Applicability of the OECD 455 in-vitro assay for determination of hERa agonistic activity of isoflavonoids

Toxicology and Applied Pharmacology 386 (2020) 114831

Contents lists available at ScienceDirect

Toxicology and Applied Pharmacology journal homepage: www.elsevier.com/locate/taap

Applicability of the OECD 455 in-vitro assay for determination of hERa agonistic activity of isoflavonoids

T

Darja Gramec Skledara, Václav Tvrdýb, Maša Kendaa, Anamarija Zegaa, Milan Pourc, ⁎ Pavel Horkýc, Přemysl Mladěnkab, Marija Sollner Dolenca, Lucija Peterlin Mašiča, a

Faculty of Pharmacy, University of Ljubljana, Aškerčeva 7, 1000 Ljubljana, Slovenia Department of Pharmacology and Toxicology, Faculty of Pharmacy, Charles University, Heyrovského 1203, 500 05 Hradec Králové, Czech Republic c Department of Organic and Bioorganic Chemistry, Faculty of Pharmacy, Charles University, Heyrovského 1203, 500 05 Hradec Králové, Czech Republic b

ARTICLE INFO

ABSTRACT

Keywords: Flavonoid Endocrine activity Estrogenic effect OECD TG 455

The Organisation for Economic Co-operation and Development (OECD)-validated transactivation assay using the human estrogen receptor alpha (hERα) Hela9903 cell line is used for activity evaluation of hERα agonists and antagonists. Due to many advantages, this assay is broadly used as an initial screening process. However, response significantly higher from that of 17-β estradiol (E2) was observed with phytoestrogens for concentrations commonly above 1 μM in previous studies. The main aim of this study was thus to ascertain the applicability of OECD protocol 455 for evaluation of estrogenic activity of natural flavonoids, including known phytoestrogens. The estrogenic activities of aglycones as well as of O-methylated and glycosylated flavonoids were evaluated. Supra-maximal luciferase activity was seen for most of the flavonoids tested at concentrations even below 1 μM. hERα-mediated luciferase expression was confirmed with the competition assay specified in OECD protocol 455. However, at concentrations above 1 μM, non-specific interactions were also observed. Instead of EC50 values, which could not be determined for most of the isoflavonoids tested, the concentrations corresponding to 10% (PC10) and 50% (PC50) of the maximum activity of the positive control, E2, were used for quantitative determination of estrogenic activities. Appropriate evaluation of the data obtained with the current OECD protocol 455 validated assay represents a valuable tool for initial screening of natural flavonoids for estrogenic activity.

1. Introduction Endocrine disrupting chemicals (EDCs) are defined as exogenous substances or mixtures that alter the function(s) of the endocrine system, and consequently can cause adverse health effects in the intact organism or its progeny, or in (sub)populations (Damstra et al., 2002). EDC occurrence is widespread, and numerous biomonitoring studies have confirmed exposure to low doses of different EDCs (Vandenberg et al., 2010; Asimakopoulos et al., 2014). EDCs are usually synthetic compounds that can be found in various everyday products, like plastic (e.g., bisphenol A, diethyl phthalate), cosmetics (e.g., benzylparabene, triclosan), and electronic devices (e.g., brominated flame retardants) (Golden et al., 2005; Gramec Skledar and Peterlin Mašič, 2016; Gramec Skledar et al., 2016; Skledar et al., 2016; Klopcic and Dolenc, 2017). However, endocrine activities have also been reported for natural compounds, like urolithins, lignans, and isoflavonoids; these show

mainly estrogenic activities, and are therefore known as phytoestrogens (Larrosa et al., 2006; Nordeen et al., 2013; Gramec Skledar et al., 2018). General concerns regarding EDCs have resulted in the development of numerous assays for their identification. Due to the ‘3Rs’ principles (i.e., replacement, reduction, refinement) and the restrictions of in-vivo testing in the cosmetics industry, efforts around the globe have been focused in the development of reliable in-vitro assays. The advantages of such in-vitro assays are their rapidity and relatively low cost. Moreover, they can provide information about the mechanisms of endocrine disruption. Nevertheless, only two in-vitro assays for estrogenic activity have been validated by the Organisation for Economic Co-operation and Development (OECD) and are included in Tier 2 of endocrine disruptor testing and assessment: the estrogen receptor (ER) binding assay (OECD TG 493), and the ER transactivation assay using the hERα-HeLa9903 cell line (OECD TG 455). hERα-HeLa-9903 cell line was derived from human cervical cancer cells and has two stably integrated

Abbreviations: E2, 17-β-estradiol; EDC, endocrine-disrupting chemicals; ER, estrogen receptor; OECD, Organisation for Economic Co-operation and Development; OHT, 4-hydroxytamoxifene; TLC, thin layer chromatography ⁎ Corresponding author. E-mail address: [email protected] (L. Peterlin Mašič). https://doi.org/10.1016/j.taap.2019.114831 Received 30 May 2019; Received in revised form 30 September 2019; Accepted 16 November 2019 Available online 20 November 2019 0041-008X/ © 2019 Elsevier Inc. All rights reserved.

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

constructs: the hERα expression construct, which encodes the full length ERα and luciferase reporter construct with estrogen responsive element. Binding of an ER agonist to ERα consequently increase expression of firefly luciferase reporter gene which resulted in higher cellular content of luciferase enzyme. The ER transactivation assay using the hERα-HeLa-9903 cell line is a reliable in-vitro assay for determination of hERα agonists and antagonists. Despite numerous advantages of the OECD validated hERαHela9903 assay for screening estrogenic agonist and antagonist activities, the main weakness of the assay remains its low sensitivity, which can be explained with non-receptor mediated mechanisms. In the invitro luciferase based reporter gene bioassays for measuring estrogenic agonist activity, the maximal response of several natural compounds

including isoflavonoids was significantly higher than that of E2 (Sotoca et al., 2010). This so called “superinduction” is not cell specific and it has already been reported for various isoflavonoids at concentrations higher than 1 μM in different cell lines, like Hela (Escande et al., 2006), Hek293 (Kuiper et al., 1998), MCF-7 (Joung et al., 2003), MVLN (Freyberger and Schmuck, 2005), CHO (Takeuchi et al., 2009) and U2OS (Sotoca et al., 2008). Superinduction was observed only in luciferase reporter gene assays and not with other reporter genes like βgalactosidase (Sotoca et al., 2010). Sotoca and co-workers showed that superinduction can be better explained with stabilization of the luciferase enzyme than with increased luciferase gene expression and is therefore probably devoid of biological relevance (Sotoca et al., 2010). Special care is needed when using luciferase reporter assays to

Table 1 Previous reports of estrogenic activities of isoflavonoids. Compound

Assay

Nonmethylated isoflavonoids Genistein HGELN-Luc MVLN-Luc E-screen MCF-7-β gal Ligand binding assay MCF-7-E10 Luc E-screen HELN-Luc Daidzein

Alkaline phosphatase activity in Ishikawa cells MCF7:D5L Luc Hek ERβ Luc MCF-7-β gal Ligand binding assay MCF-7-E10 Luc E-screen HELN-Luc

Methylated isoflavonoids Calycosin Competitive radiometric binding assay Cladrin Formononetin

Glycitein

Prunetin Biochanin A

Glycosides Ononin Puerarin Glycitin

Expression of estrogen –responsive genes in MCF-7 Huh7-Luc Huh7-Luc MCF-7-Luc E-screen Ligand binding assay Alkaline phosphatase activity in Ishikawa cells MCF7:D5L Luc Hek ERβ Luc E-screen Huh7-Luc Alkaline phosphatase activity in Ishikawa cells Binding assay β-galactosidase reporter gene assay in yeast HELN-Luc

/ E-screen Binding assay β-galactosidase reporter gene assay in yeast E-screen

Colonic metabolites of isoflavonoids S-Equol MCF-7-E10 Luc E-screen R,S-Equol Binding assay Desmethylangolensin

E-screen

Agonistic activity

Ref.

EC50 = 0.05 μM EC50 = 0.038 μM EC50 = 0.04 μM EC50 (ERα) = 0.33 μM EC50 (ERβ) = 0.013 μM RBA (ERα) = 4 RBA (ERβ) = 87 PC50 = 0.78 μM Stimulation of MCF-7 cell proliferation at > 0.015 μM EC50 (ERα) = 0.038 μM EC50 (ERβ) = 0.006 μM EC25 = 0.45 μM EC25 = 0.31 μM EC25 = 0.24 μM EC50 (ERα) = 0.63 μM EC50 (ERβ) = 0.016 μM RBA (ERα) = 0.1 RBA (ERβ) = 0.5 PC50 = 2.20 μM Stimulation of MCF-7 cell proliferation at > 0.033 μM EC50 (ERα) = 0.150 μM EC50 (ERβ) = 0.057 μM

(Gutendorf and Westendorf, 2001) (Gutendorf and Westendorf, 2001) (Gutendorf and Westendorf, 2001) (Chrzan and Bradford, 2007)

RBA (ERα) = 0.01 RBA (ERβ) = 0.027 Increased progesterone receptor and GREB1 gene expression NA NA Increased ER-mediated luciferase expression Stimulation of MCF-7 cell proliferation RBA < 0.01 EC25 = 10 μM EC25 = 5.19 μM EC25 = 7.33 μM Stimulation of MCF-7 cell proliferation at > 0.36 μM NA NA Bind to ERβ, but not to ERα Weak ERβ-dependent β-galactosidase induction EC50 (ERα) = 0.082 μM EC50 (ERβ) = 0.007 μM

(Boonmuen et al., 2016)

(Kuiper et al., 1998) (Onoda et al., 2011) (Onoda et al., 2011) (Escande et al., 2006) (Tchoumtchoua et al., 2016) (Tchoumtchoua et al., 2016) (Tchoumtchoua et al., 2016) (Chrzan and Bradford, 2007) (Kuiper et al., 1998) (Onoda et al., 2011) (Onoda et al., 2011) (Escande et al., 2006)

(Boonmuen et al., 2016) (Gautam et al., 2011) (Gautam et al., 2011) (Ji et al., 2006) (Ji et al., 2006) (Kuiper et al., 1998) (Tchoumtchoua et al., 2016) (Tchoumtchoua et al., 2016) (Tchoumtchoua et al., 2016) (Onoda et al., 2011) (Khan et al., 2015) (Khan et al., 2015) (Morito et al., 2002) (Morito et al., 2002) (Escande et al., 2006)

/ NA Poor binding to ERβ, but not to ERα NA NA

/ (Michihara et al., 2012) (Morito et al., 2001) (Morito et al., 2001) (Morito et al., 2001)

PC50 = 0.47 μM Stimulation of MCF-7 cell proliferation at > 0.006 μM IC50 (ERα) = 1.5 μM IC50 (ERβ) = 0.2 μM Stimulation of MCF-7 cell proliferation

(Onoda et al., 2011) (Onoda et al., 2011) (Mueller et al., 2004) (Kinjo et al., 2004)

RBA, relative binding affinity (E2 as 100%); PC50, concentration corresponding to 50% maximum activity of positive control; NA, not active. 2

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

Table 2 Chemical structures of the isoflavonoids tested.

determine estrogenic activities of (iso)flavonoids at concentrations higher than 1 μM. To confirm receptor mediated effects, the assay with the ER antagonist 4-hydroxytamoxifen (OHT) should be performed as described in the OECD guidelines (OECD, 2016). Flavonoids are a broad class of polyphenolic plant metabolites that have a benzo-γ-pyrone structure. In plants, flavonoids occur mostly in their glycosylated or methylated forms, which are poorly absorbed and are hydrolyzed in the gut by the intestinal microbiota (e.g., lactic acid bacteria, bifidobacterial) to form the bioavailable and bioactive aglycones (Gaya et al., 2017). As examples here, the isoflavonoids biochanin A and puerarin are the methylated and C-glycosylated precursors of daidzein, while formononetin and genistin are the methylated and O-glycosylated precursors of genistein (Gaya et al., 2017). Flavonoids have a wide range of biological activities, which include direct and indirect antioxidant, anti-inflammatory, and antiproliferative activities, and they can thus be beneficial for human health (Mladenka et al., 2010). However, some of the published data need to be used with caution because flavonoids have structural features that can generate false positives in biochemical assays (Sassano et al., 2013; Tritsch et al., 2015; Baell, 2016; Bisson et al., 2016). The structural similarity to E2 results in interactions of isoflavonoids with the ER, and therefore they are commonly used as food supplements for prevention of menopause symptoms (Chen et al., 2015). However, interactions with endocrine pathways can pose risks to human health. Indeed, long-term consumption of food supplements that contain isoflavonoids can lead to side effects, like increased breast cancer risk (EFSA, 2015).

Estrogenic activities of many isoflavonoids and their colonic absorbable metabolites have been determined previously (Table 1). Comprehensive studies that describe estrogenic activities of the soy flavonoids daidzein and genistein have been published (Table 1), although there are limited data available that describe the estrogenic activities of several methylated and glycosylated flavonoids and their above-mentioned metabolites. A study performed by Mueller et al. (2004) showed different affinities of several isoflavonoids toward ERα and ERβ (Mueller et al., 2004). Genistein showed significantly higher binding affinity to ERβ than ERα, and elimination of one (e.g., daidzein, biochanin A) or two (e.g., formononetin) hydroxyl groups significantly decreased its binding affinities toward both of these ERs (Kuiper et al., 1998). Morito et al. (2001) showed that binding of glycosylated flavonoids to ERs is weaker in comparison with their corresponding aglycones. Some selected isoflavonoids showed no estrogenic activities in previous studies, such as for the methylated forms cladrin and prunetin, and the glycosides puerarin and glycitin (Table 1). To the best of our knowledge, no studies have described the estrogenic activity of ononin. In the present study, we evaluated the applicability of the OECD 455 hERα-Hela9903 transactivation assay for evaluation of estrogenic activities of natural flavonoids, and we propose a flowchart for evaluation of natural isoflavonoids with the OECD 455 method. The in-vitro data obtained were further compared with in-silico predictions using the Endocrine Disruptome docking program. Evaluations were performed for a large number of natural isoflavonoids, as both aglycones and methylated and glycosylated isoflavonoids, and their two colonic 3

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

Fig. 1. Estrogenic activities of the different groups of isoflavonoids and their metabolites, determined using the transactivation assay with the hERα-Hela9903 cell line.

metabolites (Table 2). Estrogenic activities were determined for isoflavonoids with known and firmly established estrogenic activities, such as genistein and daidzein, and these data are compared with data from previous studies. Additionally, the present study was extended to isoflavonoids that have not been evaluated previously and to those for which reliable data are lacking. Finally, the detailed procedure for evaluation of phytoestrogens with the current assay is provided.

and kanamycin solution from Streptomyces kanamyceticus (10 mg/mL) were from Sigma-Aldrich (Germany). Calycosin (99%; CAS 25389-94-0) and cladrin (98%; CAS 24160-14-3) were from Phytolab (Vestenbergsgreuth, Germany), and R,S-equol (98%; CAS 531-95-3) and S-equol (97%; CAS 531-95-3) were from Toronto Research Chemicals (Toronto, Canada). Biochanin A (≥99%; CAS 491-80-5), formononetin (≥99%; CAS 485-72-3), glycitein (≥95%; CAS 4095783-3), glycitin (≥95%; CAS 40246-10-4), ononin (≥99%; CAS 486-624), prunetin (≥95%; CAS 552–-59-0), and puerarin (≥99%; CAS 368199-0) were from Extrasynthese (Lyon, France), and daidzein (≥98%; CAS 486-66-8) and genistein (≥98%; CAS 446-72-0) were from Sigma Aldrich (Prague, Czech Republic). The purities of the isoflavonoids were checked using high performance liquid chromatography, with the exception of R,S-equol and S-equol, where the purities were determined using thin-layer chromatography (TLC). O-desmethylangolensin (CAS

2. Material and methods 2.1. Materials 17-β-Estradiol (E2; ≥98%; CAS 50-28-2), 17-α-estradiol (≥98%; CAS 57-91-0), corticosterone (≥98.5%; CAS 50-22-6), dimethyl sulfoxide, 4-hydroxytamoxifen (OHT; ≥98%; Z isomer; CAS 68047-06-3), 4

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

21255-69-6) was synthesized in the Department of Organic and Bioorganic Chemistry, Faculty of Pharmacy in Hradec Králové, Charles University, Czech Republic, as described below.

100 mL) and saturated NaHCO3 (3× 100 mL), dried over anhydrous MgSO4, and concentrated in vacuo. The product was purified by column chromatography using hexane:ethyl acetate (7:3, v/v) as the mobile phase. The product obtained was a white amorphous solid, at 80% yield. The spectral data were in agreement with those in the literature (Goto et al., 2009). NMR analysis: 1H NMR (500 MHz, CD3OD) δ 7.78–7.76 (1H, m), 7.17–7.05 (2H, m), 6.79–6.68 (2H, m), 6.27–6.26 (1H, m), 6.23–6.22 (1H, m), 4.65 (1H, q, J = 6.8 Hz), 1.42 (3H, d, J = 6.8 Hz). 13C NMR (125 MHz, CD3OD) δ 206.64, 167.02, 166.05, 157.36, 134.24, 134.12, 129.60, 116.62, 113.26, 108.88, 103.71, 46.86, 19.61. Anal. Calcd for C15H14O4: C, 69.76; H, 4.46; found: C, 69.80; H, 5.47. HRMS (TOFESI+) m/z calcd for C15H14O4+ 258,0900; found 258.0890.

2.2. Chemical synthesis of O-desmethylangolensin All reagents were from Sigma–Aldrich and were used without further purification. Solvents were distilled prior to use (e.g., tetrahydrofuran, dichloromethane). TLC analyses were performed using silica gel TLC plates (60 F254; Merck), and the bands visualized by UV in combination with staining. Column chromatography was carried out on silica gel 60 (0.040–0.063 mm; Merck). 1H and 13C NMR spectra were recorded using Varian Mercury VxBB 300 or VNMR S500 instruments. The chemical shifts were reported relative to tetramethylsilane, and referenced to the residual solvent peaks. Infrared spectra were recorded on a Nicolet 6700 FT-IR spectrophotometer equipped with an attenuated total reflectance device. Mass spectrometry (MS) data were measured on an Agilent Tech 500 Iontrap spectrometer. High-resolution MS data were recorded on a Q-TOF mass spectrometer using electrospray ionization mode. Boron trifluoride diethyl etherate (BF3•Et2O; 0.5 mL) was added to a mixture of resorcinol (100 mg) and 2-(4-hydroxyphenyl)propanoic acid (158 mg), and the resultant solution was heated to 120 °C for 10 min. The mixture was cooled to room temperature, and cold water was added (20 mL). The product was extracted with Et2O (3× 50 mL). The combined Et2O layers were successively washed with brine (3×

2.3. Predicting flavonoid binding affinities for hERα The Endocrine Disruptome docking program was used for virtual screening of the isoflavonoids tested, according to the docking procedures described by Kolsek and coworkers (Kolšek et al., 2014). This software tool predicts the binding affinities between tested compounds and the receptor binding ligand domains of ERα and divides compounds in four classes, based on probability of binding to ERα. Every compound was docked at least twice. In cases where the docking results obtained were inconsistent (i.e. when the same compound is classified in two different classes), additional docking was applied and the results of two consistent runs out of three were included in the table.

Fig. 2. Competition assays without and with 10 μM 4-hydroxytamoxifen (OHT) and calculated difference between the nontreated wells and the wells treated with 10 μM OHT. 5

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

2.4. Stably transfected transactivation assays using the hERα-HeLa-9903 cell line

3.2. Estrogenic activities of selected isoflavonoids The estrogenic activities of the selected isoflavonoids were determined using the hERα-mediated transactivation assay with the Hela9903 cell line, following the OECD 455 guideline. First, the relevance and reliability of the selected in-vitro assay were determined using the following control compounds: E2 (strong estrogenic activity); 17-α-estradiol (weak estrogenic activity); and corticosterone (inactive). Their EC50 values were 0.032 nM for E2 and 4.87 nM for 17-α-estradiol, while corticosterone was fully inactive. With the exception of puerarin, all of the tested isoflavonoids showed dose-dependent agonistic estrogenic activities, with PC50 in the nanomolar to low micromolar range (Fig. 1, Table 4). For most of the isoflavonoids, supra-maximal luciferase activity was even seen at < 1 μM (Fig. 1). The responses to the flavonoids formononetin, ononin, and prunetin at their highest tested concentration (25 μM) were > 6-fold higher when compared to the responses to 1 nM E2. For EC50 determinations, full concentration-response curves are required. However, due to this supra-maximal luciferase activity, sigmoid dose-response curves could not be constructed for most of the flavonoids tested. Therefore, their estrogenic activities were characterized as their PC10 and PC50 values; i.e., the concentrations of these

For determination of the estrogenic activities, the Hela 9903 cell line was used (JCRB1318; Japanese Collection of Research Bioresources Cell Bank; Osaka, Japan). The estrogenic activities were evaluated according to OECD protocol 455, with small modifications. Briefly, the cells were maintained in phenol-red-free minimal essential medium (Gibco), supplemented with 10% charcoal-stripped fetal bovine serum (Sigma), 2 mM glutamine (Sigma), and 60 mg/L kanamycin (Sigma), at 37 °C and under 5% CO2. The cells were seeded on white 96-well luminometer plates (Greiner, Bio One) at 3 × 105 cells/mL, and after 3 h they were treated with the selected compounds, and also with the vehicle control (0.1% dimethylsulfoxide) and the positive control (1 nM E2), and incubated for 24 h. Cell viability was determined using resazurin (Sigma) reduction assays, and the luciferase activity was carried out using the One-glo Luciferase assay system (Promega, Madison, WI, USA), according to the manufacturer instructions (Luciferase assay system, instructions for use, 12/11; Promega). Fluorescence and luminescence were measured with an automatic microplate reader (Synergy 4 Hybrid Microplate Reader; BioTek, Winooski, VT, USA). Competition assays were performed as specified in OECD protocol 455. Briefly, serial dilutions of the tested compounds were prepared without and with addition of 10 μM OHT. After 24 h, the luciferase activities were determined (One-glo Luciferase assay system; Promega, Madison, WI, USA).

Table 3 Predicted affinities toward ERα of the isoflavonoids tested.

Isoflavanoid

Free binding energy

/metabolite

(kcal/mol)

2.5. Calculations and statistics

ERα

The data were analyzed according to the OECD protocol 455 (OCDE, 2016). GraphPad Prism 5.04 software for Windows (GraphPad Software Inc., San Diego, CA, USA) and Microsoft Excel were used to analyze the data obtained. Data collected from at least three independent experiments with triplicate wells for each experiment were fit using the GraphPad Prism software, and the representative values were calculated, as: PC10, as concentration of the test compound that corresponds to 10% of the maximum activity of the positive control (1 nM E2); PC50, as PC10, but corresponding to 50% of the maximum activity of the positive control (1 nM E2); maximum level of response induced by the test compound, expressed as percentage of the response induced by 1 nM E2 on the same plate; and EC50, as half maximal effective concentration, where possible. The flavonoids were considered as agonists at ER if their maximal response was at least 10% of the maximal response induced by the positive control (1 nM E2). The activities toward hERα were analyzed using two-sample student's t-tests, where *p < .05, **p < .01, and ***p < .001 were considered as statistically significant. 3. Results 3.1. In-silico prediction with endocrine disruptome The Endocrine Disruptome docking program was used for the binding predictions of the selected isoflavonoids to ERα (Kolšek et al., 2014). As indicated in Table 3, these data were divided into four classes that are indicated by the different colors: red, high probability; orange and yellow, intermediate probability; and green, low probability, of binding to the selected receptor (Table 3). The numbers provided by the program represent the free binding energies of the tested compounds. As can be seen from Table 3, the soy isoflavones daidzein and genistein showed the highest probabilities of binding to ERα. Desmethylangolesin, S-equol, and glycitein showed intermediate probabilities of binding, whereas for other flavonoids, the predicted binding probabilities from Endocrine Disruptome were low.

Desmethylangolesin

-8.3

S-equol

-8.8

Biochanin A

-6.5

Calycosin

-6.5

Cladrin

-6.4

Daidzein

-9.5

Formononetin

-6.8

Genistein

-9.3

Glycitin

-6.3

Glycitein

-8.1

Ononin

-5.0

Prunetin

-6.8

Puerarin

-6.5

Data were obtained with the Endocrine Disruptome docking program (Kolšek et al., 2014). Probabilities of binding to ERα: red, high (free binding energy below −9.3 kcal/mol); orange and yellow, intermediate (free binding energies from −8.8 kcal/mol to −9.3 kcal/mol and from −8.2 kcal/mol to −8.8 kcal/mol, respectively); green, low (free binding energy above −8.2 kcal/mol).

6

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

compounds that corresponded to 10% and 50% of the maximum activity of 1 nM E2 (Table 4).

also in in-vivo studies. Namely, daidzein exhibited very low uterotropic activity in mice (Farmakalidis et al., 1985) and rat (Diel et al., 2000), but significantly modulated the expression of estrogen-sensitive genes in rat uterus (Diel et al., 2000). In a study of Jefferson et al. (2002) genistein increased uterine weight in mouse uterotropic assay, but was 50,000 times less potent than diethylstilbestrol (Jefferson et al., 2002). Daidzein and biochanin A showed no activity on uterine weight in that same assay, however they both influenced some specific estrogen dependent morfological and biochemical parameters (e.g. uterine epithelial cell height, uterine gland number,..) (Jefferson et al., 2002). Then the evaluation was expanded to a large number of natural isoflavonoids and their metabolites. The data collected here showed concentration-dependent estrogenic activities of the isoflavonoids tested at nanomolar to low micromolar concentrations. These concentrations can be reached relatively easily for aglycones after consumption of flavonoid reach food, or of the food supplements that are widely used as an alternative to hormone-replacement therapy. The daidzein metabolite formed by the intestinal microbiota, S-equol, and the synthetic racemic R/S-equol showed comparable estrogenic activities to genistein in the present assays, with PC50 values of 0.14 μM and 0.12 μM, respectively. The estrogenic activities of the methylated and glycosylated flavonoids were lower, however, with PC50 values still in the low micromolar range. These data are in agreement with previous in-vitro studies. Morito et al. (2002) showed that methylation and glycosylation can lower estrogenic activities (Morito et al., 2002). Such lower activity for the methylated and glycosylated compounds were further predicted in silico, in comparison with the aglycones. Using the Endocrine Disruptome program, the highest binding affinities were predicted for the nonmethylated aglycones genistein and daidzein, and for the isoflavandiol S-equol, while the predicted binding affinities for the methylated and glycosylated isoflavonoids were low. In-silico predictions are therefore in agreement with the in-vitro data obtained. Nevertheless, the glycosides ononin and glycitin showed estrogenic activities in the present study, with PC50 values of 0.57 μM and 16.29 μM, respectively. This is in contrast with data from Morito et al. (2001), who showed no estrogenic activity for glycitin in a yeast βgalactosidase reporter gene assay, as well as in an MCF-7 proliferation assay (Morito et al., 2001). The glycoside puerarin showed no estrogenic activity in the present assay, as also in a previously conducted

3.3. Determination of nonreceptor-mediated luminiscence signals To determine whether the luciferase activities were due to specific interactions of the tested flavonoids with ERα, competition assays with the ER antagonist OHT were performed, as specified in OECD protocol 455. Addition of 10 μM OHT significantly lowered the estrogenic activities of the tested flavonoids (Fig. 2). However, the agonistic activities were not completely inhibited. At > 1 μM, some induction of the luciferase reporter gene was seen, which indicated that there were nonspecific interactions (i.e., nonreceptor-mediated interactions). The true responses can therefore be calculated as the differences between the nontreated wells and the wells treated with 10 μM OHT. 4. Discussion Despite more than two decades of extensive research into EDCs, these still raise concerns and pose challenges to both researchers and regulators. An array of different in-vitro assays for determination of estrogenic activities has been developed over the last 20 years, which include reporter gene assays, competitive binding assays, and proliferation assays. In the present study, the OECD-validated transactivation assay using the hERα-Hela-9903 cell line was used for evaluation of the estrogenic activities of various natural isoflavonoids and their metabolites. The current OECD validated assay is relevant and reliable, and thus it represents an important tool for initial screening for estrogenic activity. However, for some compounds, such as the phytoestrogens, supra-maximal luciferase activity occurs at > 1 μM, and therefore these compounds demand special attention. Estrogenic activities for several isoflavonoids have been reported previously. First, to confirm reproducibility of the hERα-Hela9903 transactivation assay, the hERα agonist activities obtained for some firmly established natural estrogens were compared with previously published data, such as for genistein and daidzein. The estrogenic activities of genistein and daidzein defined according to their PC50 values were 0.10 μM and 0.38 μM, respectively, which are in agreement with previous in-vitro studies (Table 1). In-vitro observations were confirmed

Table 4 Summary of the estrogenic activities of the different groups of flavonoids according to the transactivation assay using the hERα-Hela9903 cell line presented as means ± SD of at least three biological replicates. Compound

PC10 [μM]

PC50 [μM]

Max effect as % relative to E2-max

EC50 [μM]

Flavonoids O-methylated on B-ring Biochanin–A Calycosin Cladrin Formononetin

0.11 0.13 0.92 0.12

0.61 ± 0.42 2.10 ± 0.77 15.81 ± 0.69 0.44 ± 0.08

468.1 ± 96.2 278.8 ± 60.5 63.8 ± 0.8204 660.3 ± 137.3

NDa NDa 2.20 NDa

Non-methylated flavonoids Daidzein Genistein

0.13 ± 0.02 0.02 ± 0.003

0.38 ± 0.07 0.10 ± 0.02

427.0 ± 3.7 588.8 ± 70.8

NDa NDa

Flavonoids O-methylated on A-ring Glycitein Prunetin

5.88 ± 0.28 0.12 ± 0.01

20.09 ± 1.28 0.34 ± 0.11

75.5 ± 14.6 610.4 ± 164.2

12.42 NDa

Glycosides Ononin Glycitin Puerarin

0.12 ± 0.02 5.43 ± 0.43 NA

0.57 ± 0.33 16.29 ± 3.27 NA

650.9 ± 149.2 158.4 ± 9.4 NA

NDa NDa NA

Isoflavandiols S-Equol R,S-Equol

0.01 ± 0.003 0.01 ± 0.0006

0.14 ± 0.03 0.12 ± 0.001

175.3 ± 30.3 297.5 ± 43.9

0.41 NDa

Others Desmethylangolensin

0.21 ± 0.06

1.18 ± 0.05

105.2 ± 7.4

1.74

± ± ± ±

0.003 0.04 0.26 0.01

NA: not active; ND, not determined. a Due to overexpression of luciferase gene, sigmoid dose-response curves could not be constructed, and EC50 values could not be determined. 7

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

although at higher concentrations (> 1 μM), nonspecific interactions were also seen. This can be explained with stabilization of the firefly luciferase enzyme by tested isoflavonoids, as it was described previously by Sotoca et al. (2010) and thus have likely no biological relevance (Sotoca et al., 2010). The true estrogenic activity that is mediated through hERα should be calculated as the differences between OHT untreated and treated samples. Nevertheless, the higher transcriptional activity of the isoflavonoids compared to E2 remains even after calculation of the difference between OHT treated and untreated samples. The detailed mechanism for this observation is still not clear, but might be due to stimulated receptor or due to cofactor renewal as it was proposed previously by Legler and coworkers (Legler et al., 1999). However this topic still needs further clarification. Due to incomplete concentration-response curves, EC50 values could not be provided in cases of supra-maximal luciferase activity, and the estrogenic activities were described according to the PC10 and PC50 values, as demonstrated in Fig. 3. We have demonstrated the usefulness of the current assay for evaluation of the estrogenic activities of flavonoids, and have provided the detailed protocol for determination of their estrogenic activities with the transactivation assays using the Hela9903 cell line (Fig. 4).

Fig. 3. Definition of the PC10 and PC50 values for genistein.

5. Conclusions

MCF-7 cell proliferation assay (Michihara et al., 2012). Despite estrogenic activities being shown for the glycosides ononin and glycitin, these data are not likely to be relevant for humans due to the limited bioavailability of those compounds. Above 1 μM, the effects observed for most of the tested isoflavonoids were higher than those for the positive control (1 nM E2) (Fig. 2). This supra-maximal luciferase activity by the isoflavonoids genistein, daidzein, and biochanin-A has been described previously in different cell lines (Kuiper et al., 1998; Escande et al., 2006). This observation requires special attention, as it might indicate non hERα-mediated induction of luciferase expression. Additional assays with the ER antagonist (i.e., 10 μM OHT) were therefore conducted, as is specified in OECD protocol 455 for unambiguous confirmation of ER-mediated estrogenic activity. OHT significantly lowered these estrogenic activities,

Various in-vitro assays have been developed for the determination of estrogenic activities of natural compounds, among which the transactivation assay using the hERα-Hela9903 cell line is also validated by OECD. Due to the reported supra-maximal increases in luminescence signals observed with the phytoestrogens > 1 μM, questions arise whether this method is suitable for determination of estrogenic activities of the phytoestrogens. We have clearly demonstrated here that the transactivation assay using the hERα-Hela9903 cell line is a valuable assay for reliable determination of estrogenic activities of flavonoids, although it demands special considerations. The data obtained must be interpreted with caution, and different approaches should be applied for compounds where supra-maximal luciferase activity is seen. The detailed protocol that can be followed for the evaluation of estrogenic

Fig. 4. Proposed protocol for evaluation of the estrogenic activities of phytoestrogens with the transactivation assay using the hERα-Hela9903 cell line. 8

Toxicology and Applied Pharmacology 386 (2020) 114831

D. Gramec Skledar, et al.

activities of phytoestrogens is attached.

179–189. Ji, Z.N., Zhao, W.Y., Liao, G.R., Choi, R.C., Lo, C.K., Dong, T.T., Tsim, K.W., 2006. In vitro estrogenic activity of formononetin by two bioassay systems. Gynecol. Endocrinol. 22, 578–584. Joung, K.E., Kim, Y.W., Sheen, Y.Y., 2003. Assessment of the estrogenicity of isoflavonoids, using MCF-7-ERE-Luc cells. Arch. Pharm. Res. 26, 756–762. Khan, K., Pal, S., Yadav, M., Maurya, R., Trivedi, A.K., Sanyal, S., Chattopadhyay, N., 2015. Prunetin signals via G-protein-coupled receptor, GPR30(GPER1): stimulation of adenylyl cyclase and cAMP-mediated activation of MAPK signaling induces Runx2 expression in osteoblasts to promote bone regeneration. J. Nutr. Biochem. 26, 1491–1501. Kinjo, J., Tsuchihashi, R., Morito, K., Hirose, T., Aomori, T., Nagao, T., Okabe, H., Nohara, T., Masamune, Y., 2004. Interactions of phytoestrogens with estrogen receptors alpha and beta (III). Estrogenic activities of soy isoflavone aglycones and their metabolites isolated from human urine. Biol. Pharm. Bull. 27, 185–188. Klopcic, I., Dolenc, M.S., 2017. Endocrine activity of AVB, 2MR, BHA. and Their Mixtures. Toxicol Sci 156, 240–251. Kolšek, K., Mavri, J., Sollner Dolenc, M., Gobec, S., Turk, S., 2014. Endocrine disruptome– an open source prediction tool for assessing endocrine disruption potential through nuclear receptor binding. J. Chem. Inf. Model. 54, 1254–1267. Kuiper, G.G., Lemmen, J.G., Carlsson, B., Corton, J.C., Safe, S.H., van der Saag, P.T., van der Burg, B., Gustafsson, J.A., 1998. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 139, 4252–4263. Larrosa, M., González-Sarrías, A., García-Conesa, M.T., Tomás-Barberán, F.A., Espín, J.C., 2006. Urolithins, ellagic acid-derived metabolites produced by human colonic microflora, exhibit estrogenic and antiestrogenic activities. J. Agric. Food Chem. 54, 1611–1620. Legler, J., van den Brink, C.E., Brouwer, A., Murk, A.J., van der Saag, P.T., Vethaak, A.D., van der Burg, B., 1999. Development of a stably transfected estrogen receptormediated luciferase reporter gene assay in the human T47D breast cancer cell line. Toxicol. Sci. 48, 55–66. Michihara, S., Tanaka, T., Uzawa, Y., Moriyama, T., Kawamura, Y., 2012. Puerarin exerted anti-osteoporotic action independent of estrogen receptor-mediated pathway. J. Nutr. Sci. Vitaminol. (Tokyo) 58, 202–209. Mladenka, P., Zatloukalová, L., Filipský, T., Hrdina, R., 2010. Cardiovascular effects of flavonoids are not caused only by direct antioxidant activity. Free Radic. Biol. Med. 49, 963–975. Morito, K., Hirose, T., Kinjo, J., Hirakawa, T., Okawa, M., Nohara, T., Ogawa, S., Inoue, S., Muramatsu, M., Masamune, Y., 2001. Interaction of phytoestrogens with estrogen receptors alpha and beta. Biol. Pharm. Bull. 24, 351–356. Morito, K., Aomori, T., Hirose, T., Kinjo, J., Hasegawa, J., Ogawa, S., Inoue, S., Muramatsu, M., Masamune, Y., 2002. Interaction of phytoestrogens with estrogen receptors alpha and beta (II). Biol. Pharm. Bull. 25, 48–52. Mueller, S.O., Simon, S., Chae, K., Metzler, M., Korach, K.S., 2004. Phytoestrogens and their human metabolites show distinct agonistic and antagonistic properties on estrogen receptor alpha (ERalpha) and ERbeta in human cells. Toxicol. Sci. 80, 14–25. Nordeen, S.K., Bona, B.J., Jones, D.N., Lambert, J.R., Jackson, T.A., 2013. Endocrine disrupting activities of the flavonoid nutraceuticals luteolin and quercetin. Horm Cancer 4, 293–300. OECD, 2016. Test No. 455: Performance-Based Test Guideline for Stably Transfected Transactivation In Vitro Assays to Detect Estrogen Receptor Agonists and Antagonists. Onoda, A., Ueno, T., Uchiyama, S., Hayashi, S., Kato, K., Wake, N., 2011. Effects of Sequol and natural S-equol supplement (SE5-OH) on the growth of MCF-7 in vitro and as tumors implanted into ovariectomized athymic mice. Food Chem. Toxicol. 49, 2279–2284. Sassano, M.F., Doak, A.K., Roth, B.L., Shoichet, B.K., 2013. Colloidal aggregation causes inhibition of G protein-coupled receptors. J. Med. Chem. 56, 2406–2414. Skledar, D.G., Schmidt, J., Fic, A., Klopcic, I., Trontelj, J., Dolenc, M.S., Finel, M., Masic, L.P., 2016. Influence of metabolism on endocrine activities of bisphenol S. Chemosphere 157, 152–159. Sotoca, A.M., Ratman, D., van der Saag, P., Ström, A., Gustafsson, J.A., Vervoort, J., Rietjens, I.M., Murk, A.J., 2008. Phytoestrogen-mediated inhibition of proliferation of the human T47D breast cancer cells depends on the ERalpha/ERbeta ratio. J. Steroid Biochem. Mol. Biol. 112, 171–178. Sotoca, A.M., Bovee, T.F., Brand, W., Velikova, N., Boeren, S., Murk, A.J., Vervoort, J., Rietjens, I.M., 2010. Superinduction of estrogen receptor mediated gene expression in luciferase based reporter gene assays is mediated by a post-transcriptional mechanism. J. Steroid Biochem. Mol. Biol. 122, 204–211. Takeuchi, S., Takahashi, T., Sawada, Y., Iida, M., Matsuda, T., Kojima, H., 2009. Comparative study on the nuclear hormone receptor activity of various phytochemicals and their metabolites by reporter gene assays using Chinese hamster ovary cells. Biol. Pharm. Bull. 32, 195–202. Tchoumtchoua, J., Makropoulou, M., Ateba, S.B., Boulaka, A., Halabalaki, M., Lambrinidis, G., Meligova, A.K., Mbanya, J.C., Mikros, E., Skaltsounis, A.L., Mitsiou, D.J., Njamen, D., Alexis, M.N., 2016. Estrogenic activity of isoflavonoids from the stem bark of the tropical tree Amphimas pterocarpoides, a source of traditional medicines. J. Steroid Biochem. Mol. Biol. 158, 138–148. Tritsch, D., Zinglé, C., Rohmer, M., Grosdemange-Billiard, C., 2015. Flavonoids: true or promiscuous inhibitors of enzyme? The case of deoxyxylulose phosphate reductoisomerase. Bioorg. Chem. 59, 140–144. Vandenberg, L.N., Chahoud, I., Heindel, J.J., Padmanabhan, V., Paumgartten, F.J., Schoenfelder, G., 2010. Urinary, circulating, and tissue biomonitoring studies indicate widespread exposure to bisphenol a. Environ. Health Perspect. 118, 1055–1070.

Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements The authors acknowledge financial support of the Slovenian Research Agency (Grant No. P1-0208). VT thanks Charles University (SVV 260 414). References Asimakopoulos, A.G., Thomaidis, N.S., Kannan, K., 2014. Widespread occurrence of bisphenol A diglycidyl ethers, p-hydroxybenzoic acid esters (parabens), benzophenone type-UV filters, triclosan, and triclocarban in human urine from Athens, Greece. Sci. Total Environ. 470–471, 1243–1249. Baell, J.B., 2016. Feeling Nature’s PAINS: natural products, natural product drugs, and Pan assay interference compounds (PAINS). J. Nat. Prod. 79, 616–628. Bisson, J., McAlpine, J.B., Friesen, J.B., Chen, S.N., Graham, J., Pauli, G.F., 2016. Can invalid bioactives undermine natural product-based drug discovery? J. Med. Chem. 59, 1671–1690. Boonmuen, N., Gong, P., Ali, Z., Chittiboyina, A.G., Khan, I., Doerge, D.R., Helferich, W.G., Carlson, K.E., Martin, T., Piyachaturawat, P., Katzenellenbogen, J.A., Katzenellenbogen, B.S., 2016. Licorice root components in dietary supplements are selective estrogen receptor modulators with a spectrum of estrogenic and anti-estrogenic activities. Steroids 105, 42–49. Chen, M.N., Lin, C.C., Liu, C.F., 2015. Efficacy of phytoestrogens for menopausal symptoms: a meta-analysis and systematic review. Climacteric 18, 260–269. Chrzan, B.G., Bradford, P.G., 2007. Phytoestrogens activate estrogen receptor beta1 and estrogenic responses in human breast and bone cancer cell lines. Mol. Nutr. Food Res. 51, 171–177. Damstra, T., Barlow, S., Bergman, A., Kavlock, R., Van Der Kraak, G., 2002. Global Assessment of the State-of-the-Science of Endocrine Disruptors. World Health Organisation, Geneva, pp. 1–180. Diel, P., Schulz, T., Smolnikar, K., Strunck, E., Vollmer, G., Michna, H., 2000. Ability of xeno- and phytoestrogens to modulate expression of estrogen-sensitive genes in rat uterus: estrogenicity profiles and uterotropic activity. J. Steroid Biochem. Mol. Biol. 73, 1–10. EFSA, 2015. Risk assessment for peri- and post-menopausal women taking food supplements containing isolated isoflavones. EFSA 13. https://doi.org/10.2903/j.efsa.2015. 4246. Escande, A., Pillon, A., Servant, N., Cravedi, J.P., Larrea, F., Muhn, P., Nicolas, J.C., Cavaillès, V., Balaguer, P., 2006. Evaluation of ligand selectivity using reporter cell lines stably expressing estrogen receptor alpha or beta. Biochem. Pharmacol. 71, 1459–1469. Farmakalidis, E., Hathcock, J.N., Murphy, P.A., 1985. Oestrogenic potency of genistin and daidzin in mice. Food Chem. Toxicol. 23, 741–745. Freyberger, A., Schmuck, G., 2005. Screening for estrogenicity and anti-estrogenicity: a critical evaluation of an MVLN cell-based transactivation assay. Toxicol. Lett. 155, 1–13. Gautam, A.K., Bhargavan, B., Tyagi, A.M., Srivastava, K., Yadav, D.K., Kumar, M., Singh, A., Mishra, J.S., Singh, A.B., Sanyal, S., Maurya, R., Manickavasagam, L., Singh, S.P., Wahajuddin, W., Jain, G.K., Chattopadhyay, N., Singh, D., 2011. Differential effects of formononetin and cladrin on osteoblast function, peak bone mass achievement and bioavailability in rats. J. Nutr. Biochem. 22, 318–327. Gaya, P., Peiroten, A., Landete, J.M., 2017. Transformation of plant isoflavones into bioactive isoflavones by lactic acid bacteria and bifidobacteria. J. Funct. Foods 39, 198–205. Golden, R., Gandy, J., Vollmer, G., 2005. A review of the endocrine activity of parabens and implications for potential risks to human health. Crit. Rev. Toxicol. 35, 435–458. Goto, H., Terao, Y., Akai, S., 2009. Synthesis of various kinds of isoflavones, isoflavanes, and biphenyl-ketones and their 1,1-diphenyl-2-picrylhydrazyl radical-scavenging activities. Chem Pharm Bull (Tokyo) 57, 346–360. Gramec Skledar, D., Peterlin Mašič, L., 2016. Bisphenol a and its analogs: do their metabolites have endocrine activity? Environ. Toxicol. Pharmacol. 47, 182–199. Gramec Skledar, D., Tomašič, T., Carino, A., Distrutti, E., Fiorucci, S., Peterlin Mašič, L., 2016. New brominated flame retardants and their metabolites as activators of the pregnane X receptor. Toxicol. Lett. 259, 116–123. Gramec Skledar, D., Tomašič, T., Sollner Dolenc, M., Peterlin Mašič, L., Zega, A., 2018. Evaluation of endocrine activities of ellagic acid and urolithins using reporter gene assays. Chemosphere 220, 706–713. Gutendorf, B., Westendorf, J., 2001. Comparison of an array of in vitro assays for the assessment of the estrogenic potential of natural and synthetic estrogens, phytoestrogens and xenoestrogens. Toxicology 166, 79–89. Jefferson, W.N., Padilla-Banks, E., Clark, G., Newbold, R.R., 2002. Assessing estrogenic activity of phytochemicals using transcriptional activation and immature mouse uterotrophic responses. J Chromatogr B Analyt Technol Biomed Life Sci 777,

9