Planta Med 2013; 79(07): 580-590
DOI: 10.1055/s-0032-1328463
Women's Health
Mini Reviews
Georg Thieme Verlag KG Stuttgart · New York

Phytoestrogenic Potential of Cyclopia Extracts and Polyphenols

Ann Louw
1  Departments of Biochemistry, University of Stellenbosch, Stellenbosch, Matieland, South Africa
,
Elizabeth Joubert
2  Departments of Food Science, University of Stellenbosch, Stellenbosch, Matieland, South Africa
3  Post-Harvest and Wine Technology, ARC (Agricultural Research Council of South Africa) Infruitec-Nietvoorbij, Stellenbosch, South Africa
,
Koch Visser
1  Departments of Biochemistry, University of Stellenbosch, Stellenbosch, Matieland, South Africa
› Author Affiliations
Further Information

Correspondence

Dr. Ann Louw
Department of Biochemistry, University of Stellenbosch, Stellenbosch
Private Bag X1
Matieland (Stellenbosch) 7602
South Africa
Phone: +27 2 18 08 58 73   
Fax: +27 2 18 08 58 63   

Publication History

received 14 December 2012
revised 19 February 2013

accepted 14 March 2013

Publication Date:
22 April 2013 (online)

 

Abstract

Cyclopia Vent. species, commonly known as honeybush, are endemic to Southern Africa. The plant is traditionally used as an herbal tea but several health benefits have recently been recorded. This minireview presents an overview of polyphenols found in Cyclopia and focusses on the phytoestrogenic potential of selected polyphenols and of extracts prepared from the plant.


#

Introduction

Cyclopia species (family Fabaceae; tribe Podalyrieae) are part of the fynbos biome and endemic to the coastal and mountainous regions of the Western and Eastern Cape Provinces of South Africa. The plant may grow up to heights of 3 m in the wild and is distinguished by trifoliate leaves and sweet smelling deep yellow flowers with an indented calyx [1] ([Fig. 1]). Although more than twenty species of Cyclopia have been described [2], the commercially important species include C. genistoides, C. sessiliflora, C. intermedia, and C. subternata. Fermented (oxidised) Cyclopia is traditionally used as an herbal tea, called honeybush tea, which is acclaimed for its distinct sweet aroma and fragrant flavour. Recently, unfermented honeybush has also been added to the market. Cyclopia is one of the few South African plants to have made the transition from regional use to commercial product [3], and in 2011 a total of 174 tons of Cyclopia was exported, mostly to Germany (37 %), the Netherlands (29 %), USA (14 %), and UK (12 %) (data supplied by Soekie Snyman, SA Rooibos Council, 2012).

Zoom Image
Fig. 1 Shoots of C. subternata (left) and C. genistoides (right) with distinctive yellow flowers having an indented calyx, characteristic of Cyclopia species. (Color figure available online only.)

Cyclopia has traditionally also been used for medicinal purposes, including as a restorative, as an expectorant, and to promote appetite [4]. Research into the phenolic composition of Cyclopia spp. [5], [6], [7] has been crucial in identifying value-adding opportunities in the arena of health promoting attributes. Foremost amongst these have been the demonstration of antioxidant properties [8], [9], inhibition of tumour development [10], [11], and antidiabetic potential [12], [13]. Furthermore, scrutiny of phenolic composition coupled to anecdotal claims of Cyclopia as of use in stimulating milk production [14] and alleviating menopausal symptoms has led to recent research on the phytoestrogenic potential of Cyclopia. This minireview will focus on the polyphenol content of Cyclopia and the phytoestrogenic potential of selected polyphenols identified in this genus and extracts from the shoots and leaves of the plant.


#

Phenolic Composition of Cyclopia

The phenolic composition of a number of commercially important Cyclopia species has been investigated due to the relevance of these constituents for bioactivity of their herbal teas and extracts. In-depth studies, making use of NMR to unequivocally elucidate chemical structures, deal only with C. intermedia and C. subternata [5], [6], [7], [15]. Generally, Cyclopia species are characterised by the presence of the xanthone, mangiferin, with the co-occurrence of its 4-C-glucoside regioisomer, isomangiferin, and the flavanone, hesperidin, an O-rutinoside of hesperetin, in relatively large quantities [16]. Other classes of compounds identified in C. intermedia are flavonols, flavones, isoflavones, and coumestans, as well as some C6-C1 and C6-C2 secondary metabolites [5], [6]. Apart from luteolin, none of the latter compounds has been found in detectable quantities in C. intermedia extracts by HPLC analysis. The isoflavone orobol was isolated from C. subternata [7]. In an in vitro culture, C. subternata produces glucosides of the isoflavone aglycones, calycosin, pseudobaptigenin, and formononetin, present in C. intermedia [5], [15]. Recent investigations demonstrated the presence of benzophenones and dihydrochalcones in C. subternata [15], [17]. An iriflophenone-di-O,C-hexoside, an eriodictyol-di-C-hexoside, 3-hydroxyphloretin-3,5-di-C-hexoside, and vicenin-2 (apigenin-6,8-di-C-glucoside) were tentatively identified in C. subternata, based on UV-Vis, LC-MS, and LC-MS/MS characteristics of the compounds [17]. [Fig. 2] depicts phenolic compounds present in C. subternata.

Zoom Image
Fig. 2 Structures of major phenolic compounds of C. subternata and minor compounds with estrogenic activity present in the leaves and stems of some Cyclopia spp. (* indicates that the position or identity of the glycosyl moiety is not certain; bold text indicates the class of compound).

The abundance of C-glycosides, both in terms of content and number of compounds ([Fig. 1], [Table 1]), has implications concerning stability during processing and in vivo. The C-C bond is very stable and resistant to acid and intestinal enzymes able to hydrolyse O-glycosides, but evidence of C-C bond-cleaving reactions by human intestinal bacteria is growing [18], [19], [20].

Table 1 Phenolic composition of leaves and extracts (g · 100 g−1 dry basis) of unfermented Cyclopia subternata.

Compound

Leaves [92]
(n = 6)

Aqueous extract [16]
(n = 6)

Aqueous extract [17]
(n = 64)

Methanol extract [44]
(n = 1)

aPosition and/or identity of glycosyl moiety not certain; previous designation, bcompound 9, ccompound 8, dcompound 12, ecompound 11, funknown 2, gunknown 1

Mangiferin

1.22 ± 0.35

2.73 ± 1.65

0.93 ± 0.42

1.91

Isomangiferin

0.38 ± 0.05

0.86 ± 0.28

0.47 ± 0.12

0.77

Hesperidin

0.62 ± 0.17

0.64 ± 0.36

2.21

Eriocitrin

0.23 ± 0.06

0.32 ± 0.07

0.55 ± 0.15

1.25

Eriodictyol glucoside a

0.35 ± 0.07 b

Iriflophenone-3-C-β-glucoside

0.25 ± 0.06

0.82 ± 0.44 c

0.47 ± 0.29

3-Hydroxyphloretin-3,5-di-C-hexoside a

0.54 ± 0.13

Phloretin-3,5-di-C-glucoside

0.41 ± 0.01

0.86 ± 0.20 d

1.05 ± 0.34

1.22 f

Scolymoside

0.48 ± 0.32

0.68 ± 0.62 e

0.49 ± 0.24

2.04 g

Luteolin

0.09

Relatively high levels of certain phenolic compounds are present in the leaves of C. subternata ([Table 1]). These values could vary substantially as recently demonstrated by De Beer et al. [17] for seedling plants. Several of the compounds, including mangiferin, isomangiferin, iriflophenone-3-C-glucoside, scolymoside, the 7-O-rutinoside of luteolin, and eriocitrin, the 7-O-rutinoside of eriodictyol, occur in higher levels in aqueous extracts prepared from the leaves, while hesperidin, the 7-O-rutinoside of hesperetin, and the dihydrochalcone C-glycosides are predominant in the stems. Although natural variation is a contributing factor, trace or undetectable quantities of luteolin by HPLC-DAD in aqueous extracts, whilst present in the methanol extract ([Table 1]), are attributed to poor solubility of this aglycone in water.


#

Phytoestrogenic Potential of Cyclopia Polyphenols and Extracts

Phytoestrogenic potential may be defined in terms of the mechanism of action of the endogenous hormone 17β-estradiol (E2) [21]. According to this definition, compounds with phytoestrogenic potential would act through at least one of the main isoforms of the estrogen receptor (ER), namely ERα or ERβ [22], and act as agonists, antagonists, or selective ER modulators (SERMS) via ER signalling pathways [21] ([Fig. 3]). Phytoestrogens are, however, also considered to be endocrine disruptors and as such the definition used by regulatory bodies in both the USA and Europe could be useful [23], [24]. The European Commission State of the Art Assessment of Endocrine Disruptors, for example, defines estrogenicity in terms of “binding to the estrogen receptor(s) (ER), ER activation, cell proliferation in ER-competent cells and physiological responses (proliferation of uterine tissue in rodents, induction of vitellogenin in fish)” [24].

Zoom Image
Fig. 3 Steps in ER signalling used to evaluate estrogenicity. E = estrogenic compound, ER = estrogen receptor, ERE = estrogen response element. (1) Binding of an estrogenic ligand to the ER may be evaluated by ligand-binding assays, (2) binding of ligand-activated ER to an ERE in the promoter of an estrogen responsive gene may be evaluated by promoter-reporter studies using an ERE-containing promoter reporter or by measuring mRNA levels of select ER-responsive genes, and (3) downstream biological effects such as cell proliferation or hypertrophy of the uterus may be measured using the E-screen or uterotrophic assay, respectively.

Although several assays have been suggested to evaluate estrogenic activity [25], for the purposes of this review we will evaluate the phytoestrogenic potential of both the polyphenols shown to be present in Cyclopia and extracts prepared from Cyclopia in terms of their in vitro ability to either bind to ERα or ERβ, to induce or prevent activation of ER-responsive promoters, or to cause cell proliferation in ER-responsive cells (e.g., E-screen in MCF-7 cells, a breast cancer cell line) or in terms of their in vivo responses in known estrogenic tissues such as the uterus ([Fig. 3], [Tables 2], [3], and [4]). In addition, where it was not apparent that the ER was involved, we used evidence of loss of activity via ICI 182,782, an ER antagonist, as confirmation of ER involvement.

Table 2 Known [5], [6], [7], [15] Cyclopia polyphenols that have not been tested for estrogenic potential.

Class of compound

Specific compound(s)

Xanthone

isomangiferin

Flavanone

eriodictyol-5-O-glucoside, eriodictyol-7-O-glucoside, naringenin-5-O-glucoside, isosakuranetin

Flavone

5-deoxyluteolin, scolymoside, isorhoifolin, vicenin-2

Flavonol

kaempferol-5-O-glucoside, kaempferol-6-C-glucoside, kaempferol-8-C-glucoside

Methylinedioxyflavanol derivative

3′4′-methylinedioxyflavanol apiosyl-glucoside

Isoflavone

formononetin apiosyl-glucoside, afrormosin, rothindin, wistin

Methylinedioxyisoflavone derivative

pseudobaptigenin, fujikinetin

Coumestan

flemichapparin, sophoracoumestan B

Benzophenone

iriflophenone-3-C-β-glucoside

Dihydrochalcone

phloretin-3′,5′-di-C-β-glucoside

Benzaldehyde derivative

benzaldehyde apiosyl-glucoside

Phenylethanoid derivative

tyrosol,3-methoxy-tyrosol, 4-glucosyltyrosol, phenylethanol apiosyl-glucoside

Table 3 Phytoestrogenic potential of polyphenols found [5], [6], [7], [15] in Cyclopia.

Polyphenol

Estrogenic effect

Test for estrogenic effect

Reference

Test system

Test model

a  ICI 182,782: an estrogen receptor antagonist

Xanthones

Mangiferin

No

ER binding assay

COS-1 cells + hERα or hERβ

[30], [32]

Fluorescence ERα competitor assay kit

[45]

ERE promoter reporter assay

COS-1 cells + hERα or hERβ

[32]

Flavanones

Hesperetin

No

ER binding assay

COS-1 cells + hERα or hERβ

[30]

MCF-7 cells

[93], [94]

Yes

ERE promoter reporter assay

Yeast cells + hERα

[34]

Yeast cells + hER

[36]

U2OS cells + hERα or hERβ

[33]

Estrogen responsive genes

PC12 cells ± ICI a

[95]

Cell proliferation assay

MCF-7 cells ± ICI

[33]

Hesperidin

No

ER binding assay

COS-1 cells + hERα or hERβ

[30]

ERE promoter reporter assay

MCF-7 cells

[43]

Eriodictyol

Yes

ER binding assay

COS-1 cells + hERα or hERβ

[30]

ERE promoter reporter assay

Yeast cells + hER

[96]

Eriocitrin

Yes

ER binding assay

COS-1 cells + hERα or hERβ

[30]

Naringenin

Yes

ER binding assay

COS-1 cells + hERα or hERβ

[30], [32]

Nonisotopic ERβ-based assay

[37]

ERE promoter reporter assay

COS-1 cells + hERα or hERβ

[32]

MCF-7 cells

[43], [97]

U2OS cells + hERα or hERβ

[33]

Yeast cells + hERα; hER; ERα or ERβ

[34], [35], [95]

Estrogen responsive genes

BT-474 cells

[98]

Cell proliferation assay

MCF-7 cells ± ICI

[32], [33]

No

Uterotrophic assay

Immature rats; mice

[34], [84]

Narirutin

Yes

ER binding assay

COS-1 cells + hERα or hERβ

[30]

Prunin

Yes

ERE promoter reporter assay

MCF-7 cells

[43]

Flavones

Luteolin

Yes

ER binding assay

COS-1 cells + hERα or hERβ

[30], [32]

Nonisotopic ERβ-based assay

[37]

MCF-7 cells

[46]

ERE promoter reporter assay

MCF-7 cells

[43], [46]

COS-1 cells + hERα or hERβ

[32]

Estrogen responsive genes

BT-474 cells

[98]

Cell proliferation assay

MCF-7 cells ± ICI

[32]

Diosmetin

Yes

ERE promoter reporter assay

Yeast cells + hERα

[34]

Isoflavones

Formononetin

Yes

ER binding assay

hERα or hERβ

[38]

ERα or ERβ

[99]

COS-1 cells + hERα or hERβ

[30], [32]

Nonisotopic ERβ-based assay

[37]

No

Rabbit uterine estrogen receptor

[100]

Yes

ERE promoter reporter assay

COS-1 cells + hERα or hERβ

[32]

MCF-7 cells ± ICI

[43], [101]

Yeast cells + hERα; hERα or hERβ

[34], [40], [102]

Cell proliferation assay

MCF-7 cells ± ICI

[32], [101]

Uterotrophic assay

Ovariectomised mice

[41]

Calycosin

Yes

ER binding assay

Erα and Erβ competitor assay kit

[38]

ERE promoterreporterassay

MCF-7 cells

[42]

Uterotrophic assay

Ovariectomised mice

[41]

Calycosin-7-O-glucoside

Yes

ERE promoter reporter assay

MCF-7 cells

[43]

Orobol

Yes

ER binding assay

ERα and ERβ competitor assay kit

[103]

ERα or ERβ

[104]

ERE promoter reporter assay

Yeast cells + hERα

[105]

U2OS cells + hERα

[105]

Ononin (formononetin-7-O-glucoside)

Yes

ERE promoter reporter assay

MCF-7 cells

[43]

Flavanols

(−)-Epigallocatechin gallate

Yes

ER binding assay

hERα or hERβ

[94]

Mouse uterine estrogen receptor

[94]

Gal4 promoter reporter assay

MCF-7 cells + hERα or mERβ + 17m5-G-Luc

[94]

No

ERE promoter reporter assay

HeLa cells + hERα or hERβ

[95]

Coumestans

Medicagol

No

ER binding assay

Rabbit uterine estrogen receptor

[100]

Phenolic carboxylic acid

p-Coumaric acid

No

Uterotrophic assay

Ovariectomised rats

[106]

Table 4 Phytoestrogenic potential of polyphenols and extracts of unfermented C. genistoides and C. subternata.

Species

Extract

P104 [32]

SM6Met [44]

C. genistoides

C. subternata

a  Whole cell bindings were performed in COS-1 cells transfected with hERα or hERβ [32] and in MCF-7 cells that contain both hERα or hERβ [44]. b  RBA or relative binding affinity is expressed relative to that of E2 (100 %) and was calculated as follows: 100 × IC50 (E2)/IC50 (test compound). c  Values represent an average of values from different extractions of the same plant material. d  ERE promoter reporter assays were performed in COS-1 cells transfected with hERα or hERβ [32] or in T47D-KBluc cells that contain both hERα or hERβ [44]. e  RII or relative induction index is expressed relative to that of E2 (100 %) and was calculated as follows: 100 × EC50 (E2)/EC50 (test compound) for potencies and 100 × fold (test compound)/fold (E2) for efficacies. f  Cell proliferation assays were performed in MCF-7 cells. Verhoog et al. performed assays in the presence and absence of ICI 182,782 [32]. g  Not detected. h  Previously ʼUnknown 1’. i  Previously ʼUnknown 2

ER binding a (RBA b ± SEM c )

ERα: 0.1195 ± 0.0567 %
ERβ: 0.0004 ± 0.0001 %

0.0802 ± 0.0139 %

ERE promoter reporter assay d (RII e )

Potency ± SEM

ERβ: 1.0490 ± 0.1287 %

0.0102 ± 0.0032 %

Efficacy ± SEM

ERβ: 103.2 ± 1.1 %

57.6 ± 2.4 %

Cell proliferation assay f (RII)

Potency ± SEM

0.0072 ± 0.0069 %

0.0579 ± 0.0325 %

Efficacy ± SEM

99.1 ± 2.3 %

78.5 ± 6.6 %

Polyphenols (g · 100−1 g dry extracts ± SEM)

  • Mangiferin

3.935 ± 0.329

1.85

  • Isomangiferin

4.998 ± 0.097

0.75

  • Eriocitrin

ND g

1.25

  • Hesperidin

1.503 ± 0.226

1.87

  • Luteolin

0.097 ± 0.001

0.04

  • Scolymoside h

ND

1.82

  • Phloretin-3,5-di-C-glucoside i

ND

1.27

Although in vivo studies have been considered the “gold standard” for the evaluation of estrogenicity, many authors have not conducted such studies, and thus we have to rely on in vitro results. In terms of in vitro results, it is important to establish that a hierarchy in terms of sensitivity has been established, with the E-screen generally considered the most sensitive assay [26], [27], [28]. Furthermore, although binding to the ER may be considered a prerequisite for estrogenic activity and is certainly the most characteristic mode of action of phytoestrogens [29], receptor binding assays cannot distinguish agonists from antagonists or SERMs [26]. Assays relying on the activation of ER-responsive promoters (both of artificial ERE-containing promoter reporters and endogenous ERE-containing estrogen responsive genes) and the E-screen are more appropriate assays to distinguish agonists from antagonists and SERMs [26]. Furthermore, to distinguish activation of ERα from activation via ERβ, cell lines expressing these receptors separately have to be utilised. MCF-7 cells, used in the E-screen, contain both ERα and ERβ and thus lack the ability to discriminate between the roles of the ER isoforms [25]. In addition, the uterotrophic assay is primarily an assay to verify ERα-mediated in vivo effects, and no appropriate in vivo assay for ERβ has been established [25].

Initially, we wanted to standardise our comparison of the estrogenic potential of polyphenols in Cyclopia using the relative binding affinity (RBA) and relative induction index (RII) where binding and activation are expressed relative to the values for E2 (calculated as follows: 100 × IC50 or EC50 (E2)/IC50 or EC50 (test compound), however, we found that few papers provide quantitative data. Thus most of our comparisons of estrogenic activity of the polyphenols present in Cyclopia ([Table 3]) rest on qualitative and not quantitative data.

Most of the polyphenols present in Cyclopia have, to our knowledge, not been tested for estrogenicity ([Table 2]). For example, the dihydrochalcone phloretin-3′,5′-di-C-β-glucoside, the flavone scolymoside, and the benzophenone iriflophenone-3-C-β-glucoside, all present in relatively high concentrations in C. subternata ([Table 1]), have not been tested ([Table 2]).

[Table 3] summarises data for compounds that have been tested for estrogenicity in different assay systems. Mangiferin, the major xanthone in Cyclopia species ([Table 1]), has been shown to have no estrogenic activity both via ER binding assays and ERE-promoter reporter assays ([Table 3]). Although isomangiferin has not been tested ([Table 2]), it is unlikely to have estrogenic activity as it is a regioisomer of mangiferin ([Fig. 2]). The phenolic acid p-coumaric acid and the coumestan medicagol have both been tested but found not to be estrogenic ([Table 3]).

Of the flavanones present in Cyclopia, most have been tested for estrogenicity. Prunin (naringenin-7-O-glucoside), one of the rarer flavanones, is estrogenic, while of the glycosylated flavanones present in relatively high concentrations in Cyclopia ([Table 1]), like eriocitrin and hesperidin, only eriocitrin is estrogenic ([Table 3]). Eriodictyol and naringenin, as well as their rutinosyl derivatives, eriocitrin and narirutin bind to ER, although therutinosyl derivatives bind with a lower affinity than their corresponding aglycones. Specifically, in a competitive binding assay, eriodictyol and naringenin displaced 44 % and 70 % of 1 nM tritiated E2 from ERβ, respectively, while their corresponding rutinosyl derivatives displaced 28 % and 28 %, respectively [30]. Naringenin is interesting as it has been shown to be estrogenic in vitro using the usual array of screening assays, namely ER-binding, activation of ERE-responsive promoters both in promoter reporter studies and with endogenous genes, yet in vivo, using the immature uterotrophic assay, it does not display estrogenicity ([Table 3]). This may suggest that naringenin is not absorbed or is inactivated, either during hepatic metabolism or by gut bacteria, and highlights the importance of validating these parameters [31]. On the other hand, it may also suggest that naringenin does not transactivate via ERα, the ER responsible for uterotrophic action, but rather via ERβ, as borne out by some [32], but not by other [33], [34], [35] promoter reporter studies. Hesperetin and its rutinosyl derivative, hesperidin, do not bind ER, although hesperetin, but not hesperidin, does transactivate an ERE-containing promoter reporter, which can probably be ascribed to the lower activity of glycosalyted derivatives relative to their aglycones. Furthermore, hesperetin activates estrogen responsive genes and causes cell proliferation in the E-screen via an ER-mediated mechanism as ICI 182,782 antagonises the response. This suggests that the ER-binding assay may not be sensitive enough to evaluate weak estrogenicity, which is further borne out by the fact that in three studies where naringenin and hesperetin were directly compared, hesperetin was a weaker agonist [33], [34], [36]. Specifically, Breinholt and Larsen [36] report EC50 values of 89.6 µM and 0.3 µM, while Promberger et al. [34] report 2 % and 80 % efficacy for hesperetin and naringenin, respectively, in ERE-containing promoter reporter studies. Liu et al. [33] also clearly show that hesperetin is weaker than naringenin at causing both cell proliferation in the E-screen and activation in promoter reporter studies. The lower activity of hesperetin relative to naringenin may be ascribed to the methyl functional group found on the B-ring of hesperetin ([Fig. 2]). The flavanol (−)-epigallocatechin gallate, however, was found to be estrogenic by binding to ER and via the GAL4 promoter assay (a very artificial system in which the ER is fused to a GAL4 element), but not via the ERE-containing promoter reporter assay ([Table 3]). This suggests that, contrary to what we have suggested for hesperetin, namely that ER binding may not be sensitive enough to test for weak estrogenic activity, some compounds may bind ER but not display estrogenicity in other assays.

Of the flavones present in Cyclopia only two, luteolin and diosmetin, have been tested for estrogenicity, and both are estrogenic ([Table 3]). Luteolin is present in a methanol extract from C. subternata ([Table 1]) and has been shown to be estrogenic via ER-binding, ERE-containing promoter assays, and estrogen responsive genes, as well as by stimulating cell proliferation in the E-screen. It has, however, not been tested in vivo. Work from our laboratory suggests that luteolin binds preferentially to ERβ, with an RBA of 0.52 % for ERβ, while for ERα the RBA is 0.0025 % [30], [32] and that it has a similar affinity for ERβ as naringenin [30], [32], [37]. In promoter reporter assays, luteolin has a lower potency but higher efficacy via ERβ than naringenin, specifically it has a potency of 3.53 × 10−3 mg/mL (12.3 µM) versus the potency of 1.04 × 10−4 mg/mL (0.0382 µM) of naringenin and a efficacy of 3.69-fold versus a 2.99-fold induction by naringenin. However, unlike naringenin it does transactivate via ERα, with a potency of 1.97 × 10−3 mg/mL (6.88 µM), which is just slightly higher than via ERβ. Yet, in the E-screen, it has a lower potency (2.54 × 10−6 mg/ml or 0.00887 µM) than naringenin (3.27 × 10−8 mg/ml or 0.00012 µM) suggesting that in terms of a biological response in physiologically relevant tissues, it may favour ERβ.

Although the isoflavones shown to be present in Cyclopia are not observed in quantifiable amounts ([Fig. 2], [Table 1]), many of them are estrogenic ([Table 3]). Of these, formononetin and calycosin have been thoroughly tested, both in vitro and in vivo, and generally show a slight preference for ERβ in ER binding assays [30], [32], [38], [39]. These compounds differ only on the B-ring in that calycosin has a 3′-OH moiety. In promoter reporter studies, the ER isoform preference for formononetin is not so clear [32], [40], while both compounds are uterotrophic, with calycosin being more potent than formononetin [41], [42], suggesting that both must act via ERα. Here again we observe the phenomenon of the glycoside being less estrogenic than its corresponding aglycone, with calycosin showing greater estrogenic activity via a promoter reporter construct in MCF-7 cells than calycosin-7-O-glucoside [43]. Orobol, with OH groups at the 3′ and 4′ positions, and ononin, the 7-O-glucoside of formonentin, are also both estrogenic but here their activity appears to be similar to that of calycosin-7-O-glucoside and not to be preferentially via ERβ ([Table 3]).

The presence of polyphenols with phytoestrogenic capabilities in the plant material of Cyclopia species ([Table 3]) raised the question of whether extracts from the plant material will have phytoestrogenic capabilities. One cannot simply assume that the estrogenicity of the pure compounds will be transferred to extracts of the plant material as varying levels of polyphenols, as well as the presence of various polyphenols with varying levels of estrogenicity, might modulate the effects observed with pure polyphenols. To address this issue, examination of the phytoestrogenicity of crude extracts prepared from the plant material of various commercially cultivated Cyclopia species [30], [32], [44] as well as the HPLC analyses of these extracts to identify the polyphenols present is warranted. We chose two extracts for discussion ([Table 4]), P104 (methanol extract) from C. genistoides as it was found to have the highest binding affinity for both the ER subtypes [32], and SM6Met (methanol extract of plant material following extraction with ethyl acetate and ethanol) from C. subternata as it had the highest potency when compared to other extracts [44]. P104 bound to both ERα and ERβ, albeit with a lower potency than that of E2, and had a higher affinity for ERα. This correlates with previous studies that showed a slightly higher displacement of E2 from ERα than from ERβ by P104 [30]. Despite binding to ERα with a higher affinity, P104 was not able to activate an ERE containing promoter reporter construct through ERα, but was able to do so through ERβ with an efficacy similar to that of E2, although its potency was much lower. In addition, P104 induced cell proliferation of MCF-7 cells, but it was less potent than E2. SM6Met has also been shown to bind to the ER by performing whole cell binding assays in MCF-7 cells. Unfortunately, these results cannot distinguish between binding to specific ER isoforms as MCF-7 cells contain both ERα and ERβ. Similar to P104, SM6Met also activated an ERE containing promoter reporter construct and induced cell proliferation in MCF-7 cells and like P104, SM6Met had a lower potency than E2 in both assays. The extracts were analysed with HPLC, and [Table 4] shows the polyphenols detected. Apart from these, the extracts were also screened for narirutin, eriodictyol, naringenin, hesperetin, and formononetin. Although these polyphenols were not present in quantifiable amounts, one cannot exclude the possibility of their presence and thus the effect they may have on the estrogenicity of the whole extract. The unidentified compounds in the extract of Mfenyana et al. [44] have since been tentatively identified ([Table 4]) as the flavone, scolymoside, and the dihydrochalcone, phloretin 3′,5′-di-C-β-glucoside. The presence of unidentified compounds was also previously indicated for P104 [32], but they were not quantified. Comparison of [Tables 3] and [4] may allow the deduction of which of the polyphenols might be causing the phytoestrogenicity of the extracts. Both extracts contain the xanthones mangiferin and isomangiferin, but as they are not phytoestrogenic [30], [32], [45] ([Tables 2] and [3]), it is unlikely that they are contributing. Hesperidin also does not bind to hERα or hERβ and is unable to induce an ERE containing promoter reporter construct [30], [43], however, its aglycone hesperetin, despite showing no binding to ER, does transactivate ERE-containing promoters and causes cell proliferation in the E-screen ([Table 3]). As glycosides are likely to be metabolised to their aglycones in vivo, hesperidin should not be discounted for in vivo studies, however, for in vitro testing, it is unlikely to contribute to the estrogenicity of the extracts. Luteolin has been shown to bind to both ER isoforms [30], [32], [37], [46], to activate an ERE promoter reporter construct through both isoforms [32], [43], [46], and to induce proliferation of a breast cancer cell line ([Table 3]). The amount of luteolin present was, however, shown to be too low to explain the degree of phytoestrogenicity observed for the P104 [32] or SM6Met [44] extract. On the other hand, scolymoside, the 7-O-rutinoside of luteolin, may be important in vivo. The flavanone eriocitrin was quantified in SM6Met, but not in P104 ([Table 4]). Eriocitrin has been shown to bind to ERβ [30], but no further tests for estrogenicity have been performed ([Table 3]). To our knowledge, scolymoside and phloretin 3′,5′-di-C-β-glucoside tentatively identified in SM6Met have not been tested for phytoestrogenicity ([Table 2]). Taken together, no concrete conclusions regarding the polyphenols responsible for the phytoestrogenic effect of extracts of Cyclopia can be drawn. Some of the identified polyphenols still need to be tested for phytoestrogenicity, and the desired answer might be found in the results from these studies. We cannot, however, exclude the possibility that the effect seen with the Cyclopia extracts is the result of a fine balance between different polyphenols present in varying amounts with varying phytoestrogenic potential (agonistic, antagonistic, or SERM activity via either ERα or ERβ) and that synergism or antagonism could play a role with multiple polyphenols targeting multiple ER isoforms [47].


#

Blanket Claims for Phytoestrogenic Potential of Cyclopia

Caution should be exercised in making blanket claims for the phytoestrogenic potential of all harvestings of Cyclopia. Research indicates that variations in the polyphenol composition or content as well as the phytoestrogenic potential of individual harvestings of a specific Cyclopia species may differ ([Table 5]). For example, C. genistoides dried methanol extracts differed remarkably in their ability to induce cell proliferation in the E-screen assay with three out of the six harvestings displaying such low levels of activity that EC50 values could not be determined ([Table 5]). Even amongst the harvestings with higher activity, there was considerable variation with M7 and NP105 extracts displaying 1.4- and 3.3-fold less activity than NP104. In addition, the concentration of luteolin, a polyphenol with proven phytoestrogenic potential ([Table 3]), also varied between harvestings with a 2.6-fold difference between the harvesting with the highest concentration (M9) and that with the lowest concentration (NP104 or NP105) of luteolin ([Table 5]). This variability in polyphenol content is even more pronounced both quantitatively and qualitatively between species of Cyclopia with, for example, eriocitrin varying between undetectable in the C. genistoides aqueous extract to 0.47 % of the aqueous extract of unfermented C. subternata [8].

Table 5 Variation in phytoestrogenic potential and polyphenol content of C. genistoides harvestings.

Farm

Harvesting date

Dried methanol extract

E-screen in MCF-7 cells RII c

Luteolin (g · 100−1 g dry extracts)

a  Data from [44]; b  data from [32]; c  RII (relative induction index) = EC50 E2/EC50 extract; d  ND = RII could not be determined as activity was too low

Koksrivier/Overberg a

22 January 2002

M7

9.8 × 10−5

0.13

Reins/Albertina a

01 April 2003

M8

ND d

0.12

Reins/Albertina a

22 April 2004

M9

ND

0.25

Koksrivier/Overberg b

15 March 2001

NP104

1.4 × 10−4

0.097

Koksrivier/Overberg b

28 March 2001

NP105

4.3 × 10−5

0.097

Koksrivier/Overberg b

31 March 2003

NP122

ND

0.104

The lack of standardisation, both in terms of levels of active substances and activity levels, of botanical and dietary supplements plagues the industry. Combined with little to no regulation by national bodies regulating drug use in most countries, this has led to contrary and inconsistent findings relating to health benefits, which has damaged the credibility of the industry [48]. Thus for claims of phytoestrogenic activity in Cyclopia, individual harvestings would have to be tested for activity until such time as a marker compound(s) shown to be related to activity can be identified.


#

Potential Usage of Phytoestrogens

Estrogen plays an important role in the development of the female reproductive tract, secondary sex characteristics, and in reproductive behaviour [49]. However, estrogen also influences the growth of hormone-dependent cancers such as breast cancer [50].

Hormone replacement therapy (HRT), which includes estrogen combined with or without progesterone, is given to alleviate the symptoms of menopause, and advocates of HRT believe that it also confers long-term benefits regarding cardiovascular disease, bone preservation, and general well-being [51], [52]. Although the efficacy, superiority, and cost effectiveness of estrogen in the treatment of menopausal symptoms is accepted [53], recent large randomised clinical trials [54], [55] and observational studies [56] on HRT have modified the risk/benefit perception. Specifically, increased risk of breast cancer and cardiovascular disease has raised concerns amongst the public [57], and the Endocrine Society statement of 2010 now recommends use of HRT with the lowest effective dose and for the shortest duration possible [58].

The double-edged sword of estrogen has prompted the search for alternatives in the management of menopause, and phytoestrogens have been suggested as a viable alternative, due to their potential to modulate estrogen action [59], [60]. In addition, epidemiological studies suggest that Asian populations who consume 20–50 mg soy/day have fewer occurrences of hormone-dependent diseases, including menopausal symptoms, osteoporosis, and breast cancer and that this lower incidence is not due to under reporting or genotypic factors [53], [61], [62], [63].

Pharmacological validation of claimed health benefits for phytoestrogens has, however, only recently been undertaken and most work has focused on in vitro assays to establish biological activity while large, well-designed in vivo studies have lagged behind [64]. Molecular aspects of phytoestrogens that have been heralded as positive regarding health benefits include the fact that phytoestrogens generally have orders of magnitude lower potency than estrogen [53], [65], display estrogen agonist activities in the presence of low levels of estradiol (post-menopausal) and antagonistic activity in the presence of high levels of estradiol (premenopausal) [48], exhibit partial selectivity for ERβ, the ER isoform believed to attenuate the proliferative effect of ERα [66], [67], and many act like SERMs, making them safer for breast and endometrial tissue [29], [48], [68]. Furthermore, phytoestrogens have additional diverse beneficial biological effects, such as anti-inflammatory, antioxidant, and anticancer effects [65], [69].

Several studies and reviews have evaluated the health potential of phytoestrogens for treating post-menopausal symptoms by maintaining bone density, decreasing cardiovascular disease and hot flashes, and in preventing or treating estrogen-dependent cancers such as breast, prostate, endometrial, and colon cancer [29], [48], [53], [70], [71], [72], [73]. Although there is contradictory scientific proof of the effectiveness of phytoestrogens, specifically soy and red clover isoflavones, for the treatment of vasomotor menopausal symptoms, such as hot flushes [29], [73], [74], for other symptoms, such as osteoporosis and cardiovascular disease, the data to date strongly suggests efficacy. Specifically, phytoestrogens, such as coumestrol, genistein, daidzein and its metabolite equol as well as extracts from soy, black cohosh, and red clover, appear to slow bone loss and improve bone density [29], [48], which is positive for osteoporosis, while for cardiovascular disease, phytoestrogens, primarily from soy, are beneficial in decreasing LDL and triglycerides, while increasing HDL [48], [53]. In addition, several studies have suggested that phytoestrogen use, mainly flavones and isoflavones from soy, is associated with a reduced risk of breast cancer [67], [75], [76], [77].

Despite beneficial effects of phytoestrogens being reported, results have, however, not always been favourable or reproducible [73]. For example, although some studies suggest that soy food intake does correlate with reduced risk or recurrence of breast cancer [78], [79], other studies have found no such association between isoflavone intake and breast cancer risk [80], [81]. The diversity in results may be attributed to, amongst others, the fact that a wide variety and doses of botanicals have been used and the fact that standardisation of formulations are not currently required making comparison between studies difficult [29], [48], [70]. In addition, an evaluation of effects of phytoestrogenic preparations on health is complicated by the fact that exact formulations and concentrations of active constituents are not always known and studies are often retrospective (relying on recall of diet). Furthermore, the fact that there has never been a study comparable in size to the Million Womenʼs or WHI studies investigating side effects of phytoestrogen use should encourage caution. This is especially relevant as many consumers base their beliefs of both efficacy and safety on source rather than evidence [29]. Despite this caveat, there is no current data suggesting that dietary phytoestrogens promote hormone-dependent cancers in humans, and thus phytoestrogens can probably be used safely on a long-term basis [53], [73]. Finally, the fact that phytoestrogens are often not selected for specific attributes, such as acting only via ERβ, may have confounded studies on health effects. Some promising results regarding amelioration of hot flushes with liquiritigenin, an ERβ-selective agonist from a Chinese herbal extract, have, however, resulted in Phase 2 clinical trials to evaluate safety and efficacy for the treatment of menopausal symptoms [82], [83].


#

Conclusions

The increased public and industry interest in phytoestrogens suggests that validated health claims would contribute significantly to adding value to products such as honeybush tea. Certain extracts of Cyclopia undoubtedly display estrogenic activity ([Table 4]), and many of the major and minor polyphenols found in Cyclopia certainly have been shown to have phytoestrogenic potential ([Table 3]), but whether this translates into firm health recommendations for a “cup-of-tea” of honeybush is debatable. Firstly, harvestings of Cyclopia differ significantly in terms of estrogenic activity and polyphenol content ([Table 5]), and secondly, Cyclopia extracts have not been tested for estrogenicity in vivo. The importance of evaluating the bioavailability as well as the metabolic transformation of active compounds, both by gut microflora and hepatic enzymes, has been stressed [31], [84]. Cyclopia extracts have been tested in vivo for absorption and metabolism [85], [86]; however, the focus was on mangiferin and hesperidin, both compounds without estrogenic activity ([Table 3]). The aglycone of hesperidin, hesperetin, which does display weak estrogenic activity, was, however, one of the metabolites detected in urine [85]. This suggests that glycosylated polyphenols, of which several constitute the major polyphenols in Cyclopia extracts ([Table 1]), would probably be transformed to the corresponding aglycone with higher phytoestrogenic activity. Finally, the concept of either synergistic or even antagonistic formulations consisting of intelligent mixtures of natural products to treat disease is gaining ground [47], [87], [88], [89], [90], [91] and thus, although we have focussed on the phytoestrogenicity of individual compounds found in Cyclopia, we should consider the possibility that it is the mixture of compounds found in Cyclopia extracts, rather than an individual compound, that confers the desired estrogenic activity.


#

Acknowledgements

The authors would like to thank the Medical Research Council (MRC) and the Cancer Association of South Africa (CANSA) for financial support to A. L. (grant for projects entitled “Cyclopia Phytoestrogens” and “Cyclopia and breast cancer”) and the Department of Science and Technology as well as the National Research Foundation (NRF) for financial support to E. J. (grant 70525). The views and opinions expressed are not those of the funding agencies but of the authors of the material produced or publicised.


#
#

Conflict of Interest

The authors declare no conflict of interest.


Correspondence

Dr. Ann Louw
Department of Biochemistry, University of Stellenbosch, Stellenbosch
Private Bag X1
Matieland (Stellenbosch) 7602
South Africa
Phone: +27 2 18 08 58 73   
Fax: +27 2 18 08 58 63   


Zoom Image
Fig. 1 Shoots of C. subternata (left) and C. genistoides (right) with distinctive yellow flowers having an indented calyx, characteristic of Cyclopia species. (Color figure available online only.)
Zoom Image
Fig. 2 Structures of major phenolic compounds of C. subternata and minor compounds with estrogenic activity present in the leaves and stems of some Cyclopia spp. (* indicates that the position or identity of the glycosyl moiety is not certain; bold text indicates the class of compound).
Zoom Image
Fig. 3 Steps in ER signalling used to evaluate estrogenicity. E = estrogenic compound, ER = estrogen receptor, ERE = estrogen response element. (1) Binding of an estrogenic ligand to the ER may be evaluated by ligand-binding assays, (2) binding of ligand-activated ER to an ERE in the promoter of an estrogen responsive gene may be evaluated by promoter-reporter studies using an ERE-containing promoter reporter or by measuring mRNA levels of select ER-responsive genes, and (3) downstream biological effects such as cell proliferation or hypertrophy of the uterus may be measured using the E-screen or uterotrophic assay, respectively.