] concentrations ≤10 times than that of estradiol (E 2) . The EC 50 value of E 2 towards the ERα in the yeast assay was determined to be 1.0-1.6×10 -9 M, which corresponds to 2.7- -4 4.4×10 µg/mL. Therefore, only CPC fractions giving a response above the EC 50 at a dilution below 3 µg/mL were indicated in as active towards ER α in Figure 1 . The application of this threshold value for the ERα resulted in 9 active fractions out of 51 (see Table 1 ). -10 -10 The EC 50 value of E 2 towards the ERβ ranged from 1.1×10 to 2.1×10 M, -5 corresponding to an EC 50 of 3.2-5.9×10 µg/mL. Therefore, only CPC fractions giving a response above the EC 50 at a dilution below 0.3 µg/mL were indicated as active towards ERβ in Figure 1 . The application of this threshold value for the ER β resulted in 12 active fractions out of 51 ( Table 1 ).

Figure 1. UV-profile of the licorice root extract fractionation by CPC. Estrogenically active fractions are indicated.

83

Table 1. Estrogenic response of CPC-fractions from the EA extract of licorice roots. The estrogenicity towards the ERα and the ERβ was measured at two concentrations. The activity was standardized to the maximum induced response of 2 µM estradiol (100%). Values are the mean ± StDev (n=3). Estrogenic fractions are marked 3 µg/mL 10 µg/mL 0.3 µg/mL 1.0 µg/mL 3 µg/mL 10 µg/mL 0.3 µg/mL 1.0 µg/mL Fr 1 ERα ERα ERβ ERβ Fr ERα ERα ERβ ERβ * E2 100 100 F26 38.7(±1.9) 93.7(±6.1) 95.5(±3.4) 151.3(±15.1) F1 n.d. 2 n.d. 0.0(±2.0) 0.0(±2.0) F27 * 36.6(±0.3) 90.8(±2.1) 74.6(±5.9) 176.5(± 11.8) F2 n.d. 1.1(±1.2) 0.0(±6.5) 0.0(±4.9) F28 * 19.4(±1.9) 48.2(±2.7) 68.7(± 4.0) 104.4(± 8.3) F3 n.d. 4.9(±0.7) 1.6(±1.5) 0.0(±4.5) F29 * 19.9(±1.3) 23.0(±3.9) 57.2(±12.5) 93.5(±0.6) F4 n.d. 7.8(±0.8) 13.4(±5.0) 22.4(±7.7) F30 * 23.8(±2.7) 22.4(±1.8) 57.3(±1.1) 48.2(±2.8) F5 3.0(±2.3) 7.8(±0.3) 12.0(±5.1) 19.5(±2.2) F31 25.0(±1.5) 34.2(±1.9) 25.8(±8.5) 40.5(±17.8) F6 n.d. n.d. n.d. n.d. F32 * 26.8(±2.3) 34.5(±2.0) 57.4(±4.7) 47.4(±10.7) F7 17.1(±3.0) 21.0(±2.7) 21.4(±5.6) 12.2(±2.2) † F33 * 32.0(±1.4) 38.0(±0.4) 50.2(±9.5) 36.0(±3.2) F8 10.5(±2.2) 21.8(±6.3) 17.2(± 2.3) 58.3(±3.4) F34 29.6(±5.3) 44.9(±1.3) 27.7(±8.8) 46.1(±5.2) F9 13.3(±1.5) 7.1(±6.4) 37.8(±10.0) 49.1(±21.5) F35 30.8(±3.1) 45.5(±4.2) 43.6(±7.6) 37.5(±21.0) F10 9.7(±3.3) 9.3(±2.0) 23.0(±9.5) 57.3(±12.9) F36 42.8(±2.3) 42.9(±6.5) 19.9(±2.3) 45.3(±2.5) F11 n.d. n.d. 46.3(±12.0) n.d. † F37 13.6(±1.6) 25.3(±2.5) 24.8(±2.6) 35.1(±9.9) F12 15.6(±2.9) 15.7(±3.2) † 30.7(±0.1) 77.2(±7.3) F38 14.8(±1.7) 28.8(±0.5) 14.2(±6.8) 30.8(±15.4) F13 n.d. n.d. 40.3(±5.0) 2.4(±4.9) † F39 19.8(±1.6) 32.4( ±2.6) 21.1(±4.4) 45.6(±14.7) F14 4.8(±4.7) n.d. 42.7(±6.6) 73.3(± 4.1) F40 25.6(±4.3) 33.9(±2.3) 26.5(±5.5) 35.4(±17.4) F15 n.d. n.d. 28.4(±7.1) 77.7(± 1.8) F41 * 159.9(±9.5) 186.9(±15.0) 28.3(±4.9) 34.9(±13.5) F16 * 74.8(±4.0) 20.4(±28.9) † 40.7(± 1.0) 62.6(± 2.1) F42 21.6(±6.0) 26.2(±2.2) 19.8(±11.0) 53.8(±25.5) F17 * 109.1(±2.3) 49.3(±4.4) † 36.7(±9.5) 88.8(± 8.2) F43 26.5(±1.8) 24.7(±4.3) 25.4(±1.2) 38.9(±20.0) F18 * 105.3(±6.2) 104.6(±10.5) † 57.1(±9.4) 98.8(± 15.1) F44 18.9(±1.9) 24.6(±1.5 ) 22.7(±7.3) 33.8(±12.2) F19 * 89.1(±4.2) 131.2(±16.2) 48.3(±4.6) 123.8(±17.1) F45 20.9(±1.5) 21.0(±4.3) 19.6(±3.4) 66.3(±15.1) F20 * 60.2(±3.1) 140.5(±10.7) 60.5(±14.5) 116.7(±14.3) F46 16.9(±1.2) 20.5(±3.2) 18.0(±11.6) 73.8(±7.3) F21 * 71.9(±8.4) 125.3(±1.2) 45.2(±3.8) 145.9(±25.2) F47 21.2(±2.9) 23.3(±4.2) 31.0(±5.1) 79.4(±6.1) F22 * 101.4(±11.6) 134.1(±2.7) 87.0(±10.3) 144.4(±10.7) F48 23.2(±4.6) 21.7(±2.4) 26.8(±2.5) 49.1(±17.8) F23 16.0(±0.9) 83.0(±9.9) 26.3(±8.7) 84.8(±7.0) F49 26.4(± 0.6) 26.7(±0.3) 16.5(±3.0) 82.9(±4.9) F24 * 79.6(±1.6) 156.3(±14.4) 103.1(±2.3) 159.3(±23.1) F50 18.8(±2.7) 16.7(±3.2) 7.3(±9.3) 41.3(±3.1) F25 * 44.2(±5.7) 108.6(±3.9) 97.9(±10.6) 172.8(±20.7) F51 22.0(±3.2) 15.6(±2.5) 18.8(±2.7) 40.5(±18.3) 1 Fraction; 2 n.d. –not detected (estrogenicity values were zero or slightly negative) * Estrogenically active; † inhibited yeast growth due to cytotoxicity

Agonistic and antagonistic estrogens in licorice root

The screening for estrogenicity of the CPC-fractions on both receptor subtypes showed that the estrogenic response of several fractions substantially exceeded the maximum response of E 2 (Table 1 ). This phenomenon has been referred to as superinduction [26]. In our study, this superinduction was observed with both receptors and appeared more pronounced for ERβ. The mechanism that leads to superinduction is not well understood, but sometimes occurs with colored extracts. Such colored extracts can disturb the fluorescent measurement, as due to a decrease of the pH during the exposure period, the color can change as well. To determine whether fractions gave an increased fluorescent response as a result of acidification (change of pH 5.0 to pH 2.9) of the culture medium due to yeast growth, six representative fractions ( F4 , F13 , F22 , F27 , F30 and F44 ), with no, moderate or high estrogenic activity, were measured at different pH values in the absence of yeast. No altered fluorescent signals were observed compared to the blank, showing that the observed superinduction was not related to altered fluorescent signals due to a drop in pH. In the next series of experiments, two subtype selective antagonists were used to determine whether the observed estrogenic activities, including the superinduction, were ER-mediated. First, RU 58668 (ERα-selective) [27] and R,R-THC (ERβ-selective) [24] were tested in the yeast estrogen bioassays to confirm their antagonistic properties. Co- -6 incubation of E 2 and RU 58668 showed that 6×10 M RU 58668 was able to decrease the E2-induced response of ERα by ~60% ( Figure 2A ). RU 58668 itself showed a weak agonistic activity of ~12% in concentrations above 1×10 -5 M. RU 58668 is known as a pure ER α antagonist, but its weak agonistic activity in the yeast estrogen bioassay with ER α is not expected, as several other 11β-analogues of E 2 were shown to be selective estrogen receptor modulators (SERM) with both agonistic and antagonistic activity on the ERα in the same yeast estrogen bioassay [23].

Table 2. Inhibition of estrogenic response of representative CPC-fractions by addition of a subtype-specific antagonist. Inhibition of the estrogenic response of 6 representative CPC fractions were determined by co- incubation at 1 µg/mL with 6×10 -6 M RU 58668 for the ERα, and 1 µg/mL with 2×10 -8 M R,R-THC for the ERβ. Values are the mean ± StDev (n=3). Estrogenic Inhibition by Estrogenic Inhibition by activity ERα RU 58668 activity ERβ R,R-THC

E2 100 60% 100 55% F4 n.d. - 13.4(±5.0) 70% F13 n.d. - 40.3(±5.0) 37% F22 101.4(±11.6) 52% 87.0(±10.3) 34% F27 36.6(±0.3) 21% 74.6(±5.9) 61% F30 23.8(±2.7) 21% 57.3(±1.1) 48% F44 18.9(±1.9) 30% 22.7(±7.3) 70%

85

Chapter 4

-7 Co-incubation of E 2 and R,R-THC showed that 2×10 M R,R-THC was able to decrease the E 2-induced response of ERβ by ~55% (Figure 2B ). Meyers and co-workers reported an inhibition of ~100% at similar concentrations while using mammalian cell based assays [24]. To determine whether the observed estrogenic responses of the fractions were ER-mediated, selected fractions were co-incubated with the subtype-selective antagonists. Besides, the fractions were tested in the yeast strain that expresses no estrogen receptor and only contains the reporter construct. In all estrogen active fractions, the responses were inhibited by either RU 58668 or R,R-THC and the fluorescence response was reduced by up to 70%. This confirms that the estrogenic responses caused by the fractions on both receptors were ER-mediated ( Table 2 ). Also the controls with the yeast strain expressing no ER confirmed that the observed responses were ER-mediated, as no fluorescent signal was observed for any of the fractions or E 2.

Superinduction by stabilization of ER-mediated response

The phenomenon of superinduction has been previously observed in several assay types and the superinduction caused by genistein in human U2OS bone cells transfected with the ERα and a luciferase reporter gene was intensively investigated [26]. It was concluded that this superinduction was caused by a post-translational stabilization of the firefly luciferase reporter enzyme by genistein and not by stabilization of the ER α. To verify the hypothesis that superinduction in the yeast was caused by the stabilization of the ER and/or the yEGFP, the yeast expressing ERβ was co-incubated with E 2, genistein or the representative fractions ( F4 , F13 , F22 , F27 , F30 and F44 ) mentioned before. The estrogenic responses were measured after 6 and 24 h ( Figure 3 ). After 6 h, both E 2 and genistein showed the maximum estrogenic response, but as expected the estrogenic response of E 2 completely disappeared after 24 h. Also, the response of genistein completely disappeared, whereas the estrogenic response of the fractions was similar or even higher compared to their response measured after 6 h. This strongly indicates that the responses, including the superinduction, of the fractions were stabilized. Our results do not allow speculation on whether the ER, the yEGFP or both proteins were stabilized, but the observed estrogenic responses were without doubt ER-mediated.

86

Agonistic and antagonistic estrogens in licorice root

Figure 2. Transcription activation by ERα (A) and ERβ (B) in response to E 2 and sub-type selective antagonists RU58668 (ERα) and R,R-THC (ERβ). The antagonistic activity of both receptor-specific antagonists were assayed in the presence of the EC 70 (0.8 nM) and EC 90 (0.2 nM) E 2 for ERα and ERβ, respectively. Both graphs were normalized to E 2. The response of R,R-THC on ERβ for every concentration was lower than the minimum response of E 2. This was corrected by normalizing the lowest concentration of R,R-THC to 0%. Values are the mean ± StDev (n=3).

87

Chapter 4

LC-MS characterization of licorice fractions

Because the estrogenic responses were ER-mediated, the licorice fractions were subjected to characterization by LC-MS. Fractions F6-21 contained the main flavonoids previously annotated in the EA extract of licorice roots ( Table 3 , Figure 4 ) [11]. These flavonoids have been annotated based on UV and MS n spectra and, if possible, compared to spectra published in the literature. The estrogenic activity of a number of active fractions ( F20- 22 ,24 -30 ,32 -33 ,42 ) could not be traced back to individual components (see Appendix VIII ). Compositional analysis by LC-MS showed that the active fractions were complex mixtures, indicating that further purification of CPC fractions is prerequisite for the identification of the estrogenic compounds. In most cases, it was not possible to assign the identity of the predominant peaks by UV and MS n. Furthermore, in most fractions the majority of the compounds were prenylated, which might suggest a correlation between prenylation and superinduction ( Table 3 , Appendix VIII ). Fractions F16 -20 , rich in glabrene, showed a predominant estrogenic activity on the ERα. This is in agreement with the fact that glabrene is considered one of the principal estrogenic components of licorice root. The estrogenic potency of glabrene for both ER- subtypes, however, has not yet been established. Our results indicate that the glabrene- rich fractions had a particularly high response towards ERα. Whereas phytoestrogens generally have a more pronounced affinity to the ERβ compared to the ERα, the glabrene- rich fractions showed the opposite behavior, similar to that of 8-prenyl naringenin [27].

Figure 3. Stabilizing effect of E2, genistein and several fractions obtained from the licorice root extract on the -10 -7 relative activity measured after 6 and 24 h in the ERβ assay. E 2, 2 × 10 M estradiol; Gen, 2 × 10 M genistein; F4 to F44, licorice root fractions obtained by CPC measured at 0.3 µg/mL. *, negative response.

88

Table 3. Fractions generated by CPC-fractionation and the presence of the main flavonoids in each fraction determined by UHPLC-MS.

Fraction 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Yield (mg) 110 141 97 136 108 126 211 345 254 156 66 176 157 115 37

Number of peaks 20 6 12 25 22 18 7 4 6 6 15 14 17 30 42

Of which are: - non-prenylated 3 1 2 4 2 1 0 0 0 0 0 1 2 4 9 - prenylated 17 5 10 21 20 17 7 4 6 6 15 13 15 26 33

13 3 7 9 6 8 4 1 2 3 11 11 13 18 22 Single prenylated -Chain 7 2 4 6 4 5 2 0 1 2 7 5 9 13 21 -Pyran 6 1 3 3 2 3 2 1 1 1 4 6 4 5 1

Double prenylated 4 2 3 12 14 9 3 3 4 3 4 2 2 8 11

Characterization Glabrene ++ ++ ++ ++ +

Glabrone + + ++ +

Glabridin + + ++ ++ + + + Glabrol + ++ ++ + + 3’-OH-4’OMe-G1 + + ++ ++ 2 4’-OMe-G + + ++ + Hispaglabridin A + ++ ++

Hispaglabridin B ++ ++ +

1 4’-O-methyl-glabridin. 2 3’-hydroxy-4’-O-methyl-glabridin.

Chapter 4

In addition to glabrene, the estrogenic activity of licorice roots extract has been ascribed to the presence of glabridin and its derivatives. Despite the abundance of glabridin in F11 -15 , no significant estrogenic response on both ER subtypes was observed. -6 EC 50 values of 5×10 M have been reported for glabridin, using different mammalian proliferation assays [15-17]. Furthermore, several glabridin derivatives were shown to be moderately estrogenic compared to glabridin [14]. In our study, the pure reference standard of glabridin did not exert any estrogenic response in a concentration range of 1×10 -7 to 1×10 -4 M towards both ER subtypes (data not shown). In concentrations above 1×10 -4 M glabridin was toxic to the yeast cells. Because glabridin had been shown to interact with the ERs, and because it is known that different ER-based bioassays can generate different output, glabridin as well as the glabrene-rich fraction F18 (due to the lack of a glabrene reference standard) were tested for their antagonistic properties.

Figure 4. Main flavonoids identified in CPC fractions 6-21 obtained from the EA extract of licorice root.

90

Agonistic and antagonistic estrogens in licorice root

Antagonistic activity of glabridin and glabrene

Prenylation of isoflavonoids has been suggested to induce antagonism towards the ERα [5, 7, 8]. The glabrene-rich fraction F18 did not show antagonistic activity on both ER subtypes

(no further data shown), but increased the estrogenic response upon co-exposure with E 2, confirming its agonistic character. The reference standard of glabridin did not have antagonistic properties towards the ERβ (data not shown), but was shown to be an ERα-selective antagonist ( Figure 5 ). At -6 a concentration of 6×10 M, glabridin was able to inhibit the E 2 response by ~80% without being toxic towards the yeast cells. The agonistic activity of glabridin in the MCF-7 proliferation assay, in in vivo animal models, and the ERα-selective antagonistic activity in the yeast estrogen bioassay might imply that glabridin acts as a SERM. The estrogenic activity of glabridin is similar to kievitone and phaseollin, a prenyl-chain substituted isoflavanone and a pyran-ring substituted pterocarpan, respectively. Both kievitone and phaseollin displayed agonistic activity in the MCF-7 proliferation assay and in the human HEK 293 transactivation assay, but were antagonistic in the MCF-7 colony formation assay [28]. The assay-dependent mode of action of glabridin has also been observed with other compounds. For example, both tamoxifen and 4-hydroxy-tamoxifen act as SERMs, displaying both estrogenic and anti-estrogenic activities in mammalian breast and endometrial cells, act as agonists in yeast estrogen bioassays [27]. As mentioned before, LC- MS characterization showed that fractions F6 -15 were rich in glabridin derivatives. These compounds share prenyl-substitution on the A-ring with a pyran ring ( Table 3, Figure 4 ). It will be worthwhile to also test the purified glabridin derivatives for antagonistic activity in the yeast estrogen bioassay.

Figure 5. Antagonistic activity of glabridin on the estrogenic response in the ERα yeast estrogen bioassay.

91

Chapter 4

ACKNOWLEDGEMENTS

This work was financially supported by the Food & Nutrition Delta of the Ministry of Economic Affairs, The Netherlands. We would like to thank Daniel Piscitello, Florian Wiegand and Miroslav Bolardt (Frutarom Switzerland Ltd., Wädenswil, Switzerland) for their assistance in preparing the licorice extract and RIKILT-Institute of Food Safety for the yeast estrogen bioassays.

REFERENCES

[1] D. Barron, R. K. Ibrahim, Isoprenylated flavonoids - A survey. Phytochemistry 1996 , 43 , 921-982. [2] B. Botta, G. D. Monache, P. Menendez, A. Boffi, Novel prenyltransferase enzymes as a tool for flavonoid prenylation. Trends in Pharmacological Sciences 2005 , 26 , 606-608. [3] M. Kitaoka, H. Kadokawa, M. Sugano, K. Ichikawa, M. Taki, S. Takaishi, Y. Iijima, S. Tsutsumi, M. Boriboon, T. Akiyama, Prenylflavonoids: A new class of non-steroidal phytoestrogen (Part 1). Isolation of 8-isopentenylnaringenin and an initial study on its structure- activity relationship. Planta Medica 1998 , 64 , 511-515. [4] B. Pinto, A. Bertoli, C. Noccioli, S. Garritano, D. Reali, L. Pistelli, Estradiol-antagonistic activity of phenolic compounds from leguminous plants. Phytotherapy Research 2008 , 22 , 362-366. [5] E. M. Ahn, N. Nakamura, T. Akao, T. Nishihara, M. Hattori, Estrogenic and antiestrogenic activities of the roots of Moghania philippinensis and their constituents. Biological and Pharmaceutical Bulletin 2004 , 27 , 548-553. [6] F. Roelens, N. Heldring, W. Dhooge, M. Bengtsson, F. Comhaire, J. Å. Gustafsson, E. Treuter, D. De Keukeleire, Subtle side-chain modifications of the hop phytoestrogen 8-prenylnaringenin result in distinct agonist/antagonist activity profiles for estrogen receptors α and β. Journal of Medicinal Chemistry 2006 , 49 , 7357-7365. [7] Q. Jiang, F. Payton-Stewart, S. Elliott, J. Driver, L. V. Rhodes, Q. Zhang, S. Zheng, D. Bhatnagar, S. M. Boue, B. M. Collins-Burow, J. Sridhar, C. Stevens, J. A. McLachlan, T. E. Wiese, M. E. Burow, G. Wang, Effects of 7-O substitutions on estrogenic and anti-estrogenic activities of daidzein analogues in MCF- 7 breast cancer cells. Journal of Medicinal Chemistry 2010 , 53 , 6153-6163. [8] Y. Okamoto, A. Suzuki, K. Ueda, C. Ito, M. Itoigawa, H. Furukawa, T. Nishihara, N. Kojima, Anti- estrogenic activity of prenylated isoflavones from Millettia pachycarpa : Implications for pharmacophores and unique mechanisms. Journal of Health Science 2006 , 52 , 186-191. [9] S. M. Boué, T. E. Cleveland, C. Carter-Wientjes, B. Y. Shih, D. Bhatnagar, J. M. McLachlan, M. E. Burow, Phytoalexin-Enriched Functional Foods. Journal of Agricultural and Food Chemistry 2009 , 57 , 2614- 2622. [10] T. Nomura, T. Fukai, Phenolic constituents of licorice ( Glycyrrhiza species), Progress in the chemistry of organic natural products , Vol. 73 , Springer, 1998 . [11] R. Simons, J.-P. Vincken, E. J. Bakx, M. A. Verbruggen, H. Gruppen, A rapid screening method for prenylated flavonoids with ultra-high-performance liquid chromatography/electrospray ionisation mass spectrometry in licorice root extracts. Rapid Communications in Mass Spectrometry 2009 , 23 , 3083-3093. [12] H. Kohno, K. Kouda, R. Tokunaga, Y. Sonoda, Detection of estrogenic activity in herbal teas by in vitro reporter assays. European Food Research and Technology 2007 , 225 , 913-920. [13] D. T. Zava, M. Blen, G. Duwe, Estrogenic activity of natural and synthetic estrogens in human breast cancer cells in culture. Environmental Health Perspectives 1997 , 105 , 637-645.

92

Agonistic and antagonistic estrogens in licorice root

[14] S. Tamir, M. Eizenberg, D. Somjen, S. Izrael, J. Vaya, Estrogen-like activity of glabrene and other constituents isolated from licorice root. Journal of Steroid Biochemistry and Molecular Biology 2001 , 78 , 291-298. [15] S. Tamir, M. Eizenberg, D. Somjen, N. Stern, R. Shelach, A. Kaye, J. Vaya, Estrogenic and antiproliferative properties of glabridin from licorice in human breast cancer cells. Cancer Research 2000 , 60 , 5704-5709. [16] D. Somjen, S. Katzburg, J. Vaya, A. M. Kaye, D. Hendel, G. H. Posner, S. Tamir, Estrogenic activity of glabridin and glabrene from licorice roots on human osteoblasts and prepubertal rat skeletal tissues. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 241-246. [17] D. Somjen, E. Knoll, J. Vaya, N. Stern, S. Tamir, Estrogen-like activity of licorice root constituents: Glabridin and glabrene, in vascular tissues in vitro and in vivo. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 147-155. [18] M. Shanmugam, N. L. Krett, E. T. Maizels, R. E. Cutler Jr, C. A. Peters, L. M. Smith, M. L. O'Brien, O. K. Park-Sarge, S. T. Rosen, M. Hunzicker-Dunn, Regulation of protein kinase C δ by estrogen in the MCF-7 human breast cancer cell line. Molecular and Cellular Endocrinology 1999 , 148 , 109-118. [19] T. D. Rachner, M. Schoppet, U. Niebergall, L. C. Hofbauer, 17β-Estradiol inhibits osteoprotegerin production by the estrogen receptor-α-positive human breast cancer cell line MCF-7. Biochemical and Biophysical Research Communications 2008 , 368 , 736-741. [20] T. F. H. Bovee, R. J. R. Helsdingen, I. M. C. M. Rietjens, J. Keijer, R. L. A. P. Hoogenboom, Rapid yeast estrogen bioassays stably expressing human estrogen receptors α and β, and green fluorescent protein: A comparison of different compounds with both receptor types. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 99-109. [21] M. Nassiri Asl, H. Hosseinzadeh, Review of pharmacological effects of Glycyrrhiza sp. and its bioactive compounds. Phytotherapy Research 2008 , 22 , 709-724. [22] T. Nomura, T. Fukai, T. Akiyama, Chemistry of phenolic compounds of licorice ( Glycyrrhiza species) and their estrogenic and cytotoxic activities. Pure and Applied Chemistry 2002 , 74 , 1199-1206. [23] T. F. H. Bovee, M. Thevis, A. R. M. Hamers, A. A. C. M. Peijnenburg, M. W. F. Nielen, W. G. E. J. Schoonen, SERMs and SARMs: Detection of their activities with yeast based bioassays. Journal of Steroid Biochemistry and Molecular Biology 2010 , 118 , 85-92. [24] M. J. Meyers, S. Jun, K. E. Carlson, B. S. Katzenellenbogen, J. A. Katzenellenbogen, Estrogen receptor subtype-selective ligands: Asymmetric synthesis and biological evaluation of cis- and trans-5,11- dialkyl-5,6,11,12- tetrahydrochrysenes. Journal of Medicinal Chemistry 1999 , 42 , 2456-2468. [25] M. G. Glazier, M. A. Bowman, A review of the evidence for the use of phytoestrogens as a replacement for traditional estrogen replacement therapy. Archives of internal medicine 2001 , 161 , 1161-1172. [26] A. M. Sotoca, T. F. H. Bovee, W. Brand, N. Velikova, S. Boeren, A. J. Murk, J. Vervoort, I. M. C. M. Rietjens, Superinduction of estrogen receptor mediated gene expression in luciferase based reporter gene assays is mediated by a post-transcriptional mechanism. Journal of Steroid Biochemistry and Molecular Biology 2010 , 122 , 204-211. [27] T. F. H. Bovee, W. G. E. J. Schoonen, A. R. M. Hamers, M. J. Bento, A. A. C. M. Peijnenburg, Screening of synthetic and plant-derived compounds for (anti)estrogenic and (anti)androgenic activities. Analytical and Bioanalytical Chemistry 2008 , 390 , 1111-1119. [28] S. M. Boué, M. E. Burow, T. E. Wiese, B. Y. Shih, S. Elliott, C. H. Carter-Wientjes, J. A. McLachlan, D. Bhatnagar, Estrogenic and antiestrogenic activities of phytoalexins from red kidney bean ( Phaseolus vulgaris L.). Journal of Agricultural and Food Chemistry 2011 , 59 , 112-120.

93

Chapter 4

94

Chapter 5

Increasing soya isoflavonoid content and diversity by simultaneous malting and challenging by a fungus to modulate estrogenicity

Based on : Rudy Simons, Jean-Paul Vincken, Nikolaos Roidos, Toine F.H. Bovee, Martijn van Iersel, Marian A. Verbruggen, Harry Gruppen, Increasing soy isoflavonoid content and diversity by simultaneous malting and challenging by a fungus to modulate estrogenicity, Journal of Agricultural and Food Chemistry , 2011 , in press

Chapter 5

ABSTRACT

Soybeans were germinated on kilogram-scale, by application of malting technology used in the brewing industry, and concomitantly challenged with Rhizopus microsporus var. oryzae . In a time-course experiment, samples were taken every 24 h for 10 days and the isoflavonoid profile was analyzed by RP-UHPLC-MS. Upon induction with R. microsporus , the isoflavonoid composition changed drastically with the formation of phytoalexins belonging to the subclasses of the pterocarpans and coumestans, and by prenylation of the various isoflavonoids. The pterocarpan content stabilized at 2.24 mg daidzein equivalents (DE) per g after ~9 days. The levels of the less common glyceofuran, glyceollin IV and VI ranged from 0.18 to 0.35 mg DE/g, and were comparable to those of the more commonly reported glyceollins I, II and III (0.22-0.32 mg DE/g) and glycinol (0.42 mg DE/g). The content of prenylated isoflavones after the induction process was 0.30 mg DE/g. The total isoflavonoid content increased by a factor 10-12 on DW basis after 9 days, which suggested to be ascribable to de novo synthesis. These changes were accompanied by a gradual increase in agonistic activity of the extracts towards both the estrogen receptor α (ERα) and ERβ during the 10-day induction, with a more pronounced activity towards ERβ. Thus, the induction process yielded a completely different spectrum of isoflavonoids, with a much higher bioactivity towards the estrogen receptors. This, together with the over 10- fold increase in potential bioactives, offers promising perspectives for producing more, novel, and higher potency nutraceuticals by malting under stressed conditions.

Keywords : Prenylation, isoflavonoids, soya, estrogenicity, phytoalexins, antagonism

96

Increasing the isoflavonoid content and diversity of soya

INTRODUCTION

The Leguminosae (or Fabaceae ) is an economically important family of flowering plants. It comprises many edible legumes such as soya beans ( Glycine max ), beans ( Phaseolus ssp.), peas ( Pisum sativum ), chickpeas ( Cicer arietinum ), alfalfa ( Medicago sativa ), peanut (Arachis hypogaea ) and licorice ( Glycyrrhiza glabra ). Many edible legumes share the feature of belonging to the subfamily of Papilionoideae. All species within the subfamily Papilionoideae have in common that they accumulate isoflavonoids, a subclass of flavonoids [1]. Soybean is considered to be one of the richest sources of isoflavones, a subclass of isoflavonoids. The three main isoflavones in soya are daidzein, genistein and glycitein. These isoflavones are present as aglycone, glucoside, acetyl-glucoside or malonyl-glucoside [2]. Soya isoflavones are known for their weak estrogenic properties and it is assumed that soya consumption is associated with health-beneficial effects towards hormone-related conditions such as menopausal complaints and osteoporosis [3]. Isoflavones exert their in vitro estrogenicity by binding to the estrogen receptor (ER) and are, therefore, often referred to as phytoestrogens. Application of stress to soya during germination is known to induce changes in the isoflavonoid profile, i.e. the formation of novel isoflavonoids, also known as phytoalexins. Phytoalexins are defined as low molecular-weight, anti-microbial compounds that are synthesized and accumulated in plants in response to e.g. pathogen attack or application of chemical stimuli [4]. Soy phytoalexins comprise pterocarpans, coumestans and isoflavones [5]. The formation of coumestans and pterocarpans is thought to result from the conversion of daidzein into coumestrol and glycinol, respectively [6] . Prenylation is a common feature to these phytoalexins. Coumestrol can be converted into phaseol by enzyme-catalysed substitution with a prenyl chain on the 4-position. Glycinol can be enzymatically substituted with a prenyl chain to form glyceollidins I and II, which can subsequently ring-closed into a pyran ring to form glyceollins I-III by the action of a cyclase [4, 6, 7]. Furthermore, the occurrence of prenylated isoflavones in Rhizopus- challenged soya seedlings has been reported recently [5]. Although the estrogenic activities of several soya phytoalexins have been described, surprisingly neither the formation of phytoalexins nor the changes in estrogenic activity in response to the altered isoflavonoid composition has been quantitatively monitored in time. In this study, soya beans were germinated in the presence of the food- grade fungus Rhizopus microsporus var. oryzae . Subsequently, both their isoflavonoid profile and estrogenic activity towards ER α and ER β were assessed during 10 days with 24 h intervals. The treatment was performed on kilogram-scale with malting technology commonly employed in the brewing industry. The aim of this study is to quantify the changes in isoflavonoid composition and the concomitant changes in the estrogenicity by means of a yeast estrogen bioassay [8].

97

Chapter 5

EXPERIMENTAL

Materials

Soya beans, Glycine max (L.) Merrill, were provided by Frutarom Belgium (Londerzeel,

Belgium). The authentic standards of daidzein, genistein, coumestrol and estradiol (E 2) were purchased from Sigma Aldrich (St. Louis, MO, USA), both R,R-THC and RU 58668 were purchased from Tocris Bioscience (Bristol, United Kingdom). UHPLC/MS grade acetonitrile (ACN) was purchased from Biosolve BV (Valkenswaard, The Netherlands). Water was prepared using a Milli-Q water purification system (Millipore, Billerica, MA, USA). All other chemicals were purchased from Merck (Darmstadt, Germany). The fungus, Rhizopus microsporus var. oryzae (LU 581), was purchased from the Fungal Biodiversity Centre CBS (Utrecht, The Netherlands).

Rhizopus -challenged germination of soya beans

The soya beans were subjected to a controlled malting protocol. This treatment consists of 3 stages: soaking, germination and challenging by Rhizopus , and is further referred to as induction. All stages were performed in the absence of light. Samples were collected every 24 h and directly stored at -20 °C. Soya beans were selected by visual screening. Damaged beans and debris were discarded (2-3% w/w). Before starting the treatment, surface-sterilization was performed by soaking the soya beans in a 1% (w/v) hypochlorite solution (5 L/kg beans) for 1 h at 20 °C. After surface-sterilization the soya beans were rinsed at least 3 times with (tap) water (5 L/kg beans). The treatment was performed in a Joe White micro-malting system unit 90-102 (Perth, Australia) comprising 16 separate compartments. Compartments 1 to 3 and 14 to 16 were not used for the treatment. The micro-malting system was cleaned and surface-sterilized before use, according to a cleaning in place protocol with the application of the commercial cleaning product Chlorosept (Sopura S.A., Courcelles, Belgium). For each time course experiment, a total of 4.0 kg of non-soaked soya beans was divided into 10 portions of 400 g each. The portions were distributed over compartments 4 to 13. In the first stage, the beans were soaked for 24 h at 22 °C (air flow parameters: 30% air (100% = 3000 rpm of ventilator); 85% recirculation, 15% fresh air). After soaking, the soaking water was drained, and the beans entered the second stage in which they were germinated for 48 h at 30 °C under 100% RH (30% air; 85% recirculation). The germination stage was followed by a third stage in which the soya seedlings were inoculated with a spore suspension of R. microsporus . After inoculation, the germination was continued for another 7 days at 30 °C, with adjusted air flow parameters (40% air; 25% recirculation).

98

Increasing the isoflavonoid content and diversity of soya

For the fungal inoculation, a sporangiospore suspension was prepared by scraping off the sporangia from pure slant cultures of R. microsporus , grown on malt extract agar (CM59; Oxoid, Basingstoke, UK) for 7 days at 30 °C. Subsequently, the sporangia were suspended in sterile Milli-Q water with 0.85% (w/v) NaCl (10 8 CFU/mL). The beans were inoculated with the sporangiospore suspension (0.2 mL/g) by gently distributing the suspension manually through the germinated seeds.

Soya seedling extraction

The frozen samples were freeze-dried and subsequently milled in a Retch Ultra centrifugal mill ZM 200 (Haan, Germany) using a 0.5 mm sieve. Ten grams of soya bean powder was defatted with 250 mL hexane under continuous reflux conditions for 4 h. Defatted powder (0.5 g) was subsequently extracted with 50 mL of EtOH by sonication (20 min) and shaking (20 min; 250 rpm) at 20 oC in order to extract the isoflavonoids. The suspension was centrifuged (2500 g, 10 min at 20 °C) and the pellet was rinsed with 10 mL of EtOH. This procedure was repeated four times and the isoflavonoid contents of the supernatants was monitored by UV absorption at 280 nm to ensure that all isoflavonoids were extracted. The first four extracts and rinsing fractions were pooled, as the fifth supernatant did not contain isoflavonoids, i.e. did not show any UV absorption at 280 nm. The ethanol was removed under reduced pressure by a Savant ISS-110 SpeedVac concentrator (Thermo Fisher Scientific, Waltham, USA). The dried extracts were dissolved in EtOH to a concentration of 10 mg/mL and were stored at -20 oC. Prior to analysis, these stock solutions were thawed at room temperature (RT), vortexed and centrifuged (5 min; 14000 rpm, RT).

RP-UHPLC-MS analysis

Samples were analyzed on a Thermo Accela UHPLC system (San Jose, CA, USA) equipped with pump, autosampler and PDA detector. Samples (1 µL) were injected on an Acquity UPLC BEH shield RP18 column (2.1 x 150 mm, 1.7 µm particle size) with an Acquity UPLC shield RP18 Vanguard pre-column (2.1 x 5 mm, 1.7 µm particle size; Waters, Milford, MA, USA). Water acidified with 0.1% (v/v) acetic acid, eluent A, and ACN acidified with 0.1% (v/v) acetic acid, eluent B, were used as eluents. The flow rate was 300 µL/min, the column temperature was controlled at 35 °C, and the PDA detector was set to measure at a range of 200-400 nm. The following elution profile was used: 0-2 min, linear gradient from 10%-25% (v/v) B; 2-9 min, linear gradient from 25%-50% (v/v) B; 9-12 min, isocratic on 50% B; 12-22 min, linear gradient from 50%-100% (v/v) B; 22-24 min, isocratic on 100% B; 24-25 min, linear gradient from 100%-10% (v/v) B; 25-27 min, isocratic on 10% (v/v) B. Mass spectrometric data were obtained by analyzing samples on a LTQ-XL (Thermo

99

Chapter 5

Scientific) equipped with an ESI-MS probe coupled to the RP-UHPLC. Mass spectrometric analysis in both positive (PI) and negative ion (NI) mode (data acquisition and reprocessing) was performed as described previously [5]. Due to the lack of commercially available references, daidzein was used as a generic standard to quantify the amounts of isoflavonoids expressed as mg daidzein equivalents per g dry weight (mg DE/g). The quantification was performed at 280 nm by means of Xcalibur (version 2.0.7, Thermo Scientific).

Determination of estrogenic activity

The protocol for the yeast-based estrogen bioassay measurements was adopted from Bovee et al . [8] with slight modifications. The yeast for this assay is genetically modified. It has a strong constitutive expression vector, stably integrated in the genome, to express either the human estrogen receptor α (ERα) or the human estrogen receptor β (ERβ). The yeast genome also contains a reporter construct, with an inducible yeast-enhanced fluorescent protein (yEGFP) regulated by the activation of a minimal promoter with estrogen responsive elements (EREs). Cultures of the yeast estrogen biosensor with either ERα or ERβ were grown overnight at 30 °C with shaking at 200 rpm. At the late log phase, the cultures of both estrogen receptors were diluted in selective minimal medium (MM) supplemented with either leucine (ERα) or histidine (ERβ) to an OD value (630 nm) between 0.04 and 0.06 (ERα) and 0.1-0.2 (ERβ). For exposure, 200-μL aliquots of this diluted yeast culture were combined with 2 μL of test compound or extract (in various concentrations) in a 96-well plate to test the agonistic properties of these compounds. DMSO (blank) and control samples containing 17 β-estradiol (E2) or genistein dissolved in DMSO were included in each experiment. Dilution series of each sample were prepared in DMSO and the final concentration of DMSO in the assay did not exceed 1% (v/v). Each sample concentration was assayed in triplicate. Exposure was performed for 24 h and 6 h for the ER α and ER β, respectively, in an orbital shaker (200 rpm; 30 °C). Fluorescence and OD were measured at 0 and 24 h for the ERα and 0 and 6-8 h for the ER β in a Tecan Infinite F500 (Männedorf, Switzerland) using an excitation filter of 485 nm (bandwidth 20 nm) and an emission filter of 535 nm (bandwidth 35 nm). The fluorescence signal of the samples was corrected with the signals obtained for the background signal of the culture medium with yeast. In order to verify the viability of the yeast in each well, the absorbance was measured at 630 nm. Each extract of the different time points in the treatment of a representative Rhizopus -challenged germination experiment was measured at 4 different dilutions, and every dilution was measured in triplicates: 1, 3, 10 and 30 µg/mL in both the ERα and ERβ bioassay. EC 50 calculations were performed in Sigma Plot (8.02, SPSS Inc.). The yeast-based assay was validated with a dilution-series of reference compounds, estradiol and genistein. The EC 50 values in the ERα bioassay were 0.86 nM and 1.73 µM for estradiol and genistein, respectively, and 0.12 nM

100

Increasing the isoflavonoid content and diversity of soya

and 9.1 nM in the ERβ bioassay. All EC 50 values were in line with those reported in literature [8]. For the determination of the potential antagonistic properties of the soya extracts towards ERα and β, the yeast cells were exposed to the EC 70 or EC 90 of estradiol, respectively, in combination with the 4 different extract concentrations (1, 3, 10 and 30 µg/mL; each measured in triplicates) of the same representative Rhizopus -challenged germination experiment. For each ER antagonist test, a reference compound was used as a positive control for antagonism, RU 58688 for ERα [9], and R,R-THC for ERβ [10]. All experiments were performed in triplicates.

RESULTS

Growth parameters of soya seedlings

The percentage of germination of the soya beans and the average root length were determined on 20 randomly picked seedlings from each of the 24 h interval-samples. Root length was determined on the main root, excluding possible side-roots. Germinated beans were defined as beans of which the root had a minimal length of 2 mm. The germination was expressed as the percentage of the germinated beans of the total amount of beans sampled. The averages (± standard deviation) of both percentage of germination and root length in the time course experiment were calculated on the basis of 5 individual time course experiments. The different parameters of the soya bean treatment that were monitored are shown in Table 1 . The percentage of germination increased from 11% after 24 h to 90% after 6-7 days. The root development was characterized by a steady growth in the first 5 days after which a plateau was reached of approximately 90 mm. Furthermore, side-root formation was observed after 4 days of germination. These results are in line with similar soya bean germination experiments in which an average germination percentage of 93% after 4 days was reported [11]. The seedling development in time is illustrated in Figure 1 .

A weight loss of ~10% DW compared to the non-soaked beans was observed after 10 days (Table 1 ). To account for this loss, the isoflavonoid content was recalculated to g DW of original soya bean material at t=0. Without this correction, the total amount of isoflavonoid after germination would be overestimated.

101

Chapter 5

Table 1. Growth parameters of the soya beans/soya seedlings during the induction process, represented by the average of 5 experiments. Data are mean ± SD values of measurements performed in five-fold. Time Stage in % of Root length Absolute DW (g) ¶ Relative DW (d) treatment germination (mm) † (g/100g) ¶ 0 Non -soaked * 0 0 366.3 (±0.2) 100 1 Soaking 11 (±14) 5.3 (±0.3) 341.2 (±20.1) 93.2 (±5.5) 2 53 (±9) 30.4 (±2.3) 343.9 (±29.2) 93.9 (±8.0) Germination 3 63 (±11) 41.9 (±3.4) 322.0 (±4.6) 87.9 (±1.2) 4 75 (±8) 54.1 (±13.8) 317.8 (±43.1) 86.8 (±11.8) 5 83 (±5) 67.2 (±6.9) 314.8 (±36.8) 86.0 (±10.1) 6 86 (±4) 82.0 (±8.5) 328.5 (±3.1) 89.7 (±0.9) 7 Challenging 88 (±3) 77.5 (±9.0) 333.3 (±8.3) 91.0 (±2.3) 8 90 (±0) 78.2 (±1.4) 336.2 (±7.4) 91.8 (±2.0) 9 90 (±0) 90.5 (±1.8) 340.6 (±14.8) 93.0 (±4.0) 10 90 (±0 ) 87.7 (±7.7) 330.7 (±4.7) 90.3 (±1.3) * The weight of the experimental batch size of fresh beans was 400 g before drying. † The root length was defined as the length of the main root excluding possible side-roots. ¶ Absolute and relative DW refers to the whole seedling weight.

Identification of compounds

The UHPLC analysis of the extracts of non-soaked beans, seedlings and Rhizopus - challenged seedlings revealed that the UV-profiles changed drastically during induction (Figure 2 ). The identities of most peaks were established previously [5], whereas peaks 1, 8, 10 , 13 -15 , 24 and 28 were now identified for the first time ( Table 2 ). Figure 3 summarizes most of the structures found in this study. Furthermore, the metabolite flow from constitutive (isoflavones) to induced isoflavonoids (pterocarpans and coumestans) is visualized. Prenylation is indicated as the most down-stream event in the biosynthetic pathway, and relates to all three isoflavonoid subclasses mentioned. Compound 1 was tentatively assigned as phenylalanine based on MS analysis. In addition, three other pterocarpans were tentatively assigned at retention times (Rt) of 8.08 ( 10 ), 13.85 ( 24 ) and 14.67 ( 28 ) min with MW of 354, 336 and 354, respectively. The UV spectra of compounds 10 , 24 and 28 were characterized by two λ max values of 227(±2) nm and 288(±3) nm. Glycinol, glyceollidins I/II and glyceollins I, II and III showed the same two λ max values. In addition, the NI MS 2 spectra of compounds 10 , 24 and 28 displayed a x,y characteristic fragment combination (-H2O RDA fragments) that is also observed in the MS 2 spectra of glycinol, glyceollidins I/II and glyceollins I-III, but not in any other MS 2 spectrum in the chromatogram [5]. Moreover, in PI mode, in-source fragmentation of + compounds 10 , 24 and 28 occurred, leading to the predominance of the [M+H-H2O] fragment ion in MS 1. These observations strongly suggest that 10 , 24 and 28 belong to the subclass of the 6a-hydroxypterocarpans. Prenylation of 24 and 28 was confirmed by the characteristic neutral loss of 56 in PI mode MS 2 [12], whereas a loss of a methyl radical was

102

Increasing the isoflavonoid content and diversity of soya

Figure 1 . Soya seedling development in time (t in days) illustrated by representative samples. The induction process starts with non-soaked beans (t=0) that were soaked (t=1), germinated (t=2,3) and challenged by Rhizopus (t=4-10).

observed in NI mode MS 2 of compound 28. Based on all these findings, compounds 24 and 28 were tentatively assigned and glyceollin V/VI and glyceollin IV, 2 respectively. The absence of the characteristic neutral loss of 56 in PI mode MS for compound 10 could be explained by the presence of a hydroxyl-group on the furan ring (hydroxyl-isopropenyl furan), as in glyceofuran. This hydroxyl-group altered its fragmentation: instead of a loss of 42 [C 3H6] or 56 [C 4H8], a loss of 28 [CO] was observed. Therefore, compound 10 was tentatively assigned as glyceofuran.

Five other compounds were observed. Compound 14 was tentatively assigned as the flavanone naringenin based on comparison of the NI mode MS 2 spectrum [13]. Compound 8 was assigned as biochanin A on the basis of comparative MS spectral analysis [14-16]. Two other unknown flavonoid peaks were observed at retention times 8.85 ( 13 ) and

10.33 ( 15 ). In the UV spectra of both 13 and 15 , two λ max values were observed around 224 and 355 nm, resembling the UV spectrum of coumestrol that has λ max values at 221 and 343-360 nm (data not shown). Moreover, coumestrol, 13 and 15 have a predominant loss of 28 (CO) in PI mode MS 2 in common, and all three showed no retro-Diels-Alder (RDA) fragments in both PI and NI mode MS 2. Therefore, compound 13 was tentatively assigned as C-methylated derivative of coumestrol, methyl-coumestrol. The predominant loss of a methyl radical observed in NI mode MS 2 of compound 15 led to its tentative assignment as

103

Chapter 5

a methyl ether of hydroxylated coumestrol. The absence of RDA fragments in MS 2 does not allow further speculation on the position of the C-methyl (13 ) or methoxyl (15 ) group. The only methoxylated coumestrol isolated from soya is isotrifoliol, 3,9-dihydroxy-1- methoxy-coumestan [17]. Compound 15 was, therefore, tentatively assigned as isotrifoliol.

Changes in the isoflavonoid composition during the induction process

Isoflavones The quantification of all isoflavonoids annotated at each time point is given in Table 3 . The changes in the isoflavone profile upon soaking and germination were mainly characterized by an increase of the aglycone, glucoside and malonyl-glucoside forms of daidzein, genistein and glycitein. This increase was most pronounced after soaking (t=1). The total isoflavone content increased from 0.37 to 1.54 mg DE/g during induction. The ratio of total (free and conjugated forms of) genistein, daidzein and glycitein of approximately 0.45:0.45:0.10 remained constant throughout the whole process.

Pterocarpans The pterocarpan pool consisted of glyceollins I-IV ( 19 -21 ,28 ) and VI/V* ( 24 ), glyceollidin I/II ( 17 ), glyceofuran ( 10 ) and glycinol ( 4). The pterocarpans represent the predominant group of phytoalexins that was induced in the soya seedlings and increased throughout the treatment, reaching a maximum level of 2.2 mg DE/g. The non-prenylated glycinol was the only pterocarpan that steadily accumulated throughout the whole time course experiment. The levels of all other pterocarpans appeared to stabilize at t=9.

Coumestans The coumestans in the seedlings comprised coumestrol (18 ), C-methyl-coumestrol ( 13 ), isotrifoliol (15 ) and 4-prenyl-coumestrol ( 31 ). The levels of coumestrol showed the same trend as those of glyceollidin I/II, but they did not exceed the 0.16 mg DE/g. The total level of coumestans reached a maximum of 0.41 mg DE/g at t=9. Similar to most pterocarpans, the level of coumestans appeared to stabilize at t=9.

Estrogenic activity of extracts from induced soya beans

The estrogenic activity in response to the altered isoflavonoid profile in the extracts of soya during induction is shown in Figure 4 . With 1 and 3 µg/mL extract, ERα showed too low responses, whereas a concentration of 30 µg/mL gave too high responses for most extracts (comparable with the maximum response obtained with a high dose of 17β- estradiol). With 10 µg/mL, a clear increase in estrogenic activity in the ERα bioassay in time was observed. For ERβ, the three highest concentrations of extracts resulted in.

104

Increasing the isoflavonoid content and diversity of soya

Figure 2 . RP-UHPLC-UV-profile of EtOH extracts of non-soaked soya beans (A), soya bean seedlings at t=1 (B), Rhizopus -challenged soya bean seedlings at t=4 (C) and at t=10 (D). Peak numbers refer to compounds in Table 2 .

105

Table 2. Compounds tentatively assigned by UHPLC-ESI-MS in the extracts of Glycine max after induction. - 2 + 2 No Rt. UV max (nm) Identification Class Molecular [M -H] MS NI product ions [M+H] MS PI product ions (min) formula (relative abundance ) (relative abundance )

1 1.63 224,290 Phenylalanine Amino acid C9H11 NO 2 164 147( 100 ) 166 120( 100 ) † 2 4.46 225,248 Daidzin Isoflavone C21 H20 O9 415 253( 100 ) 417 255( 100 ) † 3 4.56 225,256 Glycitin Isoflavone C22 H22 O10 445 283( 100 ) 447 283( 100 ) † * 4 5.56 205,226,281 Glycinol Pterocarpan C15 H12 O5 271 256( 34 ),227( 28 ),161( 100 ) 255 199( 100 ),237( 52 ) † 5 5.57 225,260 Genistin Isoflavone C21 H20 O10 431 431( 100 ),269( 11 ) 433 271( 100 ) † 6 5.81 225,255 Malonyl-daidzin Isoflavone C24 H22 O12 501 253( 100 ) 503 255( 100 ) † 7 7.13 209,225,260 Malonyl-genistin Isoflavone C24 H22 O13 517 269( 100 ) 519 271( 100 )

8 7.74 224,257 Biochanin A Isoflavone C16 H12 O5 283 268( 100 ) 285 270( 100 ) † 9 7.95 226,257 Glycitein Isoflavone C16 H12 O5 283 268( 100 ) 285 270( 100 ),229( 21 ),225( 16 ) * 10 8.03 225,257,291 Glyceofuran Pterocarpan C20 H18 O6 353 335(100),161( 6),149( 25 ) 337 319( 87 ),309( 100 ),188( 25 ) † 11 8.14 226,297 Daidzein Isoflavone C15 H10 O4 253 253( 100 ) 255 227( 49 ),199( 100 ),137( 70 ) † 12 8.26 212,226,257 2'-OH-genistein Isoflavone C15 H10 O6 285 217( 100 ) 287 245( 49 ),217( 100 ),153( 57 )

13 8.80 225,286,360 C-methyl-coumestrol Coumestan C16 H10 O5 281 281( 11 ),253( 100 ) 283 255( 100 )

14 9.62 225,287 Naringenin Flavanone C15 H12 O5 271 177( 23 ),151( 100 ) 273 243(56),215(73),253(100)

15 10.30 225,351 Isotrifoliol Coumestan C16 H10 O6 297 297( 25 ),282( 100 ) 299 284(19),271( 100 ),267(15) † 16 10.52 224,260 Genistein Isoflavone C15 H10 O5 269 269( 100 ) 271 153( 100 ),215( 79 ) † * 17 10.93 209,227,285 Glyceollidin I/II Pterocarpan C20 H20 O5 339 324( 45 ),161( 100 ) 323 267( 100 ) † 18 11.15 226,303,343 Coumestrol Coumestan C15 H8O5 267 267( 100 ) 269 241( 100 ),225( 22 ),197( 23 ) † * 19 11.41 211,227,284 Glyceollin III Pterocarpan C20 H18 O5 337 319( 100 ),293( 8),149( 18 ) 321 306( 76 ),297( 79 ),251( 100 ) † * 20 11.72 211,228,283,307 Glyceollin II Pterocarpan C20 H18 O5 337 319( 100 ),293( 24 ),149( 44 ) 321 306( 74 ),297( 100 ), 251( 94 ) † * 21 11.85 210,230,282 Glyceollin I Pterocarpan C20 H18 O5 337 319( 100 ),293( 34 ),149( 95 ) 321 306( 71 ),303( 100 ), 293( 33 ) † 22 12.92 225,292 Aprenyl -2’OH-daidzein Isoflavone C20 H18 O5 337 337( 37 ),293( 100 ),282( 84 ) 339 283( 100 ) † 23 13.80 225,257 Bprenyl -glycitein Isoflavone C21 H20 O5 351 336( 100 ) 353 297( 100 ),285( 11 ) * 24 13.89 225,287,308 Glyceollin V/VI Pterocarpan C20 H16 O5 335 317( 100 ),149( 39 ) 319 319( 28 ),291( 63 ),263( 100 ) † 25 14.35 225,251,305 Aprenyl -daidzein Isoflavone C20 H18 O4 321 321( 16 ),266( 100 ) 323 267( 100 ) † 26 14.44 229,262 Aprenyl -2’OH-genistein Isoflavone C20 H18 O6 353 285( 95 ),284( 100 ),267( 37 ) 355 299( 100 ) † 27 14.62 225,257,301 Bprenyl -daidzein Isoflavone C20 H18 O4 321 321( 100 ),266( 16 ),265( 68 ) 323 267( 100 ),255( 8) * 28 14.71 226,285 Glyceollin IV Pterocarpan C21 H22 O5 353 335( 100 ),149( 20 ),148( 5) 337 281( 100 ),269( 49 ) † 29 17.12 229,265 Aprenyl -genistein Isoflavone C20 H18 O5 337 337( 34 ),283( 12 ),282( 100 ) 339 283( 100 ) † 30 17.30 229,260,334 Bprenyl -genistein Isoflavone C20 H18 O5 337 337( 100 ),282( 16 ),281( 51 ) 339 283( 100 ),271( 11 ) † 31 18.03 229,307,343 4-prenyl-coumestrol Coumestan C20 H16 O5 335 335( 48 ),281( 16 ),280( 100 ) 337 281( 100 ) n.a. – not available; NI mode - negative ion mode; PI mode - positive ion mode. * + 1 + The [M+H-H2O] dominated in the MS compared to the [M+H] . † previously assigned by Simons et al. [5].

Figure 3 . Biosynthetic pathway and structures of isoflavonoids in soya beans and Rhizopus -induced soya seedlings. * In the literature glyceollin V is often referred to as glyceollin IIIb. This scheme was compiled from Veitch (2007, 2009), Banks and Dewick (1983), Simons et al. (2011) [1, 5, 6, 39] .

Chapter 5

maximum responses in time, and no trend was observed. At 1 µg/mL of extract, saturation in signal (i.e. maximal response) was not reached yet and an increase in estrogenicity for the ERβ in time was observed. The most representative experiments with respect to the increasing estrogenicity in time are shown in Figure 4 . The estrogenicity towards the ERβ was relatively higher compared to that towards the ERα, even more so as the ERβ bar diagram was obtained with 10-fold lower extract concentrations than that of ERα. For 1 µg/mL the estrogenicity towards the ERβ appeared to reach a plateau value at t=5 The extracts were also tested for possible antagonistic activity towards both estrogen receptors. These experiments showed that the extracts were unable to suppress the agonistic response caused by the dose of 17β-estradiol that gave a 70 to 90% maximal response. On the contrary, the responses only increased by these co-exposures, showing additive agonistic effects, e.g. the addition of the extracts of t=4 to t=10 enhanced the estrogenicity of estradiol by 20-25% (data not shown).

Figure 4. Estrogenic activity of the ethanol extracts of the soya bean seedlings at different time points measured in the yeast estrogen bioassays. The extracts were measured at 10 µg/mL in the ERα bioassay, and at 1 µg/mL in the ERβ bioassay. Data are mean ± SD values of measurements performed in triplicates. E 2 is 1 µM 17β-estradiol.

108

Increasing the isoflavonoid content, diversity and estrogenicity of soya

DISCUSSION

By combining malting technology and fermentation with the food-grade R. microsporus , the induction of phytoalexins in soya seedlings was successfully performed in terms of increasing total isoflavonoid content, altering isoflavonoid composition and enhancing the estrogenic potential of the soya material.

De novo synthesis of isoflavonoids during the induction-process

The isoflavonoid content of soya increased by a factor 10-12 suggesting de novo synthesis of isoflavonoids. This is further corroborated by the appearance (t=1) of phenylalanine ( 1), the basic precursor of flavonoid biosynthesis, and naringenin ( 14 ), a precursor of genistein (16 ). The formation of the flavanone naringenin might thus explain the increase of the genistein pool [18]. These findings are supported by several reports in which de novo synthesis is suggested based on quantification of isoflavonoid content, up-regulation of enzymes involved in isoflavonoid biosynthesis and incorporation of 14 C-labeled phenylalanine, daidzein, 2’OH-daidzein and glycinol into glyceollins I-III [6, 19-22]. Data on accurate quantification of the total isoflavonoid content during the induction process are very limited. It has been reported that levels ranging from 0.01 to 1 mg/g fresh weight (FW) of glyceollins I-III can accumulate in fungus-challenged soya seedlings [23]. With an approximate water content of 65-70% (w/w) [24], the glyceollin I-III levels in soya seedlings range from 0.03 to 3 mg/g DW, whereas a content of 0.9 mg DE/g DW was found at 9 days in our study ( Table 3 ). In most studies, soya beans are wounded prior to inoculation. It has been shown that this additional treatment can increase the glyceollin I-III levels of soya beans inoculated with Aspergillus sojae by a factor 10 [25]. This extra wounding step might explain the difference between our data and those in the literature. Moreover, a number of other pterocarpans (1.3 mg DE/g DW) were found in our study ( Table 3 ), suggesting that the difference between our results and those in the literature might actually be smaller than they appear. The stabilization of the glyceollins I-III contents at t=9 might be due to the inhibition of cyclase, the enzyme responsible for the last step in glyceollin biosynthesis, as this enzyme is inhibited by its reaction products [26]. In line with this, the accumulation of glycinol did not seem to level off.

Changes in isoflavonoid profile Although inoculation was performed at t=3, small amounts of pterocarpans and coumestans already occurred at t=2. This might be due to incomplete sterilization. Alternatively, the sterilization agent itself (hypochlorite) might have acted as a chemical elicitor.

109

Chapter 5

Table 3. Quantitative analysis of the isoflavonoid profile during soya bean induction. The content of each flavonoid is expressed in mg daidzein equivalent (DE) per g DW, corrected for loss in dry weight during the induction process. Data are mean ± SD values of five independent Rhizopus -challenged germination experiments. No Isoflavonoid Prenyl t=0 t=1 t=2 t=3 5 Genistin - 0.11(±0.00) 0.30(±0.07) 0.25(±0.04) 0.18(±0.03) 7 Malonyl -genistin - 0.06(±0.03) 0.33(±0.07) 0.29(±0.04) 0.29(±0.07) 12 Genistein - 0.01(±0.00) 0.05(±0.01) 0.05(±0.02) 0.05(±0.02) 16 2' -Hydroxygenistein - - 0.01(±0.00) 0.01(±0.00) 0.02(±0.00)

26 Aprenyl -2’OH -genistein Chain - - - 0.01(±0.01) 29 Aprenyl -genistein Chain - - 0.01(±0.00) 0.02(±0.01) 30 Bprenyl -genistein Chain - - - 0.01(±0.01) Genistein pool 0.18 (±0.04) 0.70 (±0.16) 0.60 (±0.10) 0.58 (±0.14) 3 Glycitin - 0.02 (±0.01) 0.10 (±0.02) 0.09 (±0.01) 0.07 (±0.02) 9 Glyci tein - 0.01 (±0.01) 0.06 (±0.01) 0.05 (±0.01) 0.06 (±0.02)

23 Bprenyl -glycitein Chain - - - <0.01 Glycitein pool 0.03 (±0.02) 0.16 (±0.03) 0.14 (±0.02) 0.13 (±0.04) 2 Daidzin - 0.07(±0.03) 0.25(±0.04) 0.22(±0.04) 0.18(±0.03) 6 Malonyl -daidzin - 0.07(±0.04) 0.39(±0.07) 0.34(±0.04) 0.32(±0.08) 11 Daidzein - 0.02(±0.01) 0.05(±0.01) 0.05(±0.02) 0.07(±0.02)

22 Aprenyl -2’OH -daidzein Chain - - - 0.02(±0.01) 25 Aprenyl -daidzein Chain - - <0.01 0.02(±0.01) 27 Bprenyl -daidzein Chain - - <0.01 0.01(±0.00) Daidzein pool 0.16 (±0.07) 0.69 (±0.13) 0.63 (±0.10) 0.62 (±0.15) 8 Biochanin A - - - - - Isoflavone pool 0.37 (±0.13) 1.54 (±0.32) 1.38 (±0.22) 1.32 (±0.33) 4 Glycinol - - <0.01 <0.01 0.01(±0.00) 17 Glyceollidin I/II Chain - <0.01 <0.01 0.02(±0.01) 19 Glyceollin III Furan 1 - <0.01 0.01(±0.00) 0.02(±0.01) 20 Glyceollin II Pyran - - 0.00(±0.00) 0.01(±0.01) 21 Glyceollin I Pyran - <0.01 0.02(±0.00) 0.07(±0.04) 28 Glyceollin IV Chain - - <0.01 <0.01 24 Glyceollin V/ VI Furan 2 - <0.01 <0.01 <0.01 10 Glyceofuran Fur an 3 - 0.07(±0.00) 0.06(±0.00) 0.06(±0.00) Pterocarpan pool - 0.07 (±0.00) 0.09 (±0.01) 0.20 (±0.09) 18 Coumestrol - - - 0.01(±0.00) 0.02(±0.01) 13 C-methyl -coumestrol - - - - - 15 Isotrifoliol - - - - 0.01(±0.00) 31 4-prenyl -coumestrol Chain - - - 0.01 (±0.01) Coumestan pool - - - 0.03 (±0.02) Total isoflavonoids 0.37(±0.13) 1.61(±0.32) 1.48(±0.23) 1.54(±0.44)

Chain, 3,3 dimethylallyl (prenyl chain). Pyran, 2,2-dimethylchromeno. Furan 1, 2’’-isopropenyl-dihydrofuran. Furan 2, 2’’-isopropenyl-furan. Furan 3, 2’’-(2-hydroxy-)isopropenyl-dihydrofuran.

110

Increasing the isoflavonoid content, diversity and estrogenicity of soya

t=4 t=5 t=6 t=7 t=8 t=9 t=10 0.23(±0.04) 0.23(±0.03) 0.20(±0.01) 0.23(±0.03) 0.19(±0.04) 0.26(±0.06) 0.19(±0.02) 0.30(±0.07) 0.25(±0.06) 0.25(±0.04) 0.26(±0.02) 0.22(±0.05) 0.31(±0.13) 0.20(±0.04) 0.05(±0.03) 0.05(±0.04) 0.06(±0.04) 0.06(±0.03) 0.09(±0.02) 0.09(±0.04) 0.06(±0.01) 0.02(±0.01) 0.02(±0.01) 0.02(±0.01) 0.02(±0.01) 0.02(±0.00) 0.08(±0.08) 0.02(±0.00) 0.02(±0.01) 0.03(±0.02) 0.04(±0.02) 0.04(±0.01) 0.04(±0.02) 0.04(±0.02) 0.03(±0.01) 0.03(±0.01) 0.03(±0.01) 0.04(±0.02) 0.05(±0.02) 0.05(±0.02) 0.06(±0.02) 0.03(±0.02) 0.02(±0.01) 0.03(±0.01) 0.04(±0.02) 0.04(±0.01) 0.04(±0.02) 0.05(±0.02) 0.03(±0.01) 0.66 (±0.18) 0.64 (±0.18) 0.65 (±0.17) 0.70 (±0.14) 0.65 (±0.17) 0.89 (±0.38) 0.56 (±0.11) 0.07 (±0.01) 0.07 (±0.01) 0.06 (±0.01) 0.07 (±0.01) 0.05 (±0.01) 0.06 (±0.01) 0.05 (±0.01) 0.08 (±0.01) 0.06 (±0.01) 0.05 (±0.01) 0.06 (±0.01) 0.06 (±0.01) 0.07 (±0.02) 0.05 (±0.02) 0.01 (±0.00) 0.01 (±0.00) 0.01 (±0.00) 0.01 (±0.00) 0.01 (±0.00) 0.01 (±0.00) 0.01 (±0.00) 0. 15 (±0.03) 0.13 (±0.02) 0.12 (±0.03) 0.13 (±0.02) 0.12 (±0.02) 0.15 (±0.04) 0.10 (±0.03) 0.19(±0.03) 0.19(±0.01) 0.16(±0.02) 0.18(±0.03) 0.14(±0.04) 0.18(±0.03) 0.12(±0.01) 0.32(±0.06) 0.31(±0.05) 0.27(±0.05) 0.27(±0.05) 0.22(±0.06) 0.28(±0.10) 0.21(±0.05) 0.0 9(±0.04) 0.09(±0.03) 0.11(±0.05) 0.13(±0.04) 0.13(±0.02) 0.16(±0.05) 0.10(±0.01) 0.04(±0.02) 0.05(±0.02) 0.07(±0.04) 0.07(±0.02) 0.08(±0.04) 0.09(±0.04) 0.06(±0.03) 0.03(±0.01) 0.03(±0.01) 0.04(±0.02) 0.03(±0.01) 0.03(±0.02) 0.04(±0.02) 0.03(±0.01) 0.01 (±0.00) 0.01(±0.00) 0.01(±0.01) 0.01(±0.00) 0.01(±0.00) 0.01(±0.00) 0.01(±0.00) 0.68 (±0.17) 0.68 (±0.14) 0.66 (±0.18) 0.69 (±0.16) 0.62 (±0.18) 0.76 (±0.24) 0.53 (±0.12) 0.01(±0.01) 0.01(±0.01) 0.02(±0.01) 0.02(±0.01) 0.02(±0.01) 0.02(±0.01) 0.01(±0.01) 1.50 (±0.39) 1.46 (±0.34) 1.46 (±0.38) 1.54 (±0.32) 1.40 (±0.39) 1.83 (±0.67) 1.21 (±0.27) 0.04(±0.01) 0.10(±0.03) 0.17(±0.04) 0.25(±0.03) 0.27(±0.06) 0.35(±0.03) 0.42(±0.07) 0.06(±0.04) 0.12(±0.06) 0.16(±0.11) 0.18(±0.07) 0.20(±0.08) 0.22(±0.05) 0.23(±0.09) 0.07(± 0.04) 0.12(±0.07) 0.18(±0.11) 0.21(±0.09) 0.29(±0.09) 0.32(±0.07) 0.29(±0.06) 0.04(±0.02) 0.08(±0.03) 0.12(±0.07) 0.17(±0.06) 0.21(±0.04) 0.24(±0.03) 0.22(±0.04) 0.17(±0.09) 0.22(±0.10) 0.24(±0.13) 0.29(±0.11) 0.31(±0.09) 0.33(±0.05) 0.32(±0.06) 0.02(±0 .03) 0.08(±0.08) 0.16(±0.15) 0.25(±0.13) 0.27(±0.16) 0.33(±0.20) 0.35(±0.21) 0.03(±0.02) 0.07(±0.04) 0.13(±0.06) 0.18(±0.04) 0.22(±0.08) 0.24(±0.05) 0.22(±0.05) 0.04(±0.01) 0.06(±0.02) 0.08(±0.04) 0.10(±0.04) 0.15(±0.05) 0.17(±0.03) 0.18(±0.04) 0.47 (±0. 26) 0.86 (±0.43) 1.24 (±0.70) 1.63 (±0.57) 1.91 (±0.64) 2.21 (±0.51) 2.24 (±0.62) 0.06(±0.02) 0.08(±0.03) 0.10(±0.06) 0.12(±0.04) 0.15(±0.06) 0.16(±0.05) 0.13(±0.04) 0.01(±0.01) 0.03(±0.01) 0.06(±0.03) 0.09(±0.04) 0.15(±0.03) 0.16(±0.04) 0.15(±0.03) 0.01(±0.0 1) 0.03(±0.02) 0.05(±0.03) 0.06(±0.03) 0.09(±0.05) 0.10(±0.04) 0.08(±0.02) 0.01(±0.01) 0.03(±0.02) 0.05(±0.04) 0.05(±0.03) 0.07(±0.06) 0.09(±0.06) 0.06(±0.05) 0.09 (±0.05) 0.18 (±0.09) 0.26 (±0.17) 0.33 (±0.13) 0.45 (±0.20) 0.51 (±0.20) 0.42 (±0.14) 2.06(±0.70 ) 2.49(±0.85) 2.95(±1.24) 3.50(±1.03) 3.77(±1.24) 4.54(±1.38) 3.87(±1.03)

111

Chapter 5

The ratio of the commonly found glyceollins I, II and III was 1.4:1:1.3 at t=9, respectively, and differed from the ratios of 6:1:2 previously reported [27]. More strikingly, the present study shows the additional formation of glyceofuran ( 10 ), glyceollin IV ( 28 ), glyceollin VI ( 24 ), coumestrol-derivatives ( 13 ,15 ,31 ), and several prenylated isoflavones (22 ,23 ,25 -27 ,29 ,30 ). The simultaneous occurrence of these compounds has never been reported in Glycine max . However, glyceollin IV has been found in G. tomentella and glyceollin VI in G. clandestine and G. tabacina [28, 29]. The isomer of glyceollin VI has recently been isolated and given the trivial name glyceollin V [17]. However, the name glyceollin V was already used for the enantiomeric form of glyceollin III [28]. To avoid confusion, the two enantiomers of glyceollins III are indicated as glyceollin IIIa and IIIb in Figure 3 , whereas the isomer of glyceollin VI is indicated by glyceollin V. The occurrence of phaseol ( 31 ) and isotrifoliol (15 ) has been reported before in G. max [30, 31]. The formation of prenylated isoflavones has recently been reported for G. max , as well as in several other, often stress-induced, Leguminosae species [5]. After the 9 days of induction, these compounds have accumulated to an impressive 32 % (w/w) of the total isoflavonoids in the soya material. Taken together, the more extensive changes in isoflavonoid composition as observed in the present study compared to studies by others is most likely due to the combination of prolonged induction times and more favorable growth conditions in the micro-malting system. Consequently, the altered ratio between the most common glyceollins and the formation of compounds might represent the outcome of events occurring more downstream in the biosynthesis process. Possibly, omission of the wounding pretreatment also plays a role.

Accumulation of phytoestrogens during the induction process

Figure 4 shows a gradual increase in estrogenicity of the extracts during the induction process towards both the ERα and ERβ. It is likely that the increased proportion of isoflavonoids and the change in isoflavonoid composition in the extract are responsible for this increase of estrogenic activity. Deglycosylation of isoflavones by β-glucosidase might be a driver behind this [32] . The proportion of glucosides (compounds 2, 3 and 5-7) decreased from 90% (t=0) to approximately 20% (t=9). It is known that the glucosyl residues hamper binding to the ER [8]. More importantly, other isoflavonoids were formed of which several have been reported as potent agonists towards both ERs, particularly glycinol and coumestrol [8, 33]. Differences in the outcome of various bioassays described in the literature, and lack of data for binding of individual isoflavonoids for both ERs, do not allow extrapolations concerning the contribution of each isoflavonoid to the estrogenicity of the current extract. Besides agonistic activity, it is likely that the extracts contain compounds with antagonistic properties. For soya bean, this has been demonstrated for glyceollin I [34], which comprised 8% (w/w) of the total isoflavonoids at t=9. To determine to what extent

112

Increasing the isoflavonoid content, diversity and estrogenicity of soya

phytoestrogens with a known EC 50 value can account for the estrogenic activity of a tested extract, the expected estrogenicity of the extract was calculated from the contribution of each phytoestrogen (expressed in E 2 equivalents, EEQ), using their relative estrogenic potency (REP) and their quantity in the extract. This was performed for extracts of t=10 and t=4 with ER α and ER β, respectively (see Appendix IX ). The calculated values were compared to those actually measured. The expected estrogenicity (expressed as the sum of the EEQ of genistein, coumestrol and daidzein) was approximately 25% higher than the EEQ corresponding to the response measured in the assay. The same calculation was performed for the ER β showing that the total expected estrogenicity was even a factor 7 higher than the EEQ of the response measured. Thus, the measured response in the assay was lower than the one calculated, which hints at the presence of antagonists in the extract. Even more so, as the calculated EEQ based on genistein, coumestrol and daidzein

(phytoestrogens with a known EC 50 ) is most likely underestimated because glycinol, a potent phytoestrogen on both ER subtypes with an estrogenic activity similar to that of coumestrol, has not been taken into account. The data on glycinol available from the literature were obtained with a bioassay different from ours [33]. As mentioned earlier, glyceollin I has been reported to display antagonistic activity on the ER α. It has been suggested that prenyl substituents are important structural attributes for inducing antagonistic responses [35-38]. In Figure 5 , the accumulation of prenylated isoflavonoids is compared with that of non-prenylated isoflavonoids. It is clearly seen that longer induction times promote a higher proportion of prenylated isoflavonoids, reaching similar amounts as non-prenylated isoflavonoids at around t=7. This suggests that the accumulation of prenylated isoflavonoids in time increases the antagonistic activity of the extracts.

Figure 5 . Quantitative analysis of the isoflavonoid profile in the time course experiments of the development of Rhizopus -elicited soya seedlings.

113

Chapter 5

In summary, we speculate that the increased agonistic ER activity of the extracts is mainly due to the formation of the non-prenylated pterocarpan, glycinol, and the non- prenylated coumestans, particularly coumestrol. Besides, the extracts might also be rich in ER antagonists, as prenylation of isoflavonoids is known to correlate well with antagonistic properties (e.g. glyceollin I). Although antagonism was not established directly in our study, we found lower estrogenic response with the extracts than might be expected from their amounts of agonists. Further investigation is necessary to substantiate this.

CONCLUSIONS

Malting of soya beans in the presence of a food-grade fungus yielded a completely different spectrum of isoflavonoids, with a much higher bioactivity towards both estrogen receptors subtypes than the starting material (non-soaked beans). Together with the over 10-fold increase in potential bioactives, this offers promising perspectives for producing more, novel, and higher potency nutraceuticals by malting soya beans under stressed conditions.

ACKNOWLEDGEMENTS

This work was financially supported by the Food & Nutrition Delta of the Ministry of Economic Affairs, The Netherlands. We would like to thank Loes Mol (Wageningen University, Wageningen, the Netherlands) for her contribution to the estrogenicity experiments, RIKILT-Institute of Food Safety (WUR) for making the yeast estrogen bioassays available, and Hennie van de Laar (Holland Malt, Lieshout, the Netherlands) for his assistance in setting up the induction experiments.

REFERENCES

[1] N. C. Veitch, Isoflavonoids of the Leguminosae. Natural Product Reports 2007 , 24 , 417-464. [2] H. J. Wang, P. A. Murphy, Isoflavone content in commercial soybean foods. Journal of Agricultural and Food Chemistry 1994 , 42 , 1666-1673. [3] M. J. Messina, Legumes and soybeans: Overview of their nutritional profiles and health effects. American Journal of Clinical Nutrition 1999 , 70 , 439S-450S. [4] F. D. Dakora, D. A. Phillips, Diverse functions of isoflavonoids in legumes transcend anti-microbial definitions of phytoalexins. Physiological and Molecular Plant Pathology 1996 , 49 , 1-20. [5] R. Simons, J.-P. Vincken, M. C. Bohin, T. F. M. Kuijpers, M. A. Verbruggen, H. Gruppen, Identification of prenylated pterocarpans and other isoflavonoids in Rhizopus spp. elicited soya bean seedlings by electrospray ionisation mass spectrometry. Rapid Communications in Mass Spectrometry 2011 , 25 , 55-65. [6] S. W. Banks, P. M. Dewick, Biosynthesis of glyceollins I, II and III in soybean. Phytochemistry 1983 , 22 , 2729-2733. [7] J. L. Ingham, N. T. Keen, L. J. Mulheirn, R. L. Lyne, Inducibly-formed isoflavonoids from leaves of soybean. Phytochemistry 1981 , 20 , 795-798.

114

Increasing the isoflavonoid content, diversity and estrogenicity of soya

[8] T. F. H. Bovee, R. J. R. Helsdingen, I. M. C. M. Rietjens, J. Keijer, R. L. A. P. Hoogenboom, Rapid yeast estrogen bioassays stably expressing human estrogen receptors a and b, and green fluorescent protein: A comparison of different compounds with both receptor types. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 99-109. [9] T. F. H. Bovee, M. Thevis, A. R. M. Hamers, A. A. C. M. Peijnenburg, M. W. F. Nielen, W. G. E. J. Schoonen, SERMs and SARMs: Detection of their activities with yeast based bioassays. Journal of Steroid Biochemistry and Molecular Biology 2010 , 118 , 85-92. [10] M. J. Meyers, S. Jun, K. E. Carlson, B. S. Katzenellenbogen, J. A. Katzenellenbogen, Estrogen receptor subtype-selective ligands: Asymmetric synthesis and biological evaluation of cis- and trans-5,11- dialkyl-5,6,11,12- tetrahydrochrysenes. Journal of Medicinal Chemistry 1999 , 42 , 2456-2468. [11] U. K. Choi, M. H. Kim, N. H. Lee, Y. S. Jeong, O. J. Kwon, Y. C. Kim, Y. H. Hwang, The characteristics of cheonggukjang , a fermented soybean product, by the degree of germination of raw soybeans. Food Science and Biotechnology 2007 , 16 , 734-739. [12] R. Simons, J.-P. Vincken, E. J. Bakx, M. A. Verbruggen, H. Gruppen, A rapid screening method for prenylated flavonoids with ultra-high-performance liquid chromatography/electrospray ionisation mass spectrometry in licorice root extracts. Rapid Communications in Mass Spectrometry 2009 , 23 , 3083-3093. [13] N. Fabre, I. Rustan, E. de Hoffmann, J. Quetin-Leclercq, Determination of flavone, flavonol, and flavanone aglycones by negative ion liquid chromatography electrospray ion trap mass spectrometry. Journal of the American Society for Mass Spectrometry 2001 , 12 , 707-715. [14] J. K. Prasain, C. C. Wang, S. Barnes, Mass spectrometric methods for the determination of flavonoids in biological samples. Free Radical Biology and Medicine 2004 , 37 , 1324-1350. [15] J. Kang, L. A. Hick, W. E. Price, A fragmentation study of isoflavones in negative electrospray ionization by MSn ion trap mass spectrometry and triple quadrupole mass spectrometry. Rapid Communications in Mass Spectrometry 2007 , 21 , 857-868. [16] B. Klejdus, J. Vacek,L. Benešová, J. Kopecký, O. Lapčík,V. Kubáň, Rapid-resolution HPLC with spectrometric detection for the determination and identification of isoflavones in soy preparations and plant extracts. Anal. Bioanal. Chem. 2007, 389 , 2277-2285. [17] H. J. Yuk, J. H. Lee, M. J. Curtis-Long, J. W. Lee, Y. S. Kim, H. W. Ryu, C. G. Park, T. S. Jeong, K. H. Park, The most abundant polyphenol of soy leaves, coumestrol, displays potent α-glucosidase inhibitory activity. Food Chemistry 2011 , 126 , 1057-1063. [18] O. Yu, B. McGonigle, Metabolic engineering of isoflavone biosynthesis. Advances in Agronomy, Volume 86 2005 , 86 , 147-190. [19] V. V. Lozovaya, A. V. Lygin, O. V. Zernova, S. Li, G. L. Hartman, J. M. Widholm, Isoflavonoid accumulation in soybean hairy roots upon treatment with Fusarium solani . Plant Physiology and Biochemistry 2004 , 42 , 671-679. [20] T. L. Graham, J. E. Kim, M. Y. Graham, Role of constitutive isoflavone conjugates in the accumulation of glyceollin in soybean infected with Phytophthora megasperma Molecular Plant-Microbe Interactions 1990 , 3, 157-166. [21] S. Dhaubhadel, B. D. McGarvey, R. Williams, M. Gijzen, Isoflavonoid biosynthesis and accumulation in developing soybean seeds. Plant Molecular Biology 2003 , 53 , 733-743. [22] V. V. Lozovaya, A. V. Lygin, S. Li, G. L. Hartman, J. M. Widholm, Biochemical response of soybean roots to Fusarium solani f. sp. glycines infection. Crop Science 2004 , 44 , 819-826. [23] M. K. Bhattacharyya, E. W. B. Ward, Resistence, susceptibility and accumulation of glyceollins I-III in soybean organs inoculated with Phytophthora megasperma f. sp. glycinea . Physiological and Molecular Plant Pathology 1986 , 29 , 227-237. [24] H. Shi, P. K. Nam, Y. Ma, Comprehensive profiling of isoflavones, phytosterols, tocopherols, minerals, crude protein, lipid, and sugar during soybean ( Glycine max ) germination. Journal of Agricultural and Food Chemistry 2010 , 58 , 4970-4976. [25] H. J. Kim, H. J. Suh, J. H. Kim, S. C. Kang, S. Park, C. H. Lee, J. S. Kim, Estrogenic activity of glyceollins isolated from soybean elicited with Aspergillus sojae . Journal of Medicinal Food 2010 , 13 , 382-390.

115

Chapter 5

[26] R. Welle, H. Grisebach, Induction of phytoalexin synthesis in soybean: Enzymatic cyclization of prenylated pterocarpans to glyceollin isomers. Archives of Biochemistry and Biophysics 1988 , 263 , 191-198. [27] S. M. Boué, C. H. Carter, K. C. Ehrlich, T. E. Cleveland, Induction of the soybean phytoalexins coumestrol and glyceollin by Aspergillus . Journal of Agricultural and Food Chemistry 2000 , 48 , 2167- 2172. [28] N. T. Keen, R. L. Lyne, T. Hymowitz, Phytoalexin Production as a chemosystematic parameter within the Genus Glycine . Biochemical Systematics and Ecology 1986 , 14 , 481-486. [29] R. L. Lyne, L. J. Mulheirn, Minor pterocarpinoids of soybean. Tetrahedron Letters 1978 , 19 , 3127-3128. [30] H. J. Yuk, J. H. Lee, M. J. Curtis-Long, J. W. Lee, Y. S. Kim, H. W. Ryu, C. G. Park, T.-S. Jeong, K. H. Park, The most abundant polyphenol of soy leaves, coumestrol, displays potent [alpha]-glucosidase inhibitory activity. Food Chemistry 2011 , 126 , 1057-1063. [31] P. Caballero, C. M. Smith, F. R. Fronczek, N. H. Fischer, Isoflavones from an insect-resistant variety of soybean and the molecular structure of afrormosin. Journal of Natural Products 1986 , 49 , 1126-1129. [32] Suzuki, H.; Takahashi, S.; Watanabe, R.; Fukushima, Y.; Fujita, N.; Noguchi, A.; Yokoyama, R.; Nishitani, K.; Nishino, T.; Nakayama, T., An isoflavone conjugate-hydrolyzing β-glucosidase from the roots of soybean ( Glycine max ) seedlings: Purification, gene cloning, phylogenetics, and cellular localization. J. Biol. Chem. 2006, 281, (40), 30251-30259. [33] S. M. Boué, S. L. Tilghman, S. Elliott, M. C. Zimmerman, K. Y. Williams, F. Payton-Stewart, A. P. Miraflor, M. H. Howell, B. Y. Shih, C. H. Carter-Wientjes, C. Segar, B. S. Beckman, T. E. Wiese, T. E. Cleveland, J. A. McLachlan, M. E. Burow, Identification of the potent phytoestrogen glycinol in elicited soybean ( Glycine max ). Endocrinology 2009 , 150 , 2446-2453. [34] M. C. Zimmermann, S. L. Tilghman, S. M. Boué, V. A. Salvo, S. Elliott, K. Y. Williams, E. V. Skripnikova, H. Ashe, F. Payton-Stewart, L. Vanhoy-Rhodes, J. P. Fonseca, C. Corbitt, B. M. Collins-Burow, M. H. Howell, M. Lacey, B. Y. Shih, C. Carter-Wientjes, T. E. Cleveland, J. A. McLachlan, T. E. Wiese, B. S. Beckman, M. E. Burow, Glyceollin I, a novel antiestrogenic phytoalexin isolated from activated soy. Journal of Pharmacology and Experimental Therapeutics 2010 , 332 , 35-45. [35] Q. Jiang, F. Payton-Stewart, S. Elliott, J. Driver, L. V. Rhodes, Q. Zhang, S. Zheng, D. Bhatnagar, S. M. Boue, B. M. Collins-Burow, J. Sridhar, C. Stevens, J. A. McLachlan, T. E. Wiese, M. E. Burow, G. Wang, Effects of 7-O substitutions on estrogenic and anti-estrogenic activities of daidzein analogues in MCF- 7 breast cancer cells. Journal of Medicinal Chemistry 2010 , 53 , 6153-6163. [36] E. M. Ahn, N. Nakamura, T. Akao, T. Nishihara, M. Hattori, Estrogenic and antiestrogenic activities of the roots of Moghania philippinensis and their constituents. Biological and Pharmaceutical Bulletin 2004 , 27 , 548-553. [37] Y. Okamoto, A. Suzuki, K. Ueda, C. Ito, M. Itoigawa, H. Furukawa, T. Nishihara, N. Kojima, Anti- estrogenic activity of prenylated isoflavones from Millettia pachycarpa : Implications for pharmacophores and unique mechanisms. Journal of Health Science 2006 , 52 , 186-191. [38] S. M. Boué, M. E. Burow, T. E. Wiese, B. Y. Shih, S. Elliott, C. H. Carter-Wientjes, J. A. McLachlan, D. Bhatnagar, Estrogenic and antiestrogenic activities of phytoalexins from red kidney bean ( Phaseolus vulgaris L.). Journal of Agricultural and Food Chemistry 2011 , 59 , 112-120. [39] N.C. Veitch, Isoflavonoids of the leguminosae. Natural Product Reports 2009, 26 , 776-802.

116

Chapter 6

General Discussion

Chapter 6

In this thesis, the prenylated isoflavonoid profile of soya and licorice, and their relation to estrogenicity, is described. First, a liquid chromatography / mass spectrometry-based screening method was developed that enabled the identification of prenylated flavonoids (including isoflavonoids), and that also discriminated between different kinds of prenyl substitution (chain and pyran ring). This method was used to characterise the prenylated isoflavonoid profile of Rhizopus -challenged soya seedlings and licorice roots. Secondly, the estrogenic activity of soya was enhanced by induction of isoflavonoids by a combination of germination and biotic stress (challenging with Rhizopus ). Thirdly, we have tried to correlate the kind of estrogenic activity (agonism, antagonism) in extracts from licorice root and Rhizopus -challenged soya seedlings to prenylation of the constituent isoflavonoids.

MASS SPECTROMETRY AS A TOOL TO DETECT PRENYLATION

Update of guideline for detecting prenylation in flavonoids

Based on the literature, we expected the presence of a broad range of different isoflavonoids in the sources studied in this thesis. Furthermore, we also anticipated encountering a range of different types of prenyl substituents. Hence, isolation of all individual components would not be feasible. Mass spectrometry (MS), coupled to ultra- high performance liquid chromatography (UHPLC), provided a powerful analytical technique allowing rapid analysis of complex extracts. With the possibility to perform MS n fragmentations, MS is particularly suitable for the tentative identification of flavonoids. As described in Chapter 2 , fragmentation of prenylated flavonoids is characterised by a predominant degradation of the prenyl group, evidenced by neutral losses of 56 alone, or 42 and 56. It is known that the MS 2 fragmentation of e.g. non- prenylated isoflavones, such as genistein, often shows a neutral loss of 56, related to the sequential loss of two CO moieties [1]. Therefore, we introduced a mass threshold to exclude that non-prenylated flavonoid aglycones were assigned as being prenylated. The mass threshold would filter out the isoflavone aglycones, but not glycosylated flavonoids, which might result in false positives ( Chapter 2 , Figure 6 ). It is known that several species outside the Leguminosae family such as the genus Epimedium within the family of Berberidaceae contain glycosyl-conjugates of prenylated flavones [2]. However, to our knowledge no data was available on the MS n fragmentation of the glycosyl-conjugates of prenylated flavones in PI mode. A recent study on the analysis of two glycosyl-conjugates of prenylated isoflavones from the wild Mexican lupine (Lupinus reflexus ) showed that both luteone 4’,7-O-diglucoside and 2,3- didehydrokievitone 4’,7-O-diglucoside gave a predominant loss of one and two glucosyl moieties in the MS 2 spectra, respectively, in combination with a neutral loss of 56 with a

118

General discussion

relative abundance of ~50% [3]. In contrast, the MS 2 spectra of non-prenylated isoflavonoid glycosides were dominated by the neutral loss of the glycosyl substituent, whereas the neutral loss of 56 was exclusively observed in MS 3 spectra. Thus, prenylated isoflavone glycosides seem to give a neutral loss of 56 in MS 2 originating from the prenyl group

(C 4H8), whereas non-prenylated isoflavone glycosides exclusively show a neutral loss of in 56 in MS 3, which is related to the sequential loss of two CO moieties [1]. The necessity to screen for prenyl group-related neutral losses in MS 3 is still required, as e.g. prenylated isoflavans from licorice root ( Chapter 2 ) would otherwise not be detected. In Figure 1 , an updated version, incorporating the latest data [3], of the interpretation guideline as proposed in Chapter 2 is shown. With this updated interpretation guideline it is now possible to screen for prenylated flavonoids and discriminate between prenyl chain and pyran ring, no matter whether they are glycosylated or not.

Figure 1 . Updated interpretation guideline based on the screening for prenylated flavonoids from Chapter 2 . In this guideline, the combination of a neutral loss of 56 or 42 and 56 in MS 3 in combination with a neutral loss related to a loss of glycosyl residues in MS 2 indicates the presence of a non-prenylated flavonoid glycoside.

119

Chapter 6

Detection of prenyl substituents other than chains and pyran rings

In Chapters 3 and 4, the tentative assignment of three pterocarpans, glyceollin III, glyceollin VI and glyceofuran was described. These pterocarpans were substituted with the less common prenyl groups, 2’’-isopropenyl-dihydrofurano, 2’’-isopropenyl-furano, and 2’’-(2-hydroxy-) isopropenyl-dihydrofurano, respectively ( Figure 2 ). The degradation in

PI mode of glyceollin III and VI was characterised by predominant losses of 42 (C 3H6) and 2 2 56 (C 4H8) in MS . Whereas the MS spectrum of glyceollin III showed a predominant loss of 42 (80%) over 56 (30%), degradation of glyceollin VI was characterised by a major loss of 56 (100%) and a minor loss of 42 (10%). Apparently, the double bond of the furan group played an important role in the degradation. It appeared that the 2’’-isopropenyl- dihydrofuran-group (glyceollin III) shows pyran ring-like behaviour in fragmentation, whereas 2’’-isopropenyl-furano-group (glyceollin VI) shows behaviour similar to the prenyl chain. It should be noted that the minor neutral losses of 42 (2%) and 56 (3%) in the MS 2 spectrum of glyceofuran (having a hydroxylated prenyl group) were observed in combination with the dominant loss of 28 (CO), which also may originate from the 2’’-(2- hydroxy-)isopropenyl-dihydrofurano-group. The observation of the near absence of fragment ions [M+H-42] + or [M+H-56] + in the degradation of the hydroxylated prenyl group is corroborated by a similar observation for the hydroxylated prenyl group of dehydrocycloxanthohumol hydrate ( Tabel 1 ) [4]. It seems as if hydroxylation of the prenyl- group alters the degradation pathway, as the characteristic losses of 42 and 56 were observed only in trace amounts. Other examples of prenyl fragmentation were not encountered in our research.

Figure 2 . Three pterocarpans from Rhizopus -induced soya seedlings: glyceollin III, glyceollin VI and glyceofuran.

120

General discussion

Mass spectrometric analysis of prenylated flavonoids: PI or NI mode?

As mentioned in Chapter 2 , analysis of flavonoids in NI mode is preferred over PI mode, mainly due its higher sensitivity. However, it has been suggested that fragmentation of prenyl groups is hindered in NI mode due to the nucleophilic nature of the prenyl group [4].

As a result, the neutral loss characteristic of prenyl degradation, C 4H8, has not been observed in NI mode [5]. Our data on the fragmentation of prenylated flavonoids in NI mode suggests that the prenyl group is often degraded in both MS 2 and MS 3, albeit less prominently than in PI mode ( Table 1 ). In our PI mode MS 2 spectra, a loss of 68 was observed for three prenyl chain- substituted isoflavones and one prenyl chain-substituted pterocarpan ( Table 1 ). This loss has also been found by others, and was assigned to the neutral fragment C 5H8, representing the whole prenyl chain [6]. In that same study, the loss of 68 was not observed in double prenylated isoflavones and flavanones. Therefore, the neutral loss of 68 does not appear to be diagnostic for prenylation, suggesting that this loss is an inappropriate screening criterion for prenylation. Several characteristic neutral losses originating from the degradation of prenyl- • groups, including 15 (CH 3 ), 42 (C 3H6), 55 (C 4H7), 68(C 5H8) and 69 (C 5H9) have been observed in NI mode. The observation of these neutral losses was corroborated by a number of studies ( Table 2 ). Although the loss of a methyl radical in MS 2 was often observed, this loss generally suggests the presence of a methoxyl group [7], and is, therefore, not suitable as a screening criterion for prenylation. Furthermore, other neutral losses that could be related to the degradation of the prenyl group, such as 42 and 55, were observed in only few cases. Despite the observation of several neutral losses in NI mode that could be related to the degradation of the prenyl group, the occurrence of these neutral losses was limited and inconsistent. The analysis of prenylated flavonoids is preferred in PI mode, as it results in the consistent and abundant observation of diagnostic neutral losses 42 and 56. In addition, analysis in PI mode allows the discrimination between prenyl chain and pyran ring. As the same neutral losses were observed in the MS n spectra of less common prenyl- substituents, it remains unclear whether discrimination of these prenyl-substituents is possible.

121

Chapter 6

Table 1 .Overview of neutral losses of prenylated flavonoids observed in PI mode MS 2 and MS 3. MS 2 (Relative abundance) Compound Prenyl group Ring 15 42 55 56 68 Ref. Literature Xanthohumol Chain A 100 [4] 4,2’,4’,6’-Tetrahydroxy-3’- [4] Chain A 100 prenylchalcone 5’-Prenylxanthohumol Chain A 61 [4] Dehydrocycloxanthohumol Pyran A [4] Dehydrocycloxanthohumol [4] Pyran 2 A hydrate Isoxanthohumol Chain A 17 1 [4] 6-Prenylnaringenin Chain A 41 [4] 8-Prenylnaringenin Chain A 21 [4] Euchrestaflavanone B Chain + chain 100 [8] Euchrestaflavanone C Chain + pyran A,B ~25 [8] Osajaxanthone Pyran -3 ~25 [8] Macluraxanthone Chain + pyran -3 ~15 [8] This thesis Glabrone Pyran B 100 6 Ch. 2 Glabridin Pyran A 100 4 14 Ch. 2 Glabrol Chain+chain A,B 100 Ch. 2 3’-OH-4’-OMe-Glabridin Pyran A 100 4 14 Ch. 2 4’-OMe-Glabridin Pyran A 100 4 14 Ch. 2 Hispaglabridin A Pyran + chain A,B 100 Ch. 2 Hispaglabridin B Pyran + pyran A,B 100 4 14 Ch. 2 Glyceollidin I/II Chain A 2 8 100 Ch. 3 Glyceollin I Pyran A 80 28 10 24 Ch. 3 Glyceollin II Pyran A 68 100 15 34 Ch. 3 Glyceollin III Furan 5 A 80 80 9 29 Ch. 3 Glyceollin IV Chain A 6 100 52 Ch. 3 Glyceollin VI Furan 6 A 12 11 14 100 Ch. 3 Glyceofuran Furan 7 A 2 3 Ch. 3 Ap-2’-OH-Daidzein Chain A 2 100 Ch. 3 Bp-Glycitein Chain B 1 100 5 Ch. 3 Ap-Daidzein Chain A 100 Ch. 3 Ap-2’-OH-Genistein Chain A 100 Ch. 3 Bp-Daidzein Chain B 100 10 Ch. 3 Ap-Genistein Chain A 100 Ch. 3 Bp-Genistein Chain B 100 10 Ch. 3 Prenyl-coumestrol 8 Chain A 100 Ch. 3 1 More abundantly observed in MS 3. 2 2,2-dimethyl-3-hydroxy-pyran. 3 This polyphenolic compound is not a flavonoid, but belongs to the xanthones or benzophenones [9]. 4 Observed in MS 3. 5 2’’-isopropenyl-dihydrofuran. 6 2’’-isopropenyl-furan. 7 2’’-(2-hydroxy) isopropenyl-dihydrofuran. 8 Tentatively assigned as 4-prenyl-coumestrol ( Chapter 3 ).

122

General discussion

Table 2 .Overview of neutral losses of prenylated flavonoids observed in NI mode MS 2. MS 2 (Relative abundance) Compound Prenyl group Ring 15 42 55 56 68 Misc. Ref. NI mode – This thesis Glabrene Pyran B 100 12 Ch. 2 Glabrone Pyran B 17 Ch. 2 Glabridin Pyran A 12 4 Ch. 2 Glabrol Chain + chain A,B Ch. 2 3’-OH-4’-OMe-Glabridin Pyran A 38 Ch. 2 4’-OMe-Glabridin Pyran A 52 Ch. 2 Hispaglabridin A Pyran + chain A,B 23 1 100 1 69 Ch. 2 Hispaglabridin B Pyran + pyran A,B 100 1 Ch. 2 Glyceollidin I/II Chain A 46 69 Ch. 3 Glyceollin I Pyran A 15 1 9 Ch. 3 Glyceollin II Pyran A 1 100 1 85 1 9 Ch. 3 Glyceollin III Furan 2 A 1 100 1 4 Ch. 3 Glyceollin IV Chain A 5 Ch. 3 Glyceollin VI Furan 3 A 5 4 Ch. 3 Glyceofuran Furan 4 A Ch. 3 Ap-2’-OH-Daidzein Chain A 84 8 Ch. 3 Bp-Glycitein Chain B 100 80 1 Ch. 3 Ap-Daidzein Chain A 100 Ch. 3 Ap-2’-OH-Genistein Chain A 29 90 69 Ch. 3 Bp-Daidzein Chain B 100 Ch. 3 Ap-Genistein Chain A 100 Ch. 3 Bp-Genistein Chain B 100 Ch. 3 Prenyl-coumestrol 5 Chain A 100 Ch. 3 NI mode - Literature Luteone 6 Chain B
[10] Breviflavone A Chain 7 B,B 70 8 [11] Breviflavone B Chain 9 B,B 86 10 [11] M3 11 Chain 12 A 70 13 [12] Hedysarimcoumestan G Chain B

43 13 [13] Hedysarimcoumestan D Chain B

43 13 [13] Hedysarimcoumestan H Chain B

43 13 [13] Hedysarimpterocarpene B Chain B

43 13 [13] Gancaonin A Chain A

43 13 [13] Gancaonin M Chain A

43 13 [13] 5,7-Dihydroxy-4’-O-methyl- Chain A

43 13 [13] 6,8-diprenylisoflavone Licochalcone A Chain B
[14] Desmethylicaritin Chain A
[15] Misc – miscellaneous neutral loss observed in MS 2;
Neutral loss was observed, but relative abundance was not reported. 1 More abundantly observed in MS 3. 2 2’’-isopropenyl-dihydrofuran. 3 2’’-isopropenyl-furan. 4 2’’-(2-hydroxy-)isopropenyl-dihydrofuran. 5 Tentatively assigned as 4-prenyl-coumestrol ( Chapter 3 ). 6 6-prenyl-2’OH-genistein. 7 2-Hydroxy-3-metyl-but-3-enyl. 8 C4H6O was proposed as the corresponding fragment. 9 2’’-(2-hydroxy-)isopropenyl-1’’-hydroxy-dihydrofurano. 10 C4H6O2 was proposed as the corresponding fragment. 11 Human liver microsomal metabolite of xanthohumol with hydroxylation of the prenyl chain. 12 2-Hydroxy-3-metyl-but-3-enyl. 13 C4H7 was proposed as the corresponding fragment.

123

Chapter 6

MS fragmentation of pterocarpans and coumestans

As mentioned in Chapter 1 , very little is known about the fragmentation pathways of pterocarpans and coumestans. Three other studies, investigating the NI fragmentation of pterocarpans, did not report any C-ring or D-ring cleavages [13, 16, 17]. In Chapter 3 , we proposed a fragmentation pathway for pterocarpans in NI mode based on glycinol, glyceollidins and glyceollins ( Figure 3 ). The additional D-ring appeared to affect the fragmentation pathway and an adapted nomenclature was proposed for the fragmentation pathway of this subclass. To our knowledge, only one study previously reported the C-ring cleavage of a pterocarpan in PI mode [18]. In that study, fragment ions 1,3 A+ and 2,3 B+ were assigned after fragmentation in PI mode of 3-hydroxy-9,10-dimethoxy-pterocarpan via an isoflavene- intermediate. According to our proposed nomenclature for pterocarpan fragmentation (Chapter 3 ), their fragment ions 1,3 A+ and 2,3 B+ would be annotated as 1,3,5 A+ and 2,3,5 B+, respectively ( Figure 3 ). The fragmentation of glyceollins in PI mode was not investigated, because the parent ion was not available. Ionisation of glyceollins in PI mode is known to induce an in- + source water loss, resulting in the predominance of [M+H-H2O] (Chapter 3 ). This would result in a pterocarpene structure indicated in Figure 3 , which cannot convert into an isoflavene. The presence of the 6a-hydroxyl group may, therefore, result in a different fragmentation pattern. Apart from the 2 fragment ions previously proposed [18], no generic fragmentation model has been suggested for pterocarpans in PI mode. Whereas the occurrence of 6a-hydroxy-pterocarpans is mostly confined to Glycine spp., pterocarpans are more widespread throughout the Leguminosae . This is illustrated by their occurrence in the economically important legumes such as the phaseollins that are found in Phaseolus ssp., Cicer ssp., Medicago ssp. and Pisum ssp. This warrants that further investigations of the fragmentation of pterocarpans, in both PI and NI mode. From our MS n spectra of coumestrol in NI and PI mode, no A-ring or B-ring fragment ions resulting from C-ring or D-ring cleavage were observed. The fragmentation of coumestrol is mainly characterised by neutral losses of 28 (CO), 44 (CO 2) and 72 (CO- CO 2). Our observations were similar to those of a study in which the same neutral losses were observed for several different coumestans [13]. Coumestans are highly conjugated systems which makes them very stable. In MS fragmentation, this stability appears to lead to a limited number of neutral losses and no apparent A-ring or B-ring fragment ions due to C-ring or D-ring cleavage.

124

General discussion

Figure 3 . Fragmentation pathway of the parent ion and parent ion after water loss of 6a-hydroxy-pterocarpans in NI mode, based on Figure 3 from Chapter 3 .

LICORICE SPENT AS ALTERNATIVE SOURCE FOR PRENYLATED ISOFLAVONOIDS

As mentioned in Chapter 1 , licorice is commonly used in the tobacco, confectionary and pharmaceutical industries mainly as a sweetening agent [19, 20]. In order to extract the major sweet component glycyrrhizin from licorice root, the industrial processing of licorice generally involves the extraction of ground roots with water. After filtration, the water extract is concentrated into syrup. The syrup primarily contains the triterpenoid glycosides, including glycyrrhizin. The by-product after water extraction is often referred to as spent. The spent stripped from its polar compounds still contained relatively large amounts of non-polar constituents, like prenylated flavonoids (Figure 4 ). We expect relatively high amounts of glabridin in spent compared to root and, therefore, we considered spent as an interesting source for developing a phytoestrogen-rich health ingredient. These compounds might be recovered with solvents such as ethanol, methanol or ethyl-acetate (EA). Roots and spent were extracted on lab-scale with EA and the extracts were analysed on RP-UHPLC to profile and quantify their flavonoid content (Figure 4 ). The glabridin contents for root and spent were 3.81 and 1.75 mg/g, respectively. This might suggest that glabridin and glycyrrhizin are co- extracted with water during industrial processing, in accordance with Tian et al . [21]. Of the solvents tested in that study, the extraction of glycyrrhizin was optimal when performed with water, whereas ethanol appeared to be the best solvent for glabridin extraction. However, water extraction also resulted in an extraction of ~20% of the total amount of glabridin from roots. In our study, we found that spent contained only ~55% of the amount of glabridin compared to that of

125

Chapter 6

roots, suggesting that glabridin losses with water extraction might be larger than suggested by Tian et al . [21]. Furthermore, it was observed that glabrene did not occur in the UV-profile of the spent extract but in that of the syrup. This is consistent with the postulation that glabrene is more polar than of glabridin. Thus, water extraction of glycyrrhizin during industrial processing seems to co-extract glabrene and, to a lesser extent, glabridin. This observation is of importance, as glabridin and glabrene are the major estrogenic licorice root constituents. Successful exploitation of spent as a source of phytoestrogen-rich health ingredients might depend on minimising losses in glabridin and glabrene upon extraction of glycyrrhizin. Depending on the market price of glycyrrhizin and a phytoestrogen-rich extract, process innovations can be considered to avoid these losses. Several options might be explored: (1) a reversal of the extraction order in which the roots are first extracted with a non-polar solvent such as ethyl acetate to retrieve all phytoestrogens, or (2) modifying the pH to enable a more specific recovery of glycyrrhizin. With respect to (1), roots were extracted on lab-scale with 100% ethanol and 100% EA. Quantitative analysis showed that both 100% ethanol and 100% EA extracted almost all glabridin from the root. Furthermore, based on MS-analysis hardly any glycyrrhizin was present in the EA extract, whereas glycyrrhizin was found to be predominantly present in both 70% and 100% EtOH extracts. Since EA is a less polar solvent than 100% ethanol, it might provide a good alternative to more selectively extract prenylated flavonoids and not glycyrrhizin.

Figure 4 . UV-profiles of the RP-UHPLC analyses of EA extracts of licorice root (A) and spent (B).

126

General discussion

DEVELOPMENT OF NOVEL PHYTOESTROGEN-ENRICHED SOYA EXTRACTS

Pre-treatment required for optimal induction of phytoalexins?

In glyceollin research, the response of soya to the pathogenic fungus Phytophthora ssp. mostly studied, and generally high levels of glyceollins are reported. More recently other fungi, in particular food-grade fungi, were proven to be effective elicitors [22-25]. Different food-grade microorganisms such as Aspergillus spp., white rice yeast and Rhizopus oligosporus were compared regarding their ability to induce glyceollin formation [25]. All microorganisms were able to induce glyceollin accumulation, with Rhizopus oligosporus being the most effective. The maximum levels of 0.9 mg (daidzein eq.)/g DW of glyceollins I-III reached in our experiments were determined to be within the range of 0.03 to 3 mg/g DW reported in literature but not as high as we expected (Chapter 5 ) [26]. Interestingly, in addition to glyceollins I-III, a broad range of other phytoalexins accumulated in the Rhizopus - challenged soya seedlings in our experiments, such as glycinol, glyceollin IV, glyceollin VI/V* and glyceofuran, in similar levels as glyceollins I-III. This led to a total pterocarpan level of up to 2.2 mg (daidzein eq.)/g DW, suggesting that the difference between our results and those from the literature might be smaller than they appear. Furthermore, other novel phytoalexins were identified, mainly prenylated isoflavones and prenylated coumestans. These phytoalexins were reported for the first time ( Chapter 3 ) in fungi- challenged soya seedlings, leading to a total phytoalexin content of 2.9 mg (daidzein eq.)/g DW. In many studies, the exposure of soya to an elicitor did not last longer than 2 to 3 days and often the soya beans, its cotyledons or hypocotyls, were not germinated prior to elicitation. In our study, soya seedlings were pre-treated for 3 days by soaking and germination, after which the seedlings were exposed to Rhizopus spp. for a 7-day period. Our data showed that the glyceollin content reached its maximum after 6 days of exposure at day 9 of the induction process. In another study, the glyceollin content of soya cotyledons that were not pre-germinated was monitored over a 6 day period [23]. In that study, it was concluded that the maximum glyceollin content was reached after 3 days, after which the levels stabilised or slightly decreased. A major difference between our induction experiments and those reported in literature is the application of wounding to the soya tissue. It has been suggested that wounding plays an important role in the induction of glyceollins [22]. The absence of wounding prior to inoculation might, therefore, have resulted in underproduction of glyceollin I-III, when our results are compared to the highest values reported in the literature. Besides the content of glyceollins, wounding might also speed up the elicitation process, consequently reducing the time necessary to reach maximum levels of phytoalexins.

127

Chapter 6

Growth conditions and elicitors

An important difference between our induction experiments at lab-scale ( Chapter 3 ) and those performed in kilogram-scale in the micro-malting system ( Chapter 5 ) is the difference in growth conditions. Whereas the soya beans on lab-scale are grown in temperature-controlled climate chambers, the kilogram-scale germination of the soya bean was performed with malting technology used in the brewing industry. Apart from a temperature-controlled environment, malting of soya in a micro-malting system provides air circulation, controlled in-flow of fresh air and controlled humidity. Compared to our own lab-scale experiments, the soya beans grown in micro-malting system showed a faster and more even growth and less spoilage. The optimised growth conditions provided by the micro-malting system might, therefore, also affect the induction of glyceollins. The ratio of glyceollins described in Chapter 5 was roughly 6:2:1 in the first days of induction. Towards the end of the induction experiment (t=9) a ratio of approx. 1:1:1 was observed. This ratio has not been reported before for seedlings. In lab-scale experiments, as described in Chapter 3, soya beans were pre-germinated for 4 days followed by a 4-day Rhizopus -challenged germination period. The isoflavonoid profile at the end of the 8-day induction process was characterised by a ratio of glyceollin I, II and III of 6:2:1. Furthermore, only small amounts of glyceollin IV and VI/V*, prenylated isoflavones and prenylated coumestans were observed in the lab-scale experiment. Reproduction of the lab-scale induction-process in the micro-malting system with the same soya bean batch resulted in an isoflavonoid profile similar to the lab-scale induction process regarding diversity, but with different ratios of glyceollin I, II and III ( Figure 5 ). In addition, the amounts of glyceofuran, glyceollin IV and VI/V* increased relative to glyceollin I-III, in the micro-malting system. The same observation was made for the groups of prenylated isoflavones and coumestans. Possibly, the improved germination conditions provided by the micro-malting system have made the elicitation process more efficient, making pre-germination prior to elicitation obsolete. Furthermore, it would be worthwhile to investigate the combination of wounding with micro-malting to see whether this combination results in higher levels of phytoalexins, in particular glyceollins. This combination might accelerate the induction process.

128

General discussion

Figure 5 . RP-UHPLC-UV profile of isoflavonoids extracted from Rhizopus -challenged soya seedlings grown on lab- scale (A) and performed in the micro-malting system (B). Ap, A-ring prenylated; Bp, B-ring prenylated; Mal, malonyl glucoside.

PRENYLATION OF ISOFLAVONOIDS IN RELATION TO ER-ANTAGONISM

The induction of phytoalexins in soya seedlings ( Chapter 5 ) increased the total isoflavonoid content. Concomitant to this increase in isoflavonoids, an increase in estrogenic activity on both ER-subtypes was observed. However, this increase in estrogenic activity was lower than what might be expected based on the calculation of the activity in estradiol equivalents. The difference between expected and measured estrogenicity was observed for both ER-subtypes and was more pronounced for the ERβ. This might suggest the presence of antagonists in the soya extracts that bind both ERα and ERβ. It has been suggested that the beneficial effects of phytoestrogens might also be related to their antagonistic effect on the ERβ [27]. It has been hypothesised that the ERβ plays an important role in the prevention of hyper-proliferation and carcinogenesis in the breast, prostate and gastrointestinal tract [28]. The difference in tissue distribution led to

129

Chapter 6

the hypothesis that there are potential differences in biological function and tissue- selective actions of the two receptors. In literature, studies mainly focus on ERα- antagonism. The discovery of new ERα-antagonists forms the basis of the development of alternative drugs in hormone replacement therapy and cancer prevention [29]. Therefore, the relationship between prenylation of isoflavonoids and ERα-antagonism was further addressed. As mentioned before, the discrepancy in estrogenicity measured and calculated might be explained by the presence of antagonists. For soya, this has been demonstrated for glyceollin I [30], which comprised 8% (w/w) of the total isoflavonoids at t=9. Moreover, it was shown that glyceollin I docks in the ER α cavity occupying the same position as the side chain of 4-hydroxy-tamoxifen, a well-known ER α antagonist in breast cells ( Figure 6 ). This side-chain is responsible for its antagonism [30]. No interaction between His524 of ER α, binding most agonists, and glyceollin I was observed. In contrast, glycinol (similar to glyceollin I, except for the prenyl group) binds His524, and does not follow the binding path of the side chain of 4-hydroxy-tamoxifen. Thus, prenylation might be responsible for the different docking poses in ERα and the kind of activity (agonist vs. antagonist) on ER α. Investigation of the estrogenic properties of licorice isoflavonoids ( Chapter 4 ) also suggested that prenylation of isoflavonoids might induce antagonism.

Figure 6 . Structures of E 2 and 4-hydroxy-tamoxifen.

In Chapter 5 , the accumulation of prenylated isoflavonoids is compared with that of non-prenylated isoflavonoids. It is clearly seen that longer induction times promote a higher proportion of prenylated isoflavonoids, reaching similar amounts as non-prenylated isoflavonoids at around t=7. Although prenylation seems related to ER-antagonism, it does not mean antagonism per se, as the prenylated glyceollin II and III, as well as alpinumisoflavone (a prenylated isoflavone), did not seem to have such activity [30, 31]. In Chapter 4 , glabrene also did not show any antagonistic activity. Table 3 , and in more condensed form Figure 7 , summarises the literature data on the correlation of prenylation of isoflavonoids and kind of estrogenic activity (agonism / antagonism). To our knowledge, the literature does not provide evidence that non-prenylated isoflavonoids show antagonism, no matter which isoflavonoid subclass they belong to.

130

Figure 7. Summary of the relationship of the position of prenylation on the isoflavonoid skeleton and ER α activity (agonist/antagonist), based on the overview in Table 3 . According to IUPAC, the carbon numbering of isoflavones differs from that of pterocarpans and coumestans. Therefore, the 4 different prenylation positions are indicated by positions α, β, γ and δ. Weak and strong agonists are distinguished ( +…+, respectively). For antagonists, the potency is more difficult to derive from the literature (+), but it is clear that the potency can vary between compounds. A number of molecules are diprenylated. It is unclear whether prenylation at both sites is essential for antagonistic activity. In case of diprenylation with two different prenyl groups, the + symbols are connected with a dotted line. X, not applicable; ?, no information available. (1) Antagonism not tested; (2) Results on antagonism are ambiguous; 3 out of 4 compounds were antagonists, 1 out of 4 was neither agonistic nor antagonistic; (3) results on the agonistic activity of glabrene require confirmation .

Chapter 6

Table 3 . The influence of prenyl substituents, and their position, on the agonist/antagonist activity towards the ERα of various isoflavonoids. Isoflavonoid Prenyl Position Second Position Type 1 Ref . prenyl Pterocarpans Glycinol - - Agonist [32 ] Glyceollin I Pyran α Antagonist [30 , 33 ] Glyceollin II Pyran β No antagonist [30 ] Glyceollin III Furan 2 β No antagonist [30 ] Bitucarpin A Chain α Weak antagonist [34 ] Phaseollin Pyran δ Antagonist [35 ] Isoflavones Genistein - - Agonist [27 , 31 ] 8-Prenyl -genistein Chain 3 α Antagonist [36 ] Ebenosin Chain α Weak agonist 4 [37 ] Wighteone Chain β n.d. 5 [36 ] 6,8 -Diprenyl -orobol Chain α Chain β Antagonist [31 , 36 ] Isoerysenegalensein E Chain α Chain 6 β Antagonist [31 ] Erysenegalensein E Chain 6 α Chain β No antagonist [31 ] Millewanin G Ch ain α Chain 6 β Antagonist [38 ] Millewa nin H Chain 6 α Chain β Antagonist [38 ] 7,8 -(2,2 -dimethyl - Pyran α Antagonist [39 ] pyran)-Daidzein Alpinumisoflavone Pyran β No antagonist [31 ] Warangalone Chain α Pyran β Weak antagonist [31 ] Auriculasin Chain α Pyran β Weak antagonist [31 , 36 ] Furowanin A Chain α Furan 7 β Antagonist [31 ] Furowanin B Furan 7 α Chain β Antagonist [38 ] 2’,5’ -Diprenylorobol Chain γ Chain δ Antagonist [36 ] 2’ -Prenyl -4’5’ -(2,2 - Chain γ Pyran δ Antagonist [36 ] dimethyl-pyran)orobol Isoflavanones Kievitone Chain α Antagonist [35 ] Isoflavans Gla bridin Pyran α Antagonist Chapter 4 Isoflavenes Glabrene Pyran δ Agonist Chapter 4 Coumestans Coumestrol - - Agonist [27 , 40 ] Glycyrol Chain β Agonist 8 [41 ] 1 Antagonists are referred as weak antagonist, if specified as such in reference. 2 2’’-isopropenyl-furan. 3 1,1-dimethyl allyl. 4 Antagonistic activity not determined. 5 Not determined in yeast assay due to cytotoxicity. 6 2-hydroxy-3-methylbut-3-enyl. 7 2’’-(2-hydroxy-)isopropenyldihydrofuran. 8 >500 times weaker than coumestrol, antagonism not determined.

132

General discussion

Prenylation at the ααα -position seems to promote antagonism more than that at the β-position. The kind of prenylation (chain or pyran) seems to be of secondary importance. Antagonism is not confined to α-ring prenylation. Prenylation at positions γγγ and δδδ also seemed to promote antagonism, but this relationship is less clear because of less representatives. Furthermore, it is unclear whether double prenylation amplifies the antagonistic effect. It appears that single prenylation, at least at the δδδ-position, would also trigger the effect. On the one hand this is corroborated by the ERα-antagonist phaseollin (Table 3 ). On the other hand, from our study on the prenylated isoflavonoids from licorice (Chapter 4 ), glabrene appeared to be an agonist. Clearly, this needs to be confirmed by purification. In summary, prenylation of isoflavonoids seems to promote antagonistic activity on the ERα. The antagonism might be related to the presence of a bulky side chain, hindering the correct conformation of the ER [39]. In two studies, the effect of prenylation of isoflavonoids on their estrogenic activity was investigated. Although antagonism was not studied, generally a loss of estrogenicity was observed upon prenylation of genistein [42]. In another study, prenylation of genistein was shown to induce antagonistic activity [36]. This might suggest a trade-off between agonistic and antagonistic activities, in which the loss of agonism indicates the induction of antagonism. We speculate that the increased agonistic ER activity of the soya extracts is mainly due to the formation of the non-prenylated pterocarpan, glycinol, and the non- prenylated coumestans, particularly coumestrol. It is expected that the extracts are also rich in ER-antagonists, as prenylation of isoflavonoids seems to correlate well with antagonistic properties (e.g. glyceollin I). However, due to the presence of strong agonists, it is impossible to determine antagonistic activity of these extracts in the bioassays, and further fractionation is a prerequisite. Furthermore, the antagonistic properties of purified glyceollidin I, glyceollin VI, A-ring prenylated isoflavones, and 4-prenyl coumestrol, in particular, should be determined. These compounds have potentially the right molecular signature for antagonistic activity with prenylation at position α. In line with this, licorice root and spent might provide a rich source of the ERα-selective antagonists. In addition to glabridin, glabridin derivatives might also potentially possess antagonistic activity as they comply with the structural prerequisites for antagonism. This warrants further investigation regarding their antagonistic activity.

133

Chapter 6

ANTI-MICROBIAL ACTIVITY OF SOYA ISOFLAVONOIDS

As defined in Chapter 1 , phytoalexins generally possess anti-microbial activity. This was out of the scope of this thesis. Predominant phytoalexins in the soya extracts such as glycinol and coumestrol have been reported to have strong anti-bacterial activity against bacteria such as Erwinia carotovora , Pseudomonas glycinea , Escherichia coli , Xanthomonas phaseoli , and Bacillus subtilis [43, 44]. Growth inhibition of food-borne pathogenic bacteria would provide an additional application of soya phytoalexins in food products. As a preliminary experiment, the ethanol extract of Rhizopus -induced soya seedlings, prepared at kilogram-scale, was screened for anti-microbial activity against pathogenic bacterial strains. In this screening assay, 5 strains ( Staphylococcus aureus ATCC 25923, Listeria monocytogenes EGD-e, Salmonella typhimurium S2505, Bacillus cereus ATCC 14579, and Escherichia coli K12) were grown in Mueller Hinton broth at 30 °C in the absence or presence of the soya extract. The growth of the bacteria was monitored by measuring the optical density (OD) at 600 nm over a period of 24 h. The soya extract was applied to the bacteria at a final concentration of 50 mg extract per L culture medium. As shown in Figure 8 , the soya extract showed a complete inhibition of growth of Bacillus cereus ATCC 14579 during the experimental time frame. Similarly, also for Staphylococcus aureus ATCC 25923 the addition of soya extract resulted in a bacteriostatic effect (data not shown). The growth of Listeria monocytogenes EGD-e was also inhibited as indicated by a delayed growth ( Figure 8 ). The extracts did not have any inhibitory effect on the growth of Salmonella typhimurium S2505 and Escherichia coli K12. These results suggest that the soya extract is able to inhibit the growth of Gram-positive bacteria but not Gram-negative bacteria. The anti-bacterial properties of soya extract might provide possibilities for its application in food products to inhibit growth of Gram-positive pathogens. Despite the promising results, it should be taken into account that anti- bacterial activity of the active components are most likely attenuated by the food matrix due to e.g. binding to protein. Further investigation is, therefore, imperative to determine the effectiveness of the soya extract in real food conditions.

134

General discussion

Figure 8 . The inhibitory effects of soya extract on the growth of (A) Bacillus cereus ATCC 14579 and (B) Lysteria monocytogenes EGD-e.

REFERENCES

[1] F. Kuhn, M. Oehme, F. Romero, E. Abou-Mansour, R. Tabacchi, Differentiation of isomeric flavone/isoflavone aglycones by MS 2 ion trap mass spectrometry and a double neutral loss of CO. Rapid Communications in Mass Spectrometry 2003 , 17 , 1941-1949. [2] P. Shen, B. L. Guo, Y. Gong, D. Y. Q. Hong, Y. Hong, E. L. Yong, Taxonomic, genetic, chemical and estrogenic characteristics of Epimedium species. Phytochemistry 2007 , 68 , 1448-1458. [3] M. Stobiecki, A. Staszków, A. Piasecka, P. M. Garcia-Lopez, F. Zamora-Natera, P. Kachlicki, LC-MSMS profiling of flavonoid conjugates in wild mexican lupine, lupinus reflexus. Journal of Natural Products 2010 , 73 , 1254-1260. [4] J. F. Stevens, M. Ivancic, V. L. Hsu, M. L. Deinzer, Prenylflavonoids from Humulus lupulus . Phytochemistry 1997 , 44 , 1575-1585. [5] F. Cuyckens, M. Claeys, Mass spectrometry in the structural analysis of flavonoids. Journal of Mass Spectrometry 2004 , 39 , 1-15. [6] E. F. Queiroz, J. L. Wolfender, K. K. Atindehou, D. Traore, K. Hostettmann, On-line identification of the antifungal constituents of Erythrina vogelii by liquid chromatography with tandem mass spectrometry, ultraviolet absorbance detection and nuclear magnetic resonance spectrometry combined with liquid chromatographic micro-fractionation. Journal of Chromatography A 2002 , 974 , 123-134. [7] U. Justesen, Collision-induced fragmentation of deprotonated methoxylated flavonoids, obtained by electrospray ionization mass spectrometry. Journal of Mass Spectrometry 2001 , 36 , 169-178. [8] C. T. Da Costa, J. J. Dalluge, M. J. Welch, B. Coxon, S. A. Margolis, D. Horton, Characterization of prenylated xanthones and flavanones by liquid chromatography/atmospheric pressure chemical ionization mass spectrometry. Journal of Mass Spectrometry 2000 , 35 , 540-549. [9] G. J. Bennett, H.-H. Lee, Xanthones from guttiferae. Phytochemistry 1989 , 28 , 967-998. [10] A. D'Agostina, G. Boschin, D. Resta, P. Annicchiarico, A. Arnoldi, Changes of isoflavones during the growth cycle of Lupinus albus. Journal of Agricultural and Food Chemistry 2008 , 56 , 4450-4456. [11] X. Hong, X. Wang, E. L. Yong, Y. Gong, Determination of breviflavone A and B in Epimedium herbs with liquid chromatography-tandem mass spectrometry. Journal of Pharmaceutical and Biomedical Analysis 2009 , 49 , 853-857.

135

Chapter 6

[12] D. Nikolic, Y. Li, L. R. Chadwick, G. F. Pauli, R. B. Van Breemen, Metabolism of xanthohumol and isoxanthohumol, prenylated flavonoids from hops ( Humulus lupulus L.), by human liver microsomes. Journal of Mass Spectrometry 2005 , 40 , 289-299. [13] M. Yang, W. Wang, J. Sun, Y. Zhao, Y. Liu, H. Liang, D. A. Guo, Characterization of phenolic compounds in the crude extract of Hedysarum multijugum by high-performance liquid chromatography with electrospray ionization tandem mass spectrometry. Rapid Communications in Mass Spectrometry 2007 , 21 , 3833-3841. [14] Y. Wang, L. Yang, Y. Q. He, C. H. Wang, E. W. Welbeck, S. W. A. Bligh, Z. T. Wang, Characterization of fifty-one flavonoids in a Chinese herbal prescription Longdan Xiegan Decoction by high-performance liquid chromatography coupled to electrospray ionization tandem mass spectrometry and photodiode array detection. Rapid Communications in Mass Spectrometry 2008 , 22 , 1767-1778. [15] P. Shen, S. P. Wong, J. Li, E. L. Yong, Simple and sensitive liquid chromatography-tandem mass spectrometry assay for simultaneous measurement of five Epimedium prenylflavonoids in rat sera. Journal of Chromatography B 2009 , 877 , 71-78. [16] R. Liu, M. Ye, H. Guo, K. Bi, D. A. Guo, Liquid chromatography/electrospray ionization mass spectrometry for the characterization of twenty-three flavonoids in the extract of Dalbergia odorifera . Rapid Communications in Mass Spectrometry 2005 , 19 , 1557-1565. [17] D. Yang, S. Cai, H. Liu, X. Guo, C. Li, M. Shang, X. Wang, Y. Zhao, On-line identification of the constituents of Buyang Huanwu decoction in pig serum using combined HPLC-DAD-MS techniques. Journal of Chromatography B: Analytical Technologies in the Biomedical and Life Sciences 2006 , 831 , 288-302. [18] L. W. Qi, X. D. Wen, J. Cao, C. Y. Li, P. Li, L. Yi, Y. X. Wang, X. L. Cheng, X. X. Ge, Rapid and sensitive screening and characterization of phenolic acids, phthalides, saponins and isoflavonoids in Danggui Buxue Tang by rapid resolution liquid chromatography/diode-array detection coupled with time-of- flight mass spectrometry. Rapid Communications in Mass Spectrometry 2008 , 22 , 2493-2509. [19] T. Nomura, T. Fukai, Phenolic constituents of licorice ( Glycyrrhiza species), Progress in the chemistry of organic natural products , Vol. 73 , Springer, 1998 . [20] A. T. Leung, S. Foster, Encyclopedia of Common Natural Ingredients Used in Food, Drugs, and Cosmetics, Vol. 50 , Second edition, Springer New York, 1996 . [21] M. Tian, H. Yan, K. H. Row, Extraction of glycyrrhizic acid and glabridin from licorice. International Journal of Molecular Sciences 2008 , 9, 571-577. [22] H. J. Kim, H. J. Suh, J. H. Kim, S. C. Kang, S. Park, C. H. Lee, J. S. Kim, Estrogenic activity of glyceollins isolated from soybean elicited with Aspergillus sojae . Journal of Medicinal Food 2010 , 13 , 382-390. [23] S. M. Boué, C. H. Carter, K. C. Ehrlich, T. E. Cleveland, Induction of the soybean phytoalexins coumestrol and glyceollin by Aspergillus . Journal of Agricultural and Food Chemistry 2000 , 48 , 2167- 2172. [24] S. M. Boué, F. F. Shih, B. Y. Shih, K. W. Daigle, C. H. Carter-Wientjes, T. E. Cleveland, Effect of biotic elicitors on enrichment of antioxidant properties and induced isoflavones in soybean. Journal of Food Science 2008 , 73 , H43-H49. [25] S. Feng, L. S. Chin, K. L. Yuan, D. Huang, Fungal-stressed germination of black soybeans leads to generation of oxooctadecadienoic acids in addition to glyceollins. Journal of Agricultural and Food Chemistry 2007 , 55 , 8589-8595. [26] M. K. Bhattacharyya, E. W. B. Ward, Resistence, susceptibility and accumulation of glyceollins I-III in soybean organs inoculated with Phytophthora megasperma f. sp. glycinea . Physiological and Molecular Plant Pathology 1986 , 29 , 227-237. [27] S. O. Mueller, S. Simon, K. Chae, M. Metzler, K. S. Korach, Phytoestrogens and their human metabolites show distinct agonistic and antagonistic properties on estrogen receptor α (ER α) and ER β in human cells. Toxicological Sciences 2004 , 80 , 14-25. [28] T. F. H. Bovee, R. J. R. Helsdingen, I. M. C. M. Rietjens, J. Keijer, R. L. A. P. Hoogenboom, Rapid yeast estrogen bioassays stably expressing human estrogen receptors α and β, and green fluorescent protein: A comparison of different compounds with both receptor types. Journal of Steroid Biochemistry and Molecular Biology 2004 , 91 , 99-109.

136

General discussion

[29] T. F. H. Bovee, M. Thevis, A. R. M. Hamers, A. A. C. M. Peijnenburg, M. W. F. Nielen, W. G. E. J. Schoonen, SERMs and SARMs: Detection of their activities with yeast based bioassays. Journal of Steroid Biochemistry and Molecular Biology 2010 , 118 , 85-92. [30] M. C. Zimmermann, S. L. Tilghman, S. M. Boué, V. A. Salvo, S. Elliott, K. Y. Williams, E. V. Skripnikova, H. Ashe, F. Payton-Stewart, L. Vanhoy-Rhodes, J. P. Fonseca, C. Corbitt, B. M. Collins-Burow, M. H. Howell, M. Lacey, B. Y. Shih, C. Carter-Wientjes, T. E. Cleveland, J. A. McLachlan, T. E. Wiese, B. S. Beckman, M. E. Burow, Glyceollin I, a novel antiestrogenic phytoalexin isolated from activated soy. Journal of Pharmacology and Experimental Therapeutics 2010 , 332 , 35-45. [31] Y. Okamoto, A. Suzuki, K. Ueda, C. Ito, M. Itoigawa, H. Furukawa, T. Nishihara, N. Kojima, Anti- estrogenic activity of prenylated isoflavones from Millettia pachycarpa : Implications for pharmacophores and unique mechanisms. Journal of Health Science 2006 , 52 , 186-191. [32] S. M. Boué, S. L. Tilghman, S. Elliott, M. C. Zimmerman, K. Y. Williams, F. Payton-Stewart, A. P. Miraflor, M. H. Howell, B. Y. Shih, C. H. Carter-Wientjes, C. Segar, B. S. Beckman, T. E. Wiese, T. E. Cleveland, J. A. McLachlan, M. E. Burow, Identification of the potent phytoestrogen glycinol in elicited soybean ( Glycine max ). Endocrinology 2009 , 150 , 2446-2453. [33] S. L. Tilghman, S. M. Boué, M. E. Burow, Glyceollins, a novel class of antiestrogenic phytoalexins. Molecular and Cellular Pharmacology 2010 , 2, 155-160. [34] B. Pinto, A. Bertoli, C. Noccioli, S. Garritano, D. Reali, L. Pistelli, Estradiol-antagonistic activity of phenolic compounds from leguminous plants. Phytotherapy Research 2008 , 22 , 362-366. [35] S. M. Boué, M. E. Burow, T. E. Wiese, B. Y. Shih, S. Elliott, C. H. Carter-Wientjes, J. A. McLachlan, D. Bhatnagar, Estrogenic and antiestrogenic activities of phytoalexins from red kidney bean ( Phaseolus vulgaris L.). Journal of Agricultural and Food Chemistry 2011 , 59 , 112-120. [36] E. M. Ahn, N. Nakamura, T. Akao, T. Nishihara, M. Hattori, Estrogenic and antiestrogenic activities of the roots of Moghania philippinensis and their constituents. Biological and Pharmaceutical Bulletin 2004 , 27 , 548-553. [37] M. Halabalaki, X. Alexi, N. Aligiannis, G. Lambrinidis, H. Pratsinis, I. Florentin, S. Mitakou, E. Mikros, A. L. Skaltsounis, M. N. Alexis, Estrogenic activity of isoflavonoids from Onobrychis ebenoides . Planta Medica 2006 , 72 , 488-493. [38] C. Ito, M. Itoigawa, M. Kumagaya, Y. Okamoto, K. Ueda, T. Nishihara, N. Kojima, H. Furukawa, Isoflavonoids with antiestrogenic activity from Millettia pachycarpa . Journal of Natural Products 2006 , 69 , 138-141. [39] Q. Jiang, F. Payton-Stewart, S. Elliott, J. Driver, L. V. Rhodes, Q. Zhang, S. Zheng, D. Bhatnagar, S. M. Boue, B. M. Collins-Burow, J. Sridhar, C. Stevens, J. A. McLachlan, T. E. Wiese, M. E. Burow, G. Wang, Effects of 7-O substitutions on estrogenic and anti-estrogenic activities of daidzein analogues in MCF- 7 breast cancer cells. Journal of Medicinal Chemistry 2010 , 53 , 6153-6163. [40] G. G. J. M. Kuiper, J. G. Lemmen, B. Carlsson, J. C. Corton, S. H. Safe, P. T. Van Der Saag, B. Van Der Burg, J. Ã. Gustafsson, Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor β. Endocrinology 1998 , 139 , 4252-4263. [41] T. Nomura, T. Fukai, T. Akiyama, Chemistry of phenolic compounds of licorice ( Glycyrrhiza species) and their estrogenic and cytotoxic activities. Pure and Applied Chemistry 2002 , 74 , 1199-1206. [42] G. Kretzschmar, O. Zierau, J. Wober, S. Tischer, P. Metz, G. Vollmer, Prenylation has a compound specific effect on the estrogenicity of naringenin and genistein. Journal of Steroid Biochemistry and Molecular Biology 2010 , 118 , 1-6. [43] L. I. Weinstein, M. G. Hahn, P. Albersheim, Host-pathogen interactions XVIII: Isolation and biological- activity of glycinol, a pterocarpan phytoalexin synthesized by soybeans. Plant Physiology 1981 , 68 , 358-363. [44] F. D. Dakora, D. A. Phillips, Diverse functions of isoflavonoids in legumes transcend anti-microbial definitions of phytoalexins. Physiological and Molecular Plant Pathology 1996 , 49 , 1-20.

137

138

Appendix

Appendix I . Proposed fragmentation pathway of glabrene ( 10 ) in NI mode. The number (bold) inside the parenthesis refers to the compound in Table 1 , Chapter 2 .

140

Appendix II . Proposed fragmentation pathway of glabrone ( 11 ) in NI mode. The number (bold) inside the parenthesis refers to the compound in Table 1 , Chapter 2 .

141

Appendix III . Proposed fragmentation pathway of glabrol (13 ) and 3-hydroxy-glabrol ( 17 ) in NI mode. The numbers (bold) inside the parenthesis refer to the compounds in Table 1 , Chapter 1 .

142

Appendix IV . Proposed fragmentation pathway of glabridin ( 12 ), 3’-hydroxy-4’-O-methyl- glabridin ( 14 ), 4’-O-methyl-glabridin ( 15 ), hispaglabridin A ( 16 ) and B ( 18 ) in PI mode. The numbers (bold) inside the parenthesis refer to the compounds in Table 1 , Chapter 2 .

143

Appendix V . Screening results for prenylated flavonoids in the MS chromatograms of the 70% aq. EtOH, EtOH and EA extracts of Glycyrrhiza glabra in PI mode. All peaks in this table show a neutral in their MS n spectra of 56 u, alone or in combination with a loss of 42 u. (C - prenyl chain; P - pyran ring; np - non prenylated). 2 3 MS MS No * Rt ** (min) Identification [M+H] + MS neutral R.A. u loss42 neutral R.A. u loss56 (42:56) Ratio -type Prenyl neutral R.A. u loss42 neutral R.A. u loss56 (42:56) Ratio -type Prenyl

S1 6.21 257 4 - - - 6 3 2.0 np S2 6.41 257 3 - - - 5 3 1.7 np S3 6.90 257 4 - - - 6 2 3.0 np S4 7.19 271 12 3 4.0 - - 7 0.0 np S5 7.76 563 - - - - <1 32 0.0 np S6 7.88 556 - - - - <1 36 0.0 np S7 8.08 431 - - - - <1 33 0.0 np S8 8.44 255 3 100 <0.1 np S9 8.67 517 - - - - <1 34 0.0 C S10 8.80 285 - 6 0.0 np S11 9.02 315 - 4 0.0 - <1 38 0.0 C S12 9.11 709 - - - - <1 32 0.0 np S13 9.18 739 - - - - - 32 0.0 np S14 9.37 469 <1 2 <0.1 - 2 3 0.7 C S15 9.45 339 4 6 0.7 C S16 9.51 355 - - - - 20 17 1.2 P S17 9.57 473 - - - - - 36 0.0 C S18 9.97 410 - 3 0.0 C - - - - S19 10.09 285 5 100 <0.1 np S20 10.16 410 - 100 0.0 C - - -

S21 10.27 424 - 2 0.0 C - - - S22 10.36 339 <1 100 <0.1 C S23 10.44 327 - - - - 3 1 3.0 P S24 10.52 453 2 1 2.0 - 6 5 1.2 P S25 10.56 327 - - - - 3 1 3.0 P S26 10.79 321 24 25 1.0 C S27 10.89 Formononetin 269 - 33 0.0 np S28 11.18 453 1 - - - 23 24 1.0 C S29 11.27 341 <1 100 <0.1 C S30 11.33 371 36 49 0.7 C S31 11.38 339 <1 100 <0.1 C S32 11.51 323 2 100 <0.1 C S33 11.57 341 - 100 0.0 C S34 11.67 323 1 100 <0.1 C S35 11.79 355 41 100 0.4 C S36 11.90 337 1 7 0.1 C S37 12.10 355 <1 100 <0.1 C S38 12.20 325 3 100 <0. 1 C S39 12.28 321 24 28 0.9 C S40 12.40 323 - 100 0.0 C S41 12.53 322 - 7 0.0 C S42 12.68 337 28 57 0.5 C S43 12.72 339 5 17 0.3 C S44 12.88 353 - 100 0.0 C R.A. – Relative abundance of ion within MS n spectrum < 1 – between 0.5% and 1.0% n.d. - not detected, <0.5% * the ‘S’ refers to a peak found with the screening method in any of the 3 extracts ** The retention time in MS has a delay of 0.20±0.02 min compared to the UV retention time

(Appendix V . continued ) 2 3 MS MS No * Rt ** (min) Identification [M+H] + MS neutral R.A. u loss42 neutral R.A. u loss56 (42:56) Ratio -type Prenyl neutral R.A. u loss42 neutral R.A. u loss56 (42:56) Ratio -type Prenyl

S45 12.97 308 - - - - 3 4 0.8 C S46 13.16 321 63 48 1.3 P S47 13.25 355 - 100 0.0 C S48 13.31 341 - 100 0.0 C S49 13.52 409 - 21 0.0 C S50 13.64 Glabrone 337 100 6 16.7 P S51 13.77 Glabridin 325 - <1 0.0 - 100 1 100 P S52 13.88 389 2 100 <0.1 C 49 51 1.0 C S53 13.91 321 18 64 0.3 C S54 14.08 391 - 100 0.0 C 1 100 <0.1 C S55 14.13 645 1 100 <0.1 C 21 10 2.1 P S56 14.18 353 - 16 0.0 C S57 14.29 Glabrol 393 - 100 0.0 C 21 22 1.0 C S58 14.49 409 - 100 0.0 C 19 100 0.2 C S59 14.66 395 - 82 0.0 C S60 14.72 647 - 60 0.0 C S61 14.73 407 - 100 0.0 C <1 100 <0.1 C S62 14.79 3’ -hydroxy -4’ -O-methyl - 355 - - - - 100 1 100 P glabridin S63 14.80 627 2 7 0.3 np S64 14.89 409 - 100 0.0 C 30 20 1.5 P

S65 14.91 643 2 100 <0.1 C 16 67 0.2 C S66 15.04 647 - 14 0.0 C S67 15.08 661 - 100 0.0 C 1 35 <0.1 C S68 15.14 391 - 100 0.0 C <1 100 <0.1 C S69 15.27 391 - 18 0.0 C S70 15.30 389 2 100 <0.1 C 100 68 1.5 P S71 15.3 3 407 - 100 0.0 C 5 30 1.9 P S72 15.42 645 <1 100 <0.1 C 35 18 1.9 P S73 15.46 627 - 3 0.0 C S74 15.52 4’ -O-methyl -glabridin 339 - - - - 100 1 100 P S75 15.62 645 <1 57 <0.1 np S76 15.73 631 <1 100 <0.1 C 34 6 5.7 P S77 15.88 391 - 100 <0.1 C 6 7 0.9 C S78 15.97 Hispaglabridin A 393 - 100 <0.1 C 21 3 7.0 P S79 16.00 407 - 100 0.0 C 6 16 0.4 C S80 16.08 3-hydroxy -glabrol 409 - 100 0.0 C 1 9 0.1 C S81 16.18 391 1 39 <0.1 C S82 16.34 409 - 42 0.0 C S83 16.44 643 2 63 <0.1 np S84 16.53 393 - 100 0.0 C 10 8 1.3 P S85 16.72 659 - 26 0.0 np S86 16.81 469 - - - - 4 14 0.3 C S87 16.89 388 1 2 0.5 - 11 6 1.8 P S88 17.21 Hispaglabridin B 391 - 3 - - 100 1 100 P R.A. – Relative abundance of ion within MS n spectrum < 1 – between 0.5% and 1.0% n.d. - not detected, <0.5% * the ‘S’ refers to a peak found with the screening method in any of the 3 extracts ** The retention time in MS has a delay of 0.20±0.02 min compared to the UV retention time

Appendix VI . Schematic representation of the different fragmentation pathways of glycinol, glyceollidin I/II and glyceollins I-III in NI mode MS 2. The weight of the arrow indicates the likelihood of the pathway.

148

Appendix VII. MS 2 spectra of daidzein and 2'-hydroxy-daidzein and the MS 3 spectra of A- ring and B-ring prenylated daidzein. The difference of 16 u between the MS 2 fragments of 2'-hydroxydaidzein and daidzein are explained by the presence of an additional hydroxyl- group in 2'-hydroxydaidzein. Similarly, the difference of 12 u between the MS 2 fragments of daidzein and the MS 3 fragments of both A-ring and B-ring prenylated daidzein can be accounted for by the remaining CH group after the degradation of the prenyl group.

149

Appendix VIII. Fractions generated by CPC-fractionation and the presence of the main flavonoids in each fraction determined by UHPLC-MS Fraction 1-5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Yield (mg) 1422 110 141 97 136 108 126 211 345 254 156 66 176 157 115 37 16 94 16 67 86 44 71 80 Peaks n.d. 1 20 6 12 25 22 18 7 4 6 6 15 14 17 30 42 36 30 25 24 18 33 31 22 Non-prenylated - 3 1 2 4 2 1 0 0 0 0 0 1 2 4 9 3 4 2 2 2 4 4 2 Prenylated - 17 5 10 21 20 17 7 4 6 6 15 13 15 26 33 33 26 23 22 16 29 27 20 Single - 13 3 7 9 6 8 4 1 2 3 11 11 13 18 22 21 20 23 19 16 22 20 16 -Chain - 7 2 4 6 4 5 2 0 1 2 7 5 9 13 21 17 18 20 16 14 19 17 15 -Pyran - 6 1 3 3 2 3 2 1 1 1 4 6 4 5 1 4 2 3 3 2 3 3 1 Double - 4 2 3 12 14 9 3 3 4 3 4 2 2 8 11 12 6 0 3 0 7 7 4

Characterization Glabrene ++ ++ ++ ++ + Glabrone + + ++ + Glabridin + + ++ ++ + + + Glabrol + ++ ++ + + 3’-OH-4’OMe-G2 + + ++ ++ 4’-OMe-G3 + + ++ + Hispaglabridin A + ++ ++ Hispaglabridin B ++ ++ + 1 n.d. – not detected. 2 4’-OMe-glabridin 3 3’-OH-4’OMe-glabridin

(Appendix VIII . continued )

Fraction 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

Yield (mg) 80 84 9 6 87 84 54 62 16 51 32 42 5 24 35 27 28 103 90 94 142 88 233 71 Peaks 22 22 12 16 12 9 19 20 15 14 18 21 28 50 45 13 9 5 36 14 28 30 23 24 Non - 2 3 0 0 0 1 4 5 4 4 4 4 4 11 8 4 3 4 21 10 21 28 22 23 prenylated Prenylated 20 19 12 16 12 8 15 15 11 10 14 17 24 39 37 9 6 1 15 4 7 2 1 1 Single 16 17 9 13 9 6 12 12 10 8 12 16 22 33 30 7 5 1 15 4 7 1 1 1 -Chain 15 17 9 13 9 5 9 7 7 5 8 11 13 22 21 5 5 1 14 2 5 0 0 0 -Pyran 1 0 0 0 0 1 3 5 3 3 4 5 9 11 9 2 0 0 1 2 2 1 1 1 Double 4 2 3 3 3 2 3 3 1 2 2 1 2 6 7 2 1 0 0 0 0 1 0 0

Appendix IX. Calculation of the estradiol equivalents (EEQ)

To determine to what extent phytoestrogens with known EC 50 value can account for the estrogenic activity of a tested extract, the measured estrogenic response of the extract was compared to the estrogenic response calculated from the contribution of all known phytoestrogens in the extract. This was performed for extracts of t=10 and t=4 with ER α and ER β, respectively. Both extracts were chosen because they exhibited an estrogenic response between 50 and 90% of the maximum response (the steep part of the E 2 standard curve). Measured at 10 µg/mL, the extract at t=10 displayed a relative response on ER α of 60% compared to the maximum response of E 2. This response corresponds to a -10 concentration of 7.8×10 M E 2. Based on the known contents of genistein, coumestrol and daidzein in the t=10 extract, 5.0, 4.2, and 5.6 µg/mg, respectively, the concentration of genistein, coumestrol and daidzein could be calculated in M, based on the concentration of the extract in the assay (10 mg/L) and the molecular weight (MW) of the three isoflavonoids. Their concentrations in the assay were 1.8×10 -7, 1.6×10 -7, and 2.2×10 -7

M, respectively. In order to determine their E 2 equivalents in M, the concentration (M) of each phytoestrogen was multiplied by its relative estrogenic potency (REP). The sum of -10 the E 2 equivalents of genistein, coumestrol and daidzein was 9.8×10 M, approximately 25% higher than the E2 equivalents measured in the assay. The same calculation was performed for the ER β showing that the total E 2 equivalents contributed by genistein, coumestrol and daidzein was a factor 7 higher than the E 2 equivalents measured in the assay.

152

(Appendix IX. continued )

The estrogenic potential of the extracts of t=10 and t=4 on ER α and ER β, respectively, measured in the assay and calculated on the basis of their genistein, coumestrol and daidzein contents (expressed in E2 equivalents), phytoestrogens with known EC 50 .

2 ER ααα Content in Extraction Content in Final conc. of Conc. in assay MW Conc. RE P E2 equivalents (t=10) mg/g DW yield 1 extract extract in assay (g/L) (M) (M) beans * (mg/g) (µg/mg) (mg/L)

Genistein 0.22 5.0 5.0×10 -5 270 1.8×10 -7 5.0×10 -4 9.2×10 -11 Coumestrol 0.18 43.9 4.2 10 4.2×10 -5 268 1.6×10 -7 5.7×10 -3 8.9×10 -10 Daidzein 0.24 5.6 5.6×10 -5 254 2.2×10 -7 -3 n.a. Total 9.8×10 -10 Actual measurement 7.8 ×10 -10

2 ER βββ Content in Extraction Content in Final conc. of Conc. in assay MW Conc. REP E2 equivalents (t=4) mg/g DW yield 1 extract extract in assay (g/L) (M) (M) beans * (mg/g) (µg/mg) (mg/L) Genistein 0.32 6.7 6.7×10 -6 270 2.4×10 -8 1.1×10 -2 2.7×10 -10 Coumestrol 0.09 47.4 1.9 1 1.9×10 -6 268 7.1×10 -9 2.7×10 -2 1.9×10 -10 Daidzein 0.32 6.7 6.7×10 -6 254 2.6×10 -8 1.1×10 -4 2.9×10 -12 Total 4.7×10 -10 Actual measurement 6.6×10 -11 * Values represent results of a single induction experiment. 1 The amount of extract (mg) that is recovered from beans DW (g) 2 Relative estrogenic potency; REP is normalized to E 2 (REP Estradiol = 1.0) (1) 3 The EC 50 of daidzein was not determined due to low binding affinity MW, molecular weight n.a., not applicable

1. Bovee, T. F. H.; Helsdingen, R. J. R.; Rietjens, I. M. C. M.; Keijer, J.; Hoogenboom, R. L. A. P., Rapid yeast estrogen bioassays stably expressing human estrogen receptors a and b, and green fluorescent protein: A comparison of different compounds with both receptor types. J. Steroid Biochem. Mol. Biol. 2004, 91, (3), 99-109.

154

Summary

Epidemiological and clinical studies have shown that the dietary intake of phytoestrogens is associated with beneficial health effects in the prevention or the treatment of hormone- dependent deficiencies including osteoporosis, cancer, cardiovascular diseases and menopausal symptoms. A large part of the dietary intake of phytoestrogens is contributed by legumes such as soya, either as vegetable, as health ingredient in functional foods, or as dietary supplement (nutraceuticals). Many phytoestrogens belong to the isoflavonoid subclasses of isoflavones, pterocarpans and coumestans, although also other phenolic compounds, such as lignans and stilbenes, are known to display estrogenic activity.

Prenylation refers to the substitution with an C 5-isoprenoid or prenyl group. In a broad sense, prenylation can also refer to the substitution with other groups, including

C10 -isoprenoid (geranyl) or C 15 -isoprenoid (farnesyl). This thesis will be limited to C 5- prenylation and derivatives thereof. Prenylation of isoflavonoids has been associated with modification of their structure-activity relationship with respect to induction of estrogenic responses. Prenylated isoflavonoids are predominantly found within the Leguminosae , and are mostly inducible metabolites as part of the plant’s defensive strategy. Licorice (Glycyrrhiza glabra ) is a well-known legume, the roots of which are also rich in prenylated isoflavonoids, mainly represented by the isoflavan and isoflavene subclasses. Although soya ( Glycine max ) is known as a rich source of non-prenylated isoflavonoids, germination of soya beans under biotic stress (i.e. in the presence of fungus) can induce the formation of prenylated isoflavonoids. To determine whether licorice and soya extracts that are rich in prenylated and non-prenylated isoflavonoids can lead to novel and unique phytoestrogen-rich health ingredients, research on their characterisation, induction and estrogenic activity is imperative. Therefore, the aims of this thesis were (1) to provide a characterisation of isoflavonoids, in particular of the prenylated isoflavonoids occurring in soya and licorice, (2) to increase the estrogenic activity of soya beans by the application of malting in the presence of a food-grade fungus, and (3) to correlate the agonistic/antagonistic estrogenicity of extracts, and fractions of the extracts, with the presence of prenylated isoflavonoids. An overview of the family of flavonoids is presented in Chapter 1 . This overview mainly focuses on the isoflavonoid class, its structural classification and biosynthesis. The identification of the various flavonoids by mass spectrometry is reviewed and the various kinds of prenyl-substitution of flavonoids are addressed. Flavonoids are known to exhibit estrogenic activity and a large part of the dietary intake is contributed by economically important legume species such as soya and licorice. Both soya and licorice are also known as rich sources of prenylated isoflavonoids. Special attention is paid to methods that enhance the estrogenic potential of soya including the challenging of germinated soya beans. Although many prenylated flavonoids have been described in the literature, there were no methods available that specifically screen for their presence in complex mixtures,

156

prior to purification. In Chapter 2 , a method is presented that allows such screening. This method was based on the combined chromatographic separation and mass spectrometric analysis (RP-UHPLC/MS), and was applied to an ethyl acetate extract of licorice root. The prenylated flavonoids showed a preferential degradation of the prenyl-group in positive- 2 3 ion (PI) mode MS or MS , resulting in the characteristic neutral losses 42 (C 3H6) and 56 + (C 4H8). It appeared that the ratio between the relative abundances of [M+H-42] and [M+H-56] + could be used for determining the kind of prenylation, i.e. chain and ring-closed (pyran ring). The application of biotic stress on soya seedlings is known to induce the formation of glyceollins, which are estrogenically active prenylated pterocarpans. The process in which these bioactive pterocarpans are formed in soya seedlings is called induction, and consisted of 3 steps: soaking, germination and fungus-challenged germination. Chapter 3 describes the characterisation of novel prenylated isoflavonoids that were induced in addition to prenylated pterocarpans (glyceollins I, II and III) in Rhizopus -challenged soya seedlings grown on laboratory scale. By means of the screening method described in Chapter 2 , it was shown that in addition to glyceollins, comprising ~40% of the total isoflavonoid content, also prenylated isoflavones and prenylated coumestans were formed, comprising ~7% of the total isoflavonoid content. The prenylated isoflavones and coumestans were substituted with a prenyl-chain on either the A-ring or the B-ring. Furthermore, the fragmentation pathway for 6a-hydroxy- pterocarpans in negative ion (NI) mode was proposed, showing A-ring and B-ring fragment ions resulting from C-ring and D-ring cleavages. The presence of a ring-closed prenyl chain (pyran or furan ring) yielded predominantly x,y RDA fragments; its absence gave exclusively x,y -H2O RDA fragments. In Chapter 4 , fractions obtained from the fractionation of a licorice root extract by centrifugal partitioning chromatography were screened for estrogenic activity using an in vitro yeast estrogen receptor (ER) bioassay. About one third of all tested fractions were considered estrogenically active on one or both ER-subtypes (α and β). Some of the fractions displayed superinduced responses, being higher than that of the 17β-estradiol

(E 2) standard. This superinduction observed did not appear to be an artefact as co- incubation with an antagonist attenuated the estrogenic response indicating that the signal was ER-mediated. The observed superinduction was most likely caused by stabilisation of the ER, the yeast-enhanced green fluorescent protein or a combination of both. The estrogenic activity of some fractions was associated with the presence of glabrene, a prenylated isoflavene. Although the predominant phytoestrogen of licorice root, glabridin, did not show any agonistic activity on both ER-subtypes, glabridin was found to be a potent ERα antagonist, reducing the response of E 2 by approximately 80% at a concentration of 6 × 10 -6 M. To accurately monitor the changes in both isoflavonoid profile as well as in estrogenic activity, the induction experiment was performed at kilogram-scale by

157

application of malting technology used in the brewing industry ( Chapter 5 ). In a time- course experiment, performed in five-fold, the total isoflavonoid content was shown to increase by a factor 10-12 on dry weight (DW) basis in 9 days. Also, a substantial increase in the diversity of the isoflavonoid profile was observed. The pterocarpan content stabilized at 2.24 mg daidzein equivalents (DE) per g after 9 days. In addition to the glyceollins I, II and III, other less common pterocarpans, glycinol, glyceollidin I/II, glyceofuran, glyceollin IV and VI, were formed. All pterocarpans were formed in similar amounts ranging from 0.18 to 0.42 mg DE per g after 10 days. The content of prenylated isoflavones and prenylated coumestans reached up to 0.30 and 0.06 mg daidzein equivalents (DE) per g, respectively, after 10 days. These changes in the chemical composition were accompanied with changes in bioactivity, as the extracts showed a gradual increase in agonistic activity towards both the ERα and ERβ during the 10-day induction process. The measured estrogenicity of the extracts on both ERs was lower than the theoretically calculated estrogenicity, indicating the presence of antagonists. The antagonistic activity of the extract was associated with the presence of prenylated isoflavonoids. In the last chapter ( Chapter 6 ), the fragmentation of prenyl-groups in both PI and NI mode MS n are discussed and an update of the interpretation guideline, presented in Chapter 2 , was provided. In addition, the fragmentation of prenyl groups other than the prenyl chain and pyran ring was addressed, as well as the current knowledge on the fragmentation of pterocarpans and coumestans. Regarding the induction of prenylated isoflavonoids in soya, specific attention is focussed on the growth conditions of challenged soya seedlings. Speculations on the choice of elicitor and pre-treatment in relation to optimal induction of phytoalexins are given. Apart from its estrogenic activity, the anti- bacterial activity of the extract from Rhizopus -challenged soya seedlings was briefly addressed. The by-product of industrial licorice processing, spent, is evaluated as an alternative source for prenylated isoflavonoids with estrogenic activity. Furthermore, prenylation of isoflavonoids and its effect on estrogenic activity are discussed. An overview is presented, based on literature and data from this thesis, on the association between prenylation of isoflavonoids and agonism / antagonism. We hypothesize that prenylation of isoflavonoids reverses the agonistic activity into antagonism. The implications of employing prenylated isoflavonoid-enriched extracts from licorice and induced soya-seedlings in the development novel phytoestrogen-enriched health ingredients are discussed.

158

Samenvatting

Epidemiologische en klinische studies hebben aangetoond dat de dagelijkse inname van phytoestrogenen in verband kan worden gebracht met gezondheidseffecten met betrekking tot de preventieve behandeling van hormoongerelateerde condities zoals bijvoorbeeld botontkalking, kanker, hart- en vaatziekten en menopausale klachten. Het leeuwendeel van de phytoestrogenen in ons dieet zijn afkomstig van peulvruchten zoals sojabonen, geconsumeerd als groente, maar ook als gezondheidsingredient in zgn. functional foods of in voedingssupplementen. Veel phytoestrogenen behoren tot de isoflavonoide subklassen isoflavonen, pterocarpanen en coumestanen, maar ook andere phenolische verbindingen, zoals lignanen en stilbenen, staan bekend om hun estrogene activiteit.

Prenylering verwijst naar de substitutie met een C 5-isoprenoid of prenyl groep. In de brede zin van het woord, omvat prenylering ook de substitutie met andere groepen zoals de C10 -isoprenyl (geranyl) of C 15 -isoprenoid (farnesyl). Dit proefschrift zal zich beperken tot C 5-prenylering en afgeleiden daarvan. Prenylering van isoflavonoïden worden in verband gebracht met veranderingen met betrekking tot hun structuur-functie relatie betreffende de inductie van estrogene activiteit. Geprenyleerde isoflavonoïden worden voornamelijk aangetroffen in de Leguminoseae (Vlinderbloemigen) en zijn over het algemeen induceerbare stoffen die fungeren als onderdeel van het afweer- mechanisme van de plant. Zoethout (Glycyrrhiza glabra ) is een bekende plantensoort binnen de Vlinderbloemigen, waarvan de wortels rijk zijn aan geprenyleerde isoflavonoïden, met name vertegenwoordigd door de isoflavan en isoflaveen subklassen. Hoewel soja ( Glycine max ) bekend staat als een rijke bron van niet-geprenyleerde isoflavonoïden, leidt de ontkieming van soja in de aanwezigheid van een biologische vorm van stress tot de vorming van geprenyleerde isoflavonoïden. Om te bepalen of extracten van zoethoutwortel en soja, rijk aan zowel geprenyleerde als niet-geprenyleerde isoflavonoïden, geschikt zijn als basis voor de ontwikkeling van nieuwe, innonatieve phytoestrogeen-rijke gezondheidsingredienten, is onderzoek naar karakterisering, inductie en de bepaling van estrogene activiteit een vereiste. Dit resulteerde in de volgende doelen van het proefschrift (1) de karakterisering van isoflavonoïden, met name geprenyleerde isoflavonoïden uit soja en zoethout, (2) het verhogen van de estrogene activiteit van sojabonen door middel van vermouting in de aanwezigheid van een voedselveilige schimmel, en (3) verband leggen tussen de estrogene en antiestrogene activiteit van zowel de extracten als de fracties ervan met de aanwezigheid van geprenyleerde isoflavonoïden. Hoofdstuk 1 wordt ingeleid met een overzicht van de familie der flavonoïden. Dit overzicht richt zich met name op de klassen van de isoflavonoïden, hun classificatie op basis van structuur en biosynthese. Vervolgens wordt ingegaan op de identificatie van de verschillende flavonoïden door middel van massa spectrometrie (MS), alsmede de identificatie van de verschillende prenyl-substituties van flavonoïden. Flavonoïden staan erom bekend dat zij estrogeen actief kunnen zijn en dat ze in grote mate aanwezig zijn in

160

onze dagelijkse voeding in de vorm van economische belangrijke gewassen zoals zoethout en soja. Zowel zoethout als soja staan bekend als rijke bronnen van geprenyleerde isoflavonoïden. Nadruk wordt met name gelegd op methoden die het estrogene potentieel van soja kunnen verhogen inclusief het onder stress ontkiemen van sojabonen. Ondanks hun uitgebreide beschrijving in de wetenschappelijke literatuur, bestaan er geen methoden die het mogelijk maken om te screenen voor de aanwezigheid van geprenyleerde flavonoïden in complexe mengsels alvorens ze op te zuiveren. Hoofdstuk 2 beschrijft een methode die deze screening mogelijk maakt. Deze methode is gebaseerd op een combinatie van chromatografische scheiding met massa spectrometrische analyse (RP-UHPLC/MS), toegepast op een ethylacetaat extract van zoethoutwortels. De prenyl- groepen van de geprenyleerde flavonoïden breken bij voorkeur af in positive ion (PI) mode 2 3 MS of MS , hetgeen leidde tot de vorming van karakteristieke neutral losses 42 (C 3H6) en 56 (C 4H8). Op basis van de verhouding tussen de relatieve hoeveelheid karakteristieke fragmenten [M+H-42] + en [M+H-56] + kon worden bepaald of de betreffende prenyl- substitutie een zgn. chain of een pyraan-ring (een ring-gesloten chain ) was. Het is bekend dat de toepassing van biotsche stress op soja zaailingen leidt tot de vorming van glyceollinen. Glyceollinen zijn estrogene pterocapranen die gesubstitueerd zijn met een prenyl-groep. Het process dat leidt tot de vorming van deze bioactieve pterocarpanen wordt ook wel inductie genoemd en bestaat uit 3 stappen: weken, ontkieming en ontkieming in de aanwezigheid van een schimmel ( fungus-challenged germination ). Hoofdstuk 3 beschrijft de inductie van sojabonen op laboratoriumschaal met behulp van Rhizopus microsporus var. oryzae waarbij de vorming van nieuwe geprenyleerde isoflavonoïden wordt beschreven, naast de vorming van de bekende geprenyleerde pterocarpanen glyceollin I, III en III. Met behulp van de screeningsmethode die wordt beschreven in Hoofdstuk 2 , kon worden aangetoond dat behalve de glyceollinen, die ~40% van het totale isoflavonoïden gehalte uitmaken, ook geprenyleerde isoflavonen alsmede geprenyleerde coumestanen worden gevormd, die samen ~7% van het totaal uitmaken. De geprenyleerde isoflavonen en coumestanen bleken gesubstitueerd met een prenyl-chain op óf de A-ring óf de B-ring. Daarnaast werd een overzicht van MS fragmentatie routes voorgesteld voor de fragmentatie van 6a-hydroxy- pterocarpanen in negative ion (NI) mode, die de vorming laat zien van verschillende A-ring en B-ring fragmenten via de splijting van zowel C als D-ring. De substitutie met een pyraan of furaan ring resulteerde in voornamelijk x,y RDA fragmenten terwijl de afwezigheid van x,y zo’n substituent uitsluitend resulteerde in -H2O RDA (retroDiels-Alder) fragmenten. Het screenen van estrogeniciteit van fracties van het zoethout extract die werden verkregen na fractionering van het extract middels centrifugal partitioning chromatography (CPC) wordt beschreven in Hoofdstuk 4 . De estrogeniciteit van de fracties werd bepaald door middel van een in vitro estrogeniciteitsassay bestaande uit transgene gistcellen die de estrogeen receptor (ER) bevatten. Een derde van de fracties bleek estrogeen te zijn op ERα, de ERβ of beide receptoren. Sommige fracties bleken een

161

hogere response te genereren dan de standaard 17β-estradiol (E 2). Deze zgn. superinductie bleek geen artefact te zijn aangezien de estrogene activiteit met behulp van een antagonist (gedeeltelijk) kon worden uitgedoofd. Dit toonde aan dat de activeteit tot stand kwam door middel van binding aan de ER. Een logische verklaring is dat de stoffen uit de geteste fracties in staat waren de stabiliteit van de receptor en/of het reporter-gen te vergroten. De estrogene activiteit van bepaalde fracties werd in verband gebracht met de aanwezigheid van glabreen, een geprenyleerde isoflaveen. Alhoewel de meest voorkomende phytoestrogeen in de zoethoutwortel, glabridin, geen estrogene activiteit vertoonde op beide estrogeen receptoren, bleek glabridin een sterke antagonist op de

ERα waarbij de response van E 2 met ~80% werd verminderd met een concentratie van 6 × 10 -6 M. Om de veranderingen in het isoflavonoïden profiel alsmede de estrogene activiteit nauwkeurig te volgen, werd het soja inductie experiment uitgevoerd op kilogramschaal door de toepassing van vermoutingstechnologie die wordt gebruikt in bierbrouwerijen ( Hoofdstuk 5 ). In een tijdsverloopexperiment, dat werd uitgevoerd in vijfvoud, bleek het totale isoflavonoïdengehalte toe te nemen met een factor 10-12 in 9 dagen op basis van drooggewicht. Verder was er een significante toename in de diversiteit van het isoflavonoïden profiel. Het gehalte aan pterocarpanen stabiliseerde rond de 2.24 mg daidzein equivalenten (DE) per g drooggewicht bonen na 9 dagen. Behalve glyceollinen I, II en III werden ook andere, minder bekende pterocarpanen zoals glycinol, glyceollidin I/II, glyceofuran, glyceollinen IV en VI gevormd. De gehaltes van deze minder bekende pterocarpanen varieerden tussen de 0.18 en 0.42 mg DE per g drooggewicht bonen na 10 dagen. Het gehalte aan geprenyleerde isoflavonen en coumestanen behaalden waardes van, respectievelijk, 0.30 en 0.06 mg DE per g drooggewicht bonen na 10 dagen. Deze veranderingen in samenstelling van isoflavonoïden leidde tot een toename van de bioactiviteit, hetgeen zich vertaalde in een toename in estrogeniciteit van de extracten op beide estrogeen receptoren tijdens het 10-daagse inductie experiment. Desondanks werd er een lagere estrogeniciteit gemeten dan verwacht op basis van de theorische berekende estrogeniciteit, hetgeen de aanwezigheid van antagonisten suggereerde. De antagonitische activiteit van het extract werd in verband gebracht met de aanwezigheid van geprenyleerde isoflavonoïden. In het laatste hoofdstuk ( Hoofdstuk 6 ) wordt de MS fragmentatie van prenyl groepen in zowel PI en NI mode bediscussieerd en wordt er een bijgewerkte versie van de zgn. interpretation guideline uit Hoofdstuk 2 besproken. Daarnaast werd tevens de fragmentatie van minder voorkomende prenylgroepen besproken alsmede de huidige stand van zaken omtrent de kennis van fragmentatie van pterocarpanen en coumestanen. Wat betreft de inductie van geprenyleerde isoflavonoïden in soja wordt er met name ingegaan op de ontkiemingscondities van gestresste soja zaailingen. Er wordt gespeculeerd over de beste keuze voor (a)biotische stimulatie en ontkiemingscondities die leiden tot de optimale inductie van phytoalexinen. Behalve de estrogene activiteit wordt

162

ook de antiestrogene activiteit besproken van de extracten van sojabonen die zijn ontkiemd in de aanwezigheid van Rhizopus microsporus var. oryzae . Ook wordt het spent, het bijproduct van industriële extractie van zoethoutwortel, geëvalueerd als alternatieve bron van geprenyleerde isoflavonoïden met estrogene activiteit. Verder wordt het effect van prenylering van isoflavonoïden op hun estrogene activiteit besproken. Op basis van een literatuuroverzicht wordt een mogelijk verband tussen prenylering en agonisme/antagonisme samengevat en besproken. Voortvloeiend uit dit overzicht stellen wij de hypothese dat prenylering van isoflavonoïden leidt tot een omkering van agonisme in antagonisme. De gevolgen hiervan voor het gebruik van extracten die rijk zijn aan geprenyleerde isoflavonoïden uit zoethout en geïnduceerde soja zaailingen in de ontwikkeling van nieuwe, innovatieve phytoestrogeenrijke gezondheidsingredienten worden verder uiteengezet.

163

164

Acknowledgements

At first I considered a PhD program as a one-person’s project supervised by very clever people. I’ve come to realise that the further you tumble down the rabbit hole the more you appreciate and need the input and help from the people around you, both personally and professionally.

First and foremost I offer my sincerest gratitude to Marian and Jean-Paul. Thank you Marian for having the faith and confidence in me by offering me this wonderful opportunity to become a PhD. Your comments during every meeting we’ve had always helped me to stay focused. Jean-Paul, I consider myself fortunate to have you as my co- promoter. Our meetings were often a combination of science, humour and inspiration. Your ideas, often born from a different point of view, helped me to think out of the box and set my priorities. I could always count on you and walk in your office, especially the past few months when I was under constant pressure of many deadlines. Both you and Marian have been a source of inspiration for me.

Harry, as my promoter you were always guarding the quality of my work as well as my deadlines and planning. Although sometimes confronting, you made me realise some of my weak points. However, both you and Jean-Paul were always ready to explain and advise me but also to challenge me to re-define myself.

My work on the soya germination experiments wasn’t without help. Petra, Judith, Ingrid, Heidy and Rob from the 4 th floor (Food Microbiology), thanks for the hands-on and expertise. Hennie, your explanation, enthusiasm, patience and care were indispensable during the malting experiments at Holland Malt. I always enjoyed having pleasant conversations in Lieshout or on the phone. Nikos, without you I could never have pulled off all those malting experiment at Holland Malt. You were a quick learner and a hard worker. The estrogenicity screening would not have been possible without the help of Toine, Gerrit en Richard. Thanks for patiently teaching me how to master the yeast assay at the RIKILT. Teus, Johan, Sam, Tom, Karel, Martijn, as members of the FND project team, thanks for all the valuable discussions we’ve had the past 4 years. Daniel, Florian and Miroslav, my dear colleagues from Frutarom Wädenswil, thanks for making my life a little easier making those nice licorice and soya extracts. I’m looking forward to our next challenge!

At Food chemistry I’ve made many wonderful colleagues with whom I could share my happiness over successful experiments and my frustrations and struggles with things like the UHPLC-MS system. Thanks Karin and Maaike for the heads-up, tips and tricks in the beginning of my glorious Ph.D. career. Maaike, my meetings with Jean-Paul were never really complete without you behind your computer, commenting on our discussion topics guarding the quality of the discussions. Roomies from 509! Koen, Takao, Lieke, Marijn, Stéphanie, Jianwei, Jun, Anne and Uttara, thanks for all the wonderful times and lovely talks we’ve had the past 4 years. I’m going to miss you a lot! Koen, thanks for the

166

(pep)talks and support. Marijn, I will never forgot all our talks on Utrecht CS, in the train to Ede-Wageningen (or visa versa) and on the bike to Wageningen and back. Weer of geen weer, het was nooit een saai ritje op de fiets met jou ! Tomas, Maxime, Olga, Erika, Susan and Loes, I enjoyed working together with you on my projects. As my precious (former) students, you’ve taught me a great deal and I appreciate all the hard the work and dedication. Koos, Stefan, Yvonne V, Yvonne W, organising and leading our Ph.D. study-trip to China was one of the most extraordinary experiences in the past 4 years. From the walking through the Forbidden City, climbing the Great Wall to eating noodle soup from a plastic bag in the streets of Beijing: I’m glad to have shared it together with you. Jolanda, als spil van de afdeling was je mijn rots in de branding and oase van rust in al het tumult. Raluca, Carlos, Wibke, Tomas, Maxime and Yannick, I cherish our magnificent phytonutrient meetings on Monday morning. Those time are like the twist of lemon in my martini. Milou good luck with the follow up of my project as a Ph.D. student and thanks for critically reading my thesis. Tomas and Maxime, you have a special place in my heart. Starting as my first, respectively, second student kickstarting my soya research, becoming my dear colleagues and laboratory mates and being my paranymphs. Thanks for all the good times and being a true friend.

Winter and Young. As my elder brothers, you were there when I was standing on the crossroads of my life. Your words of wisdom have been and always will be important for me. Thank you for everything. I’m grateful that I can call you my family. Patrick and Bayu, my fellow martial artist masters! Thanks for being there during the good and bad days. You brought balance and feedback allowing me to let go of life’s hardships and enjoy the moment. Yu, our talks about life over lunch were my breathing room, my moment of rest in the day. You helped me to relativise things in life. I’ll be in Beijing next year to attend your wedding!

Family. To have them is one of the most important things in the world. Mama and Donny, you have been very important for me before, during and after my PhD work. We have been though a lot together and it strengthened our bonds. Thank you for understanding all the times I said I was ‘too busy’. In the end I know I can always count of you. Thank you mama for giving me a life full of chances and choices. Thank you for your endless support and love. Mootje, ich bin ech heel blie mit dich .

Joop. Papa. Despite our differences and fights we are more alike than we would like admit. During my life as a PhD, I was also confronted with my strengths, strengths of which I know some were yours as well. Even after your passing, you were able to push me and stimulate me to excel. Thank you for everything and bringing out the good in me.

167

Carlos. As my step-father you helped me to open my eyes. You’ve helped me through hard and difficult times with your advice and care. You were a true father to me teaching me as a son. In the difficult moments of my PhD time I often recollect our conversations and the lessons you’ve taught me. Despite the fact that you and Joop had to leave this life too early, I’m grateful for everything that I’ve learned from you both.

Ye-Ji, 자기야, thanks baby for all your love, support and patience. We’ve been through a lot together, you and me. You’ve been there from the beginning and followed me throughout my journey. With your background in linguistics, a passion for languages and no affinity for science, you always tried your best to understand me when I was talking about science (again). Sorry for the times I tried to explain molecular ionisation by electrospray of isoflavonoids in mass spectrometry to you ^^;; You always believed in me and never doubted for a second. When I needed you, you were always there. I admire your courage and I am grateful for your trust in me, in us. 잘 들어 그리고 꼭 기억해. 난 죽을 때까지 너만 선택해...

Looking back, forward and where I am right now: without a doubt, I consider myself a fortunate and rich man.

Thanks everyone,

Rudy

168

About the author

CURRICULUM VITAE

Ruud (Rudy) Simons was born on the 11 th on June 1980 in Gouda. Starting his secondary education at the Goudse Scholen Gemeenschap in Gouda in 1992, he graduated from the Kalsbeek College in Woerden in 1998. After his graduation he directly started his study biology at Leiden University, specialising in molecular biology and analytical chemistry. After obtaining the degree of Master of Sciences ( Doctorandus ) in September 2003, he started working at TNO in the department of Applied Plant Sciences. In 2005, Rudy started working for Frutarom Netherlands as a Product Technologist. In March 2007 he was appointed as PhD candidate at the Laboratory of Food Chemistry of Wageningen University. The results of this project, under the auspices of the Food and Nutrition Delta, are summarised in this thesis. From the 5 th of April 2011, Rudy continued his work at Frutarom Netherlands in Veenendaal as a member of the Medical Science and Regulatory Affairs team.

170

LIST OF PUBLICATIONS

Arno Hazekamp, Ruud Simons , Anja Peltenburg-Looman, Melvin Sengers, Rianne van Zweden and Robert Verpoorte, Preparative isolation of cannabinoids from Cannabis sativa by centrifugal partition chromatography, Journal of Liquid Chromatogromatography and Related Technologies , 2004 , 27 , 2421-2439

Rudy Simons , Jean-Paul Vincken, Edwin J. Bakx, Marian A. Verbruggen and Harry Gruppen, A rapid screening method for prenylated flavonoids with ultra-high-performance liquid chromatography/electrospray ionisation mass spectrometry in licorice root extracts, Rapid Communications in Mass Spectrometry , 2009, 23, 3083–3093

Rudy Simons , Jean-Paul Vincken, Maxime C. Bohin, Tomas F.M. Kuijpers, Marian A. Verbruggen and Harry Gruppen, Identification of prenylated pterocarpans and other isoflavonoids in Rhizopus spp. elicited soya bean seedlings by electrospray ionisation mass spectrometry, Rapid Communications in Mass Spectrometry , 2011 , 25 , 55–65

Rudy Simons , Jean-Paul Vincken, Loes A.M. Mol, Susan A.M. The, Toine F.H. Bovee, Teus J.C. Luijendijk, Marian A. Verbruggen, Harry Gruppen, Agonistic and antagonistic estrogens in licorice root ( Glycyrrhiza glabra ), Analytical and Bioanalytical Chemistry , 2011 , in press

Rudy Simons , Jean-Paul Vincken, Nikolaos Roidos, Toine F.H. Bovee, Martijn van Iersel, Marian A. Verbruggen, Harry Gruppen, Increasing the isoflavonoid content, diversity and estrogenicity of soy upon simultaneous malting and challenging by fungus, Journal of Agricultural and Food Chemistry , 2011 , in press

171

OVERVIEW OF COMPLETED TRAINING ACTIVITIES

Discipline specific activities

Courses Metabolomics, Wageningen, the Netherlands (2007) Mass Frontiers, Dreieich, Germany (2008) Ecophysiology of the gastro-intestinal tract, Helsinki, Finland (2009)

Conferences and meetings XXIV International Conference on Polyphenols, Salamance, Spain (2008) IUFOST 15 th World Congress on Food Science and Technology, Shianghai, China (2008) Food Allergy Symposium, Hangzhou, China (2008) Jiangnan University Doctoral Symposium, Wuxi, China (2008) Ghent University Food Chemistry Symposium, Ghent, Belgium (2009) International Conference on Polyphenols and Health, Harrogate, United Kingdom (2009) Food Chemistry colloquia, Wageningen, the Netherlands (2007-2010) Phillip Morris International Symposium, Lausanne, Switzerland (2010) Frutarom Workshop, Wädenswil, Switzerland (2010) XXV International Conference on Polyphenols, Montpellier, France (2010)

General courses

Vitafoods Europe, Geneva, Switzerland (2006) CTP, CEP and Active Substance Master File, Heidelberg, Germany (2006) International Soy and Health Conference, Düsseldorf, Germany (2006) Project and Time Management, Wageningen, the Netherlands (2007) Ph.D. Introduction Week, De Bilt, the Netherlands (2007) Science, the Press and the General Public, Wageningen, the Netherlands (2009) Mobilising your - Scientific - Network, Wageningen, the Netherlands (2009)

Additional activities

Preparation Ph.D. research proposal, Wageningen, the Netherlands (2006) Phytonutrient meetings, Wageningen, the Netherlands (2007-2011) Preparation of Ph.D. study trip China, Wageningen, the Netherlands (2008) Ph.D. study trip China (2008) Ph.D. study trip Belgium (2009) Ph.D. study trip Switzerland and Italy (2010)

172

173

174

175

Research presented in the Ph.D. dissertation was financially supported by the Food & Nutrition Delta of the Ministry of Economic Affairs, the Netherlands.

Printing: Ponsen & Looijen b.v., Ede

Cover design: Diana van Houten, www.dianavanhouten.nl

Rudy Simons, 2011

176