ELSEVIER

Contents lists available at SciVerse ScienceDirect

Toxicology and Applied Pharmacology

journal homepage: www.elsevier.com/locate/ytaap

T

Toxicology and Applied Pharmacology

====

Flavonoids exhibit diverse effects on CYP11B1 expression and cortisol synthesis

Li-Chuan Cheng, Lih-Ann Li *

Division of Environmental Health and Occupational Medicine, National Health Research Institutes, Zhunan, Miaoli, 35053 Taiwan, ROC

ARTICLE INFO

Article history: Received 17 August 2011

Revised 24 November 2011 Accepted 25 November 2011 Available online 8 December 2011

Keywords: Flavonoid Cortisol CYP11B1 Ad5 Signaling pathway

ABSTRACT

CYP11B1 catalyzes the final step of cortisol biosynthesis. The effects of flavonoids on transcriptional expression and enzyme activity of CYP11B1 were investigated using the human adrenocortical H295R cell model. All tested nonhydroxylated flavones including 3’,4’-dimethoxyflavone, o-naphthoflavone, and ß-naphthoflavone upregu- lated CYP11B1 expression and cortisol production, whereas apigenin and quercetin exhibited potent cytotoxicity and CYP11B1 repression at high concentrations. Nonhydroxylated flavones stimulated CYP11B1-catalyzed corti- sol formation at transcriptional level. Resveratrol increased endogenous and substrate-supported cortisol pro- duction like nonhydroxylated flavones tested, but it had no effect on CYP11B1 gene expression and enzyme activity. Resveratrol appeared to alter cortisol biosynthesis at an earlier step. The Ad5 element situated in the -121/-106 region was required for basal and flavone-induced CYP11B1 expression. Overexpression of COUP-TFI did not improve the responsiveness of Ad5 to nonhydroxylated flavones. Although COUP-TFI overex- pression increased CYP11B1 and CYP11B2 promoter activation, its effect was not mediated through the common Ad5 element. Treating cells with PD98059 (a flavone-type MEK1 inhibitor) increased CYP11B1 promoter activity, but not involving ERK signaling because phosphorylation of ERK1/2 remained unvarying throughout the course of treatment. Likewise, AhR was not responsible for the CYP11B1-modulating effects of flavonoids because incon- sistency with their effects on AhR activation. 3’,4’-dimethoxyflavone and 8-Br-cAMP additively activated CYP11B1 promoter activity. H-89 reduced 3’,4’-dimethoxyflavone-induced CYP11B1 promoter activation but to a lesser extent as compared to its inhibition on cAMP-induced transactivation. Our data suggest that constant ex- posure to nonhydroxylated flavones raises a potential risk of high basal and cAMP-induced cortisol synthesis in consequence of increased CYP11B1 expression.

@ 2011 Elsevier Inc. All rights reserved.

Introduction

Flavonoids are a diverse group of polyphenolic compounds that ex- hibit structural similarity to steroids. In nature, they are present in large quantities in the fruit peel, seeds, bark, and flowers of plants. There are growing interests in flavonoids in recent years because of their wide range of biological effects including anti-oxidant, anti-inflammatory, and anti-cancer activity (Croft, 1998). While flavonoid intake is consid- ered to be beneficial for cancer prevention, evidence suggests a poten- tial harm of flavonoids as endocrine disruptors. It has been shown that some flavonoid phytochemicals exhibit inhibitory effects on the ac- tivity of aromatase, a key enzyme for estrogen biosynthesis (Adlercreutz et al., 1993; Kellis and Vickery, 1984; Le Bail et al., 1998). Soy isoflavones daidzein, genistein, and biochanin A possess steric structures similar to the substrates of 3ß- and 17ß-hydroxysteroid de- hydrogenases. They can competitively inhibit the steroidogenic reac- tions catalyzed by these two dehydrogeases (Keung, 1995; Ohno et al., 2004). These soy isoflavones also inhibit 5x-reductase activity in genital

skin fibroblasts and prostate tissues, hence decreasing the local conver- sion of testosterone to dihydrotestosterone (a more potent androgen) (Evans et al., 1995). In addition, flavonoids have been implicated to modulate steroid metabolism by altering the activity of metabolizing enzymes responsible for steroid hydroxylation, glucuronidation, and sulfonation (Androutsopoulos et al., 2011; Harris and Waring, 2008; Pfeiffer et al., 2006).

Although endocrine toxicity of flavonoids historically centers on go- nadal hormones due to concerns on reproduction and development, in- creasing evidence shows that flavonoids may inhibit ACTH-stimulated cortisol production by adrenocortical cells (Kaminska et al., 2007; Mesiano et al., 1999; Ohno et al., 2002). Cortisol is the primary form of glucocorticoid in human. Glucocorticoids are originally defined for their importance in glucose metabolism. Indeed, they have profound in- fluences in a battery of physiological processes. The burst of cortisol se- cretion in response to ACTH stimulation is essential for fighting stress or overcoming traumatic situations. Cortisol levels in the absence of ACTH stimulation, i.e., basal cortisol levels, are also crucial to human health. Consistent high basal cortisol levels impair cognition, suppress thyroid function, decrease bone density and muscular mass, but increase obesi- ty. On the other hand, consistent low basal cortisol levels cause constant fatigue, depression, and vulnerability to allergies and infections. Little is known about the effects of flavonoids on basal cortisol synthesis.

* Corresponding author at: Division of Environmental Health and Occupational Medicine, National Health Research Institutes, R1-5034, 35 Keyan Rd., Zhunan, Miaoli, 35053 Taiwan, ROC. Fax: + 886 37 587406.

E-mail address: lihann@nhri.org.tw (L .- A. Li).

Cortisol is mainly produced by adrenal cortex. Minor amounts of cortisol are also synthesized in extra-adrenal tissues, such as brain, heart and vessels. Both adrenal and extra-adrenal cortisol syntheses require an enzyme 11ß-hydroxylase (CYP11B1) that catalyzes the final conversion of cortisol from 11-deoxycortisol in the biosynthetic pathway (Davies and Mackenzie, 2003). Alteration of CYP11B1 ex- pression or activity would disrupt cortisol homeostasis. Our previous study found that 3’,4’-dimethoxyflavone (3’,4’-DMF) induced basal CYP11B1 mRNA expression in human adrenocortical H295R cells (Lin et al., 2006). To learn the effects of flavonoids on basal cortisol synthesis, we tested several natural and synthetic flavonoids in the H295R cell line, a simple but effective model recommended for ste- roid biosynthesis study (Harvey and Everett, 2003). Diverse effects were detected. Nonhydroxylated flavones 3’,4’-DMF, &x- naphthoflavone («-NF), and ß-naphthoflavone (B-NF) significantly increased intracellular CYP11B1 mRNA abundance as well as cortisol synthesis. In contrast, hydroxylated flavonoids tested had no effect or mild inhibition on CYP11B1 expression. An Ad5 element located in the proximal upstream region of the human CYP11B1 gene was in- dispensable for flavone-induced upregulation. The involvement of a variety of signaling mediators was examined in this study.

Materials and methods

Cell culture and treatment. Human adrenocortical H295R cells were cultured in phenol red-free DMEM/F12 medium (Sigma-Aldrich, St. Louis, MO, USA) plus 10% charcoal/dextran (Sigma-Aldrich)-treated fetal bovine serum (Invitrogen, Carlsbad, CA, USA). 3’,4’- dimethoxyflavone, «-naphthoflavone, ß-naphthoflavone, apigenin, chrysin, quercetin, daidzein, resveratrol, and PD98059 (Sigma-Aldrich) were dissolved in DMSO as 1000-fold stocks. H295R cells were treated with 0-100 UM of flavonoids for 24h in the dose-response experi- ments, and 10 uM for 0-24 h in the time course study. The DMSO con- centration in all treatments was 0.1%.

Gene expression analysis. CYP11B1 mRNA expression level was measured by real-time RT-PCR. The procedures of RNA extraction and RT-PCR, including primer sequences, were described in a previ- ous study (Lin et al., 2006). Transcript abundance was determined by calibration against a standard curve and normalized to the expres- sion level of the house-keeping gene porphobilinogen deaminase.

Cortisol measurement. H295R cells were incubated in the serum-free medium with or without 1 mg/ml 11-deoxycortisol for 1 h after being treated with 10 UM of the tested flavonoids for the indicated durations. The medium and cells were harvested for cortisol and protein measure- ments, respectively. Debris in the medium was removed by micro cen- trifuge at 10,000xg for 10 min. 10 ng/ml of cortisol-9,11,12,12-d4 (C/ D/N Isotopes, Pointe-Claire, Quebec, Canada) was added to the medium as an internal control before the cortisol content was measured using PE200 HPLC (Perkin-Elmer, Norwalk, CT, USA) followed by API3000 tan- dem mass spectrometry (Applied Biosystems, Foster city, CA, USA). Each medium sample was online pre-cleaned with a LiChrospher RP-18 col- umn (25 mm× 4 mm, 25 um, Merk, Darmstadt, Germany) and analyzed using a Merk C18 analytic column (55 mm × 4 mm, 3 um) with a gradi- ent mobile phase of 50-95% methanol containing 0.1% formic acid at a flow rate of 0.8 ml/min for 18 min. Cortisol was protonated (M+H+) by electrospray ionization and scanned at m/z =363.1 ±121.1 after collision. Cortisol production in the presence and absence of 11- deoxycortisol was normalized to cellular protein content. Total protein in each cell lysate was measured by Micro BCA protein assay (Pierce Bio- technology, Rockford, IL, USA). CYP11B1-catalyzed substrate conversion, i.e., CYP11B1 activity, was calculated by subtraction of the mean cortisol production in the absence of 11-deoxycortisol (substrate) from the mean production in the presence of 11-deoxycortisol.

Transfection analysis. The effects of flavonoids on CYP11B1 and CYP11B2 promoter activities were assessed using luciferase reporters driven by varying lengths of CYP11B1 and CYP11B2 promoters (Lin et al., 2006), whereas the effects on AhR activity were assessed using the AhR responsive reporter 4xDRE-TATA-Luc. Regulation of CYP11B1 and CYP11B2 by COUP-TFI was assayed by cotransfection of a COUP-TFI ex- pression plasmid (Open Biosystems, Huntsville, AL, USA) with indicated reporters. Transfection analysis was performed as described previously (Lin et al., 2006). The transfected cells were treated with 10 UM of the tested flavonoids for 24 h or 72 h before reporter activity was assayed. Cyclic AMP-induced transactivation was measured by addition of 1 mM 8-Br-cAMP (Sigma-Aldrich) alone or together with 10 uM 3’,4’-DMF to the cells transfected with the CYP11B1(-394) reporter for 24 h. To block cAMP-dependent PKA activity, 20 µM H-89 (InvivoGen, San Diego, CA, USA) was added to the transfected cells 1 h before addition of 8-Br-cAMP or 3’,4’-DMF. Because cAMP positively regulated expres- sion of the internal control as well, normalization was not performed in the experiments involving 8-Br-cAMP.

Site-directed mutagenesis. The Ad5, SF-1, and AP-1 sites of CYP11B1 were mutated by two-step PCR. In the first-step PCR, sense and antisense mutation-containing oligonucleotides were paired with a downstream antisense oligonucleotide and an upstream sense oligonucleotide, re- spectively, to amplify the sequences flanking the desired mutation. The PCR products were then 1:1 mixed and used as the templates for the second-step PCR. The upstream and downstream oligonucleotides were used as primers in the second reaction. Mutation was confirmed by sequencing. The sequences of wild type and mutant sites are listed in Table 1.

Western blot analysis. Protein extraction and Western blotting were performed as described previously (Tsou et al., 2005). Rabbit anti- phospho-ERK1/2 (Cell Signaling Technology, Danvers, MA, USA) and mouse anti-ß-actin (Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used as primary antibodies. Peroxidase-conjugated goat anti- rabbit IgG and anti-mouse IgG (Santa Cruz Biotechnology) were used to recognize the corresponding primary antibodies.

Statistical analysis. All data are presented as mean ±SE. The dose- response relationship was analyzed using one-way ANOVA followed by Bonferroni’s post hoc test. The significance of the effect on cortisol production was determined by the independent two-sample t-test, whereas the significance of the effect on CYP11B1 activity was deter- mined by z-test.

Results

Effects of flavonoids on CYP11B mRNA expression

To learn the effect of flavonoid treatment on CYP11B1 gene ex- pression, six flavones including synthetic flavones 3’,4’-dimethoxy- flavone (3’,4’-DMF), a-naphthoflavone («-NF), and ß- naphthoflavone (ß-NF) and natural flavones apigenin, chrysin, and quercetin (Fig. 1) were tested in the human adrenocortical H295R cells at concentrations ranging from 0 to 100 uM. Cytotoxicity was ob- served with @-NF, B-NF, apigenin, and quercetin at and above 30 uM after a 24-h treatment. One-way ANOVA analysis indicated that all the tested flavones except apigenin had a dosage effect on CYP11B1

Table 1 Sequences to be mutated by site-directed mutagenesis.
SiteLocationWild typeMutant
Ad5-121/-106CCTGACCTCTGCCCTCCCTGcagTCgaCCCTC
SF-1-242/-234CCAAGGCTCCCAgatCTC
AP-1-251/-242ATGAATAATCgaattcAATC

mRNA expression level under permissible doses (p<0.01). 3’,4’-DMF, @-NF, and ß-NF exhibited a significant dose-dependent stimulation on CYP11B1 expression. In contrast, chrysin (30 and 100 µM) and quercetin (3 and 10 µM) repressed CYP11B1 expression by about 30-50%. The CYP11B1 gene was most sensitive to the ß-NF treatment. A 24-h treat- ment with 1 µM B-NF raised CYP11B1 mRNA level by 3.65±0.27-fold, while the other five flavones caused little changes at this concentration. 3’,4’-DMF exhibited the widest range of influence among these fla- vones. The maximal CYP11B1 induction was detected with 30 uM 3’,4’-DMF. CYP11B1 mRNA level increased 11.86±0.32-fold after a 24-h treatment with 30 µM 3’,4’-DMF (Fig. 2A).

We further examined the temporal response of CYP11B1 to 3’,4’- DMF, &-NF, and ß-NF treatments in a 24-h course. 10 µM was selected as the test dose based on the above dose-effect study. These three fla- vones raised CYP11B1 mRNA level approximately 2-fold after 2 h of treatment. The abundance of CYP11B1 mRNA under @-NF treatment remained steady since then. 3’,4’-DMF and ß-NF continued elevating CYP11B1 expression when treatment was lengthened to 10 h. The in- duction raised by B-NF was 7.58 ± 0.28-fold at this time point, almost twice higher than that raised by 3’,4’-DMF. B-NF-induced CYP11B1 ex- pression sharply declined when treatment was further prolonged. In contrast, lengthening of the 3’,4’-DMF treatment to 24 h did not change CYP11B1 mRNA abundance much (Fig. 2B). In addition to 3’,4’-DMF, a-NF, and B-NF, the response of CYP11B1 to soy isoflavone daidzein and grape stilbene resveratrol (Fig. 1) was determined. As the vehicle control, 10 uM daidzein and resveratrol had no remarkable effect on CYP11B1 mRNA expression throughout the course (Fig. 2B).

Effects of flavonoids on cortisol synthesis and CYP11B1 enzymatic activity

Cortisol production in response to flavonoid treatment was deter- mined with and without addition of exogenous CYP11B1 substrate

Fig. 2. CP11B1 mRNA expression in response to flavonoid treatment. (A) Dose-response. H295R cells were treated with assorted concentrations of flavonoids for 24 h before RNA was isolated (n=3). Data at cytotoxic doses were excluded. (B) Time course of effects. Each flavonoid was added to cells at 10 uM (n=3). * p<0.01 as compared to 0 M.

A

14

3’,4’-DMF

mRAN expression (fold)

12

a-NF

*

10

B-NF

apigenin

8

chrysin

*

*

6

quercetin

*

*

4

*

*

*

2

L

*

*

*

*

0

88

0

1

3

10

30

100

concentration (LM)

B

- DMSO

3’,4’-DMF

a-NF

9

B-NF

daidzein

resveratrol

8

mRNA expression (fold)

7

6

5

4

3

T

2

L

1

0

0

5

10

15

20

25

time of treatment (h)

Fig. 1. Chemical structures of flavonoids to be tested.

CH3

o

o

CH3

0

O

O

o

0

0

3’,4’-dimethoxyflavone

a-naphthoflavone

B-naphthoflavone

OH

OH

OH

HO

O

HO

o

HO

O.

OH

OH

o

OH

0

OH

0

apigenin

chrysin

quercetin

OH

HO

O.

HO

O

OCH3

NH2

O

OH

OH

0

daidzein

resveratrol

PD98059

11-deoxycortisol (1 mg/ml). The amount of cortisol secreted to the me- dium during the hour after a 24-h treatment with 10 uM flavonoid was measured and normalized to the cellular protein content. 3’,4’-DMF, &x- NF, and ß-NF raised the normalized cortisol production far above the vehicle control regardless of substrate addition (Fig. 3A). Resveratrol also increased endogenous and substrate-supported cortisol synthesis above the vehicle control (Fig. 3A). However, the increase raised by res- veratrol was not due to the effect on CYP11B1. Comparable CYP11B1 enzyme activities calculated as the difference of cortisol synthesis with and without addition of exogenous CYP11B1 substrate were detected when H295R cells were treated with vehicle, daidzein, and resveratrol. In contrast, a 24-h treatment with 10 µM 3’,4’-DMF, &-NF, and ß-NF significantly elevated CYP11B1 enzyme activity from 3.14± 0.11 (vehicle control) to 26.54± 0.73, 10.73 ± 0.46, and 23.93 ± 1.16 ng cortisol/mg protein, respectively (Fig. 3B).

The temporal response in cortisol production was also examined. 11-deoxycortisol (1 mg/ml) was given to H295R cells after 0, 2, 10, and 24 h of flavonoid treatment (10 uM). The substrate-supported cortisol production by vehicle, daidzein, and resveratrol-treated cells remained steadily low at all time points (Fig. 3C), similar to the findings in CYP11B1 mRNA expression (Fig. 2B). Incubation of H295R cells with 10 µM 3’,4’-DMF, &-NF, or B-NF for 2 h significantly induced CYP11B1 mRNA expression (Fig. 2B), but their stimulation on cortisol production was visible later (Fig. 3C). Based on the curves, 3’,4’-DMF- and ß-NF-induced increases of cortisol production appeared after about 6 h of incubation, whereas more than 10 h of treatment were required for «-NF to increase cortisol synthesis. After 24 h of treatment, 3’,4’-DMF, &-NF, and ß-NF raised cortisol production in the following hour to 10.23±0.28, 4.25 ±0.14, and 10.66 ±0.33 ng, respectively (Fig. 3C).

Responsive site in human CYP11B1 promoter

To locate the flavone responsive site in the human CYP11B1 promot- er, 105 to 4511-bp upstream sequences were cloned in front of a firefly luciferase reporter. Although we could not assess ß-NF-mediated tran- scriptional regulation using this reporter system due to the inhibition of ß-NF on firefly luciferase activity (Wang, 2002), transfection analysis indicated that the - 394/- 105 upstream region was important for the transcriptional response to 3’,4’-DMF and @-NF in H295R cells. Com- pared with vehicle, 3’,4’-DMF and @-NF (10 uM, 24 h) raised similar levels of luciferase activity expressed from the 105-bp CYP11B1 promot- er, but both flavones significantly induced luciferase activity from the 394-bp promoter. Although lengthening of the CYP11B1 promoter did not further increase expression of reporter activity, the stimulatory ef- fects of 3’,4’-DMF and @-NF remained considerable (Figs. 4A and B).

The - 394/- 105 region contains three known regulatory sites, Ad5, SF-1 binding site, and AP-1 binding site (Fig. 4A). We mutated these sites by site-directed mutagenesis in addition to shortening of the 394-bp promoter to 294 bp and 215 bp. Transfection analysis further mapped the flavone responsive site to the -215/- 105 region, where the Ad5 site is located. Mutation of the Ad5 site diminished basal and 3’,4’-DMF-induced reporter activities to the level comparable with the 105-bp promoter, suggesting that the Ad5 site was essential for both basal and flavone-induced CYP11B1 expression. Mutation of the SF-1 and AP-1 binding sites also reduced reporter activity under vehicle and 3’,4’-DMF treatments. However, the inhibition was much weaker than the Ad5 mutation. The SF-1 and AP-1 binding sites appeared irre- sponsible for flavone-induced expression because mutation of these two sites did not block 3’,4’-DMF from upregulation of reporter expres- sion (Fig. 4C).

Role of ERK, AhR, and PKA

PD98059 is a flavone-type mitogen-activated protein kinase ki- nase 1 (MEK1) inhibitor (Fig. 1). Treating H295R cells with 10 uM

Fig. 3. Cortisol synthesis and CYP11B1 activity in response to flavonoid treatment. (A) Cortisol synthesis with and without addition of exogenous CYP11B1 substrate. After 24 h of treatment with 10 uM flavonoid, H295R cells were incubated in serum- free medium with and without 1 mg/ml 11-deoxycortisol (CYP11B1 substrate) for 1 h (n=6). (B) CYP11B1 activity. Cortisol formation catalyzed by CYP11B1 was calcu- lated as the difference between production with substrate addition and without sub- strate addition (based on panel A data). (C) Time course of substrate-supported cortisol synthesis. At each indicated time point during the 24-h course of flavonoid treatment (10 µM), 6 dishes of cells were removed from each flavonoid group and in- cubated with 1 mg/ml 11-deoxycortisol for one additional hour. * p<0.01 as compared to vehicle control.

A

35

*

*

1

ng cortisol/mg protein

30

- substrate

L

25

+ substrate

20

15

*

I

10

*

5

*

*

*

*

0

DMSO

3’,4’-DMF

a-NF

B-NF

daidzein

resveratrol

B

30

*

CYP11B1 activity (ng cortisol/mg protein)

25

*

L

20

15

*

10

5

0

DMSO 3’,4’-DMF

a-NF

B-NF

daidzein resveratrol

C

12

DMSO

synthesized in one hour

10

3’,4’-DMF

a-NF

ng cortisol

8

B-NF

daidzein

6

resveratrol

4

2

0

0

5

10

15

20

25

time of treatment (h)

PD98059 for 24 h after transfection with CYP11B1-luciferase reporters also induced reporter expression from the 394-bp and 4511-bp CYP11B1 promoters but not from the - 105 promoter (Fig. 5A) as the treatments with 3’,4’-DMF and @-NF (Fig. 4B). However, Western blot analysis showed that phosphorylation of extracellular signal- regulated kinase 1 and 2 (ERK1/2 that are MEK1 targets) was not blocked by the PD98059 treatment. ERK1/2 remained steadily phos- phorylated during the 24-h course (Fig. 5B). Inhibition of ERK signal- ing was not the cause for the flavone-induced CYP11B1 upregulation.

It has been known that flavonoids can modulate expression of some cytochrome P450 genes (CYPs) through aryl hydrocarbon re- ceptor (AhR) by serving as a ligand (Moon et al., 2006). To check the possibility of AhR as the transcription factor mediating the stimu- lation of flavonoids on CYP11B1, we examined the effects of flavo- noids on AhR activity. The transcriptional activity of AhR in H295R cells was assessed by transfection of an AhR responsive reporter

Fig. 4. Localization of flavonoid responsive site in CP11B1 promoter. (A) Human CP11B1 promoter. (B-C) Transfection analysis of CP11B1 promoter activity. H295R cells were treated with vehicle (DMSO) or 10 uM flavonoid for 24 h after CP11B1-luciferase re- porter transfection (n=4). Number indicates the position of the 5' end of a promoter relative to the transcriptional start site (+1). Ad5m, AP1m, and SF1m mean mutations occurring at the Ad5, AP-1, and SF-1 sites, respectively.

A

-4511

-3527

-1987

-394

+3

AP1/SF1

Ad5

cre

-394

-294

-215

-105

+3

B

60

relative reporter activity

50

DMSO

3’4’-DMF

40

a-NF

30

20

2

I

1

I

10

I

I

0

vector

-105

-394

-1987

-3527

-4511

CYP11B1 promoters

C

180

relative reporter activity

150

DMSO

I

3’.4’-DMF

120

90

I

I

60

I

£

30

0

vector

-105

-215

-294

-394

Ad5m

AP1m

SF1m

CYP11B1 promoters

into the cells. Expression of the reporter in response to AhR activation was repressed when the transfected cells were treated with 10 µM 3’,4’-DMF for 24 h. ß-NF and resveratrol treatments (10 uM, 24 h) in- creased reporter expression but to different degrees. In contrast, 10 µM &-NF and daidzein had little effects on AhR activity (Fig. 6). Ap- parently, these flavonoids had discrete effects on AhR activity (Fig. 6) and CYP11B1 mRNA expression (Fig. 2B). It was unlikely that 3’,4’- DMF, &-NF, and ß-NF upregulated CYP11B1 through AhR.

Quercetin and genistein have been shown to increase the intracellu- lar cAMP concentration and, in turn, induce mRNA expression of aroma- tase (CYP19) through cAMP-dependent protein kinase A (PKA) in H295R cells (Sanderson et al., 2004). The human CYP11B1 gene also contains a cAMP responsive DNA element (cre) at the - 71/- 64 region (Wang et al., 2000). However, the cre site is not responsible for nonhy- droxylated flavone-induced CYP11B1 upregulation since the 105-bp cre-containing CYP11B1 promoter did not respond to the stimulation of 3’,4’-DMF, c-NF, and PD98059 in transfection analysis (Figs. 4 and 5A). Even so, we could not rule out the possibility of involvement of PKA. Therefore, we examined the impact of blockage of PKA signaling on nonhydroxylated flavone-mediated CYP11B1 regulation. Data showed that the 394-bp CYP11B1 promoter responded to 1 mM 8-Br- CAMP (cAMP analogue) and 10 µM 3’,4’-DMF in an additive manner

Fig. 5. Effects of PD98059 on CP11B1 promoter activity (A) and ERK phosphorylation (B). CP11B1 promoter activity was assayed by transfection of CP11B1-luciferase reporters into H295R cells (n=4). Cells were treated with vehicle (DMSO) or 10 µM PD98059 for 24 h after transfection. Phosphorylation of ERK1/2 during the 24-h PD98059 treatment was examined by Western blotting.

A

100

relative reporter activity

80

DMSO

I

PD98059

60

I

40

I

20

I

I

I

0

vector

-105

-394

-4511

CYP11B1 promoters

B

0

1

2

4

6

10

24 h

p-ERK1/2

B-actin

in H295R cells (Fig. 7A). Addition of a PKA inhibitor H-89 (20 µM) to cells effectively removed the positive effect of 8-Br-cAMP by ~73% but only reduced that of 3’,4’-DMF by ~ 36% (Figs. 7B and C). Taken together, 3’,4’-DMF appeared to regulate CYP11B1 independently of PKA.

Coordinate regulation of CYP11B1 and CYP11B2 by 3’,4’-DMF via the Ad5 site

The Ad5 element is also present in the upstream regions of other steroidogenic genes, for example, CYP11B2 (Clyne et al., 1997; Morohashi et al., 1992). Transfection analysis showed that the CYP11B2 promoter responded to the 3’,4’-DMF treatment (10 UM, 24 h) in a pattern similar to CYP11B1. When the Ad5 site was not in- cluded in the promoters tested, both CYP11B1 and CYP11B2 pro- moters had a fairly small response to 3’,4’-DMF. Extension of both promoters beyond the Ad5 site did not only increase basal promoter activity but also grant the sensitivity to 3’,4’-DMF. Further extension to enclose the SF-1 binding site increased reporter activity but did not additionally increase the responsiveness to 3’,4’-DMF. Relative to the vehicle treatment, the increase of reporter activity induced by 3’,4’-DMF halted (Figs. 8A and B).

Fig. 6. AhR activity in response to flavonoid treatments. AhR activity in H295R was assayed by transfection of the cells with the reporter 4xDRE-TATA-Luc. After transfection, the cells were treated with vehicle (DMSO) or 10 uM flavonoid for 72 h (n=6). * p<0.01 as compared to vehicle control.

6

*

relative AhR activity

5

I

4

3

2

*

1

E

*

0

DMSO

3’,4’-DMF

a-NF

B-NF

daidzein

resveratrol

Fig. 7. 3',4'-DMF and 8-Br-cAMP activated CYP11B1 promoter activity in a separate and additive manner. The CYP11B1(-394) reporter was transfected into H295R cells to assess the effect of 3',4'-DMF and 8-Br-CAMP on CYP11B1 promoter activity (n=4). After transfection, cells were treated with either 1 mM 8-Br-cAMP, 10 uM 3',4'-DMF or both for 24 h (A). To test the involvement of PKA, PKA activity was inhibited by addition of 20 uM H-89 to cells 1 h before the 24-h treatment with 1 mM 8-Br-cAMP (B) or 10 uM 3',4'-DMF (C).

A

12

B

12

C

6

relative reporter activity

10

10

I

5

8

8

4

6

6

3

4

4

2

2

2

1

0

0

0

8-Br-cAMP

-

+

-

+

8-Br-cAMP

-

+

+

3’,4’-DMF

-

+

-

+

3’,4’-DMF

-

-

+

+

H-89

+

-

+

H-89

-

-

+

+

The Ad5 sites of CYP11B1 and CYP11B2 contain the binding half site of COUP-TFI (5’-AGGTCA) in the antisense strand (Fig. 8A) (Montemayor et al., 2010). It has been reported that overexpression of COUP-TFI upregulates CYP11B2 promoter activity through the Ad5 site (Shibata et al., 2004). However, our analysis demonstrated that transfection of a COUP-TFI expression plasmid into H295R cells raised the activity of CYP11B1(-105) and CYP11B2(-94) promoters in a similar magnitude as to CYP11B1(-152) and CYP11B2(-213) promoters, respectively (Fig. 8C). COUP-TFI appeared to exert its regu- latory effect at a site downstream of Ad5 in both promoters. In addition, COUP-TFI overexpression did not enhance the stimulation of 3’,4’-DMF on the Ad5-containing CYP11B1 and CYP11B2 promoters. In contrast, COUP-TFI increased the response of Ad5-deficient CYP11B1(-105) promoter to 3’,4’-DMF by near 2-fold (Fig. 8D). These results led to a conclusion that COUP-TFI was not the Ad5-binding protein responsible for 3’,4’-DMF-elicited gene induction.

Discussion

The intake of flavonoids from food is unlikely to have a major impact on adrenal steroidogenesis. Even though Asian diets contain more veg- etables and soy products than typical Western diets, the mean daily soy isoflavone intake was 8-50 mg in Asian countries. The mean plasma isoflavone concentration is about 0.8-0.9 uM for Japanese on a tradi- tional soy-rich diet (Mortensen et al., 2009). However, these days there is a growing trend towards consumption of flavonoids as nutri- tional supplements with the intention to reduce the risk of cancer, oste- oporosis, Alzheimer’s disease, heart problem, and so on. In addition, more and more synthetic flavonoids are created for therapeutic pur- poses. High doses of exposure may take place when flavonoids are ingested in the form of nutritional supplement or medicine. This study investigates the effects of flavonoids on basal cortisol biosynthesis with an emphasis on the final rate-limiting step catalyzed by CYP11B1.

Fig. 8. Role of COUP-TFI in regulation of CYP11B1 and CYP11B2 promoter activities. (A) Structures of the human CYP11B1 and CYP11B2 promoters to be tested. Number indicates the position upstream of the transcriptional start site (+1). The Ad5 sequences of both genes are shown at the bottom. (B) Response to 3',4'-DMF treatment. H295R cells were treated with vehicle (DMSO) or 10 µM 3',4'-DMF for 24 h after CP11B1-luciferase reporter transfection (n =4). (C) Response to COUP-TFI overexpression. A COUP-TFI expression plasmid or its vector was transfected into H295R cells along with indicated CP11B1-luciferase reporters (n =4). The effect of COUP-TFI on reporter activity was measured by comparison of the COUP-TFI cotransfection to the vector control. (D) Response to 3',4'-DMF treatment in the absence and presence of exogenous COUP-TFI. Cotransfection was performed as described in panel C (n =4). The effect of a 24-h 10-uM 3',4'-DMF treatment on reporter activity was determined relatively to the vehicle treatment.

A

B

-394

-215

-105

35

35

relatvie reporter activity

DMSO

CYP11B1

30

3’,4’-DMF

30

25

25

-512

-213

-94

20

20

CYP11B2

SF-1

Ad5 .cre

15

15

10

10

11B1-Ad5 5’ -TGACCTCTG

11B2-Ad5 5’ -TGACCTTCG

5

5

0

-105

-152

-394

0

-94

-213

-521

CYP11B1

CYP11B2

C

8

8

D

5

5

COUP-TFI effect (fold) on reporter activity

7

7

vector

3’,4’-DMF effect (fold) on reporter activity

6

6

4

COUP-TFI

4

5

5

3

3

4

4

3

3

2

2

2

2

1

1

1

1

0

0

0

-105

-152

-394

-94

-213

-521

0

-105

-152

-394

-94

-213

-521

CYP11B1

CYP11B2

CYP11B1

CYP11B2

All synthetic flavonoids tested, including PD98059, were nonhy- droxylated flavones (Fig. 1). These nonhydroxylated flavones exhibited a capability to upregulate CYP11B1 expression. In contrast, all hydroxyl- ated flavonoids tested (Fig. 1) had little or mild inhibitory effects on basal CYP11B1 expression. We speculated that cytotoxicity caused the decline of CYP11B1 expression, even though chrysin at 30 and 100 µM did not cause obvious cell death as apigenin and quercetin. Time- effect analysis showed that nonhydroxylated flavone-induced substrate-supported cortisol production occurred hours later than CYP11B1 gene induction. The lag indicated that the rise of basal cortisol synthesis in response to the nonhydroxylated flavone treatments was attributed to CYP11B1 upregulation rather than a direct positive effect on the enzyme activity of CYP11B1. Resveratrol (3,5,4’-trihydroxylstil- bene, Fig. 1) also increased basal cortisol synthesis at 10 µM, but to a much lesser extent than 3’,4’-DMF, &-NF, and ß-NF. Whereas these non-hydroxylated flavones increased basal cortisol synthesis mainly by upregulating CYP11B1 expression, resveratrol had no effect on either CYP11B1 gene expression or CYP11B1 enzyme activity. Resveratrol was likely to affect an earlier step in the cortisol biosynthesis pathway. Pre- vious studies using rodent adrenocortical and testicular cells as models suggested that StAR and CYP21, in charge of the mitochondrial transfer of cholesterol and the formation of cortisol precursor, respectively, might be resveratrol responsive genes (Chen et al., 2007; Supornsilchai et al., 2005).

Despite little information about the in vivo adrenocortical effect of nonhydroxylated flavones in humans and other mammals, rainbow trout studies showed that a single intraperitoneal injection of B-NF at 10 or 50 mg/kg body mass increased plasma cortisol levels (Aluru and Vijayan, 2004; Tintos et al., 2008). Addition of @-NF in feed to provide ~10 mg/kg body mass/day for 5 days also raised plasma cor- tisol by 30% in rainbow trout. Similar feeding with ß-NF did not ele- vate plasmid cortisol level, but H-pregnenolone metabolism analysis in head kidney tissue slices showed an increased synthesis of cortisone, a dehydrogenated cortisol metabolite that could be con- verted back to cortisol in target tissues. Simultaneous feeding with œ- NF and ß-NF increased the synthesis of both cortisol and cortisone in head kidney, suggesting a cumulative effect for combined flavone treatment (Aluru et al., 2005). Intraperitoneal injection of @-NF and B-NF did not significantly increase CYP11B1 mRNA expression in head kidney (Aluru and Vijayan, 2004). This finding suggests that a different regulatory mechanism may exist in fish. On the other hand, since fish lacks a true adrenal gland and adrenal-like cells inter- mingle with renal cells in head kidney, the intermingling may lower quantitative accuracy and lead to a result of lack of significance.

The surge of plasma cortisol in rainbow trout declined with time after dosing of B-NF, reflecting that the in vivo effect of flavonoids depended on bioavailability. Flavonoids, for example, ß-NF and 3’,4’-DMF, can regulate transcriptional expression of the CYP1 family of metabolizing genes by modulating AhR activity. Flavonoids also af- fect the activity of phase I and II metabolizing enzymes (Moon et al., 2006). While non-substituted «-NF stimulates CYP3A4-catalyzed ox- idation, hydroxylated flavonoids inhibit CYP3A4 activity and the in- hibitory potency increases with the number of hydroxyl groups (Ho et al., 2001; Ueng et al., 1997). Methoxyl substitution also influences the susceptibility of flavonoids to CYP-mediated metabolism (Androutsopoulos et al., 2011; Walle and Walle, 2007). In general, methoxylated flavonoids are more resistant to metabolism. On the contrary, hydroxyl substitution renders flavonoids susceptible to glu- curonidation and sulfation, hence facilitating the excretion of flavonoids.

Our transfection analysis showed that the Ad5 site was indispens- able for nonhydroxylated flavone-elicited CYP11B1 gene induction. Deletion or mutation of the Ad5 site eliminated the responsiveness of the CYP11B1 promoter to nonhydroxylated flavones. This regula- tion was not specific for CYP11B1. CYP11B2 (aldosterone synthetase) was subject to similar regulation, suggesting that nonhydroxylated

flavones might coordinate cortisol and aldosterone synthesis by cohe- sive regulation of steroidogenic gene expression via the common Ad5 site. COUP-TFI and SF-1 had been reported to bind to the Ad5 site of CYP11B2 in a mutually exclusive manner. The binding of COUP-TFI resulted in transactivation, whereas the binding of SF-1 led to trans- repression (Shibata et al., 2004). However, our data suggested that COUP-TFI increased CYP11B1 and CYP11B2 promoter activation through a site downstream of Ad5. In addition, overexpression of COUP-TFI did not improve the responsiveness of Ad5 to nonhydroxy- lated flavones. AhR regulated stress-induced cortisol response in rain- bow trout (Aluru and Vijayan, 2006), but the AhR-modulating effects of 3’,4’-DMF, a-NF, B-NF, and resveratrol were not coherent with their effects on CYP11B1 expression and cortisol synthesis in human adrenocortical H295R cells. Both COUP-TFI and AhR were not likely to be the transcription factor responsible for the nonhydroxylated flavone-induced regulation at the Ad5 site.

Our data also indicated that PKA and ERK signaling pathways were not involved in nonhydroxylated flavone-elicited CYP11B1 induction. However, some other protein kinase pathway might be involved. H- 89 is a potent PKA inhibitor, but this compound also moderately an- tagonizes other protein kinases, e.g., PKG, by competitive binding to the ATP pocket (Engh et al., 1996). Treating H295R cells with H-89 significantly suppressed cAMP-signaled CYP11B1 promoter activation but only moderately reduced 3’,4’-DMF-activated promoter activity. The latter reduction appeared not due to blockade of PKA. Further- more, nonhydroxylated flavones might have an effect on ACTH- controlled cortisol synthesis in addition to basal synthesis since 3’,4’-DMF and 8-Br-cAMP increased CYP11B1 promoter activity in an additive manner. The mechanism underlying nonhydroxylated flavone-induced CYP11B1 and CYP11B2 upregulation is yet to be elucidated.

In conclusion, this study demonstrated that induction of CYP11B1 gene expression could have a profound consequence in cortisol pro- duction. Although bioavailability varies widely dependent on the na- ture of flavonoids, constant exposure to nonhydroxylated flavones raises a potential risk of high basal and cAMP-induced cortisol syn- thesis as a result of elevated CYP11B1 expression. The dietary and pharmaceutical usage of nonhydroxylated flavones deserves alertness.

Conflict of interest statement

The authors declare no conflicts of interest.

Acknowledgments

This work was supported by grants from the National Health Research Institutes (EO094-PP01) and the National Science Council (NSC98-2627- B400-001), Republic of China (Taiwan). We thank Mr. Chien-Jen Wang and Ms. Chun-Ju Lin of the National Health Research Institutes for assis- tance in LC-MS-MS analysis and real-time PCR, respectively.

References

Adlercreutz, H., Bannwart, C., Wahala, K., Makela, T., Brunow, G., Hase, T., Arosemena, P. J., Kellis Jr., J.T., Vickery, L.E., 1993. Inhibition of human aromatase by mammalian lignans and isoflavonoid phytoestrogens. J. Steroid Biochem. Mol. Biol. 44, 147-153.

Aluru, N., Vijayan, M.M., 2004. beta-Naphthoflavone disrupts cortisol production and liver glucocorticoid responsiveness in rainbow trout. Aquat. Toxicol. 67, 273-285. Aluru, N., Vijayan, M.M., 2006. Aryl hydrocarbon receptor activation impairs cortisol response to stress in rainbow trout by disrupting the rate-limiting steps in steroidogenesis. Endo- crinology 147, 1895-1903.

Aluru, N., Renaud, R., Leatherland, J.F., Vijayan, M.M., 2005. Ah receptor-mediated impair- ment of interrenal steroidogenesis involves StAR protein and P450scc gene attenuation in rainbow trout. Toxicol. Sci. 84, 260-269.

Androutsopoulos, V.P., Papakyriakou, A., Vourloumis, D., Spandidos, D.A., 2011. Comparative CYP1A1 and CYP1B1 substrate and inhibitor profile of dietary flavonoids. Bioorg. Med. Chem. 19, 2842-2849.

Chen, Y.C., Nagpal, M.L., Stocco, D.M., Lin, T., 2007. Effects of genistein, resveratrol, and quercetin on steroidogenesis and proliferation of MA-10 mouse Leydig tumor cells. J. Endocrinol. 192, 527-537.

Clyne, C.D., Zhang, Y., Slutsker, L., Mathis, J.M., White, P.C., Rainey, W.E., 1997. Angiotensin II and potassium regulate human CYP11B2 transcription through common cis-elements. Mol. Endocrinol. 11, 638-649.

Croft, K.D., 1998. The chemistry and biological effects of flavonoids and phenolic acids. Ann. N. Y. Acad. Sci. 854, 435-442.

Davies, E., Mackenzie, S.M., 2003. Extra-adrenal production of corticosteroids. Clin. Exp. Pharmacol. Physiol. 30, 437-445.

Engh, R.A., Girod, A., Kinzel, V., Huber, R., Bossemeyer, D., 1996. Crystal structures of catalytic subunit of cAMP-dependent protein kinase in complex with isoquinolinesulfonyl protein kinase inhibitors H7, H8, and H89. Structural implications for selectivity. J. Biol. Chem. 271, 26157-26164.

Evans, B.A., Griffiths, K., Morton, M.S., 1995. Inhibition of 5 alpha-reductase in genital skin fibroblasts and prostate tissue by dietary lignans and isoflavonoids. J. Endocrinol. 147, 295-302.

Harris, R.M., Waring, R.H., 2008. Sulfotransferase inhibition: potential impact of diet and environmental chemicals on steroid metabolism and drug detoxification. Curr. Drug Metab. 9, 269-275.

Harvey, P.W., Everett, D.J., 2003. The adrenal cortex and steroidogenesis as cellular and molecular targets for toxicity: critical omissions from regulatory endocrine disrupter screening strategies for human health? J. Appl. Toxicol. 23, 81-87.

Ho, P.C., Saville, D.J., Wanwimolruk, S., 2001. Inhibition of human CYP3A4 activity by grapefruit flavonoids, furanocoumarins and related compounds. J. Pharm. Pharm. Sci. 4, 217-227.

Kaminska, B., Opalka, M., Dusza, L., 2007. Phytoestrogens alter cortisol and androstenedione secretion by porcine adrenocortical cells. Acta Vet. Hung. 55, 359-367.

Kellis Jr., J.T., Vickery, L.E., 1984. Inhibition of human estrogen synthetase (aromatase) by flavones. Science 225, 1032-1034.

Keung, W.M., 1995. Dietary estrogenic isoflavones are potent inhibitors of beta- hydroxysteroid dehydrogenase of P. testosteronii. Biochem. Biophys. Res. Commun. 215, 1137-1144.

Le Bail, J.C., Laroche, T., Marre-Fournier, F., Habrioux, G., 1998. Aromatase and 17beta- hydroxysteroid dehydrogenase inhibition by flavonoids. Cancer Lett. 133, 101-106.

Lin, T.C., Chien, S.C., Hsu, P.C., Li, L.A., 2006. Mechanistic study of polychlorinated biphenyl 126-induced CYP11B1 and CYP11B2 up-regulation. Endocrinology 147, 1536-1544.

Mesiano, S., Katz, S.L., Lee, J.Y., Jaffe, R.B., 1999. Phytoestrogens alter adrenocortical function: genistein and daidzein suppress glucocorticoid and stimulate androgen production by cultured adrenal cortical cells. J. Clin. Endocrinol. Metab. 84, 2443-2448.

Montemayor, C., Montemayor, O.A., Ridgeway, A., Lin, F., Wheeler, D.A., Pletcher, S.D., Pereira, F.A., 2010. Genome-wide analysis of binding sites and direct target genes of the orphan nuclear receptor NR2F1/COUP-TFI. PLoS One 5, e8910.

Moon, Y.J., Wang, X., Morris, M.E., 2006. Dietary flavonoids: effects on xenobiotic and carcinogen metabolism. Toxicol. In Vitro 20, 187-210.

Morohashi, K., Honda, S., Inomata, Y., Handa, H., Omura, T., 1992. A common trans-acting factor, Ad4-binding protein, to the promoters of steroidogenic P-450s. J. Biol. Chem. 267, 17913-17919.

Mortensen, A., Kulling, S.E., Schwartz, H., Rowland, I., Ruefer, C.E., Rimbach, G., Cassidy, A., Magee, P., Millar, J., Hall, W.L., Kramer Birkved, F., Sorensen, I.K., Sontag, G., 2009. Analytical and compositional aspects of isoflavones in food and their biolog- ical effects. Mol. Nutr. Food Res. 53 (Suppl 2), S266-S309.

Ohno, S., Shinoda, S., Toyoshima, S., Nakazawa, H., Makino, T., Nakajin, S., 2002. Effects of flavonoid phytochemicals on cortisol production and on activities of steroido- genic enzymes in human adrenocortical H295R cells. J. Steroid Biochem. Mol. Biol. 80, 355-363.

Ohno, S., Matsumoto, N., Watanabe, M., Nakajin, S., 2004. Flavonoid inhibition of over- expressed human 3beta-hydroxysteroid dehydrogenase type II. J. Steroid Biochem. Mol. Biol. 88, 175-182.

Pfeiffer, E., Graf, E., Gerstner, S., Metzler, M., 2006. Stimulation of estradiol glucuronidation: a protective mechanism against estradiol-mediated carcinogenesis? Mol. Nutr. Food Res. 50, 385-389.

Sanderson, J.T., Hordijk, J., Denison, M.S., Springsteel, M.F., Nantz, M.H., van den Berg, M., 2004. Induction and inhibition of aromatase (CYP19) activity by natural and synthetic flavonoid compounds in H295R human adrenocortical carcinoma cells. Toxicol. Sci. 82, 70-79.

Shibata, H., Kobayashi, S., Kurihara, I., Suda, N., Yokota, K., Murai, A., Ikeda, Y., Saito, I., Rainey, W.E., Saruta, T., 2004. COUP-TF and transcriptional co-regulators in adrenal steroidogenesis. Endocr. Res. 30, 795-801.

Supornsilchai, V., Svechnikov, K., Seidlova-Wuttke, D., Wuttke, W., Soder, O., 2005. Phytoestrogen resveratrol suppresses steroidogenesis by rat adrenocortical cells by inhibiting cytochrome P450 c21-hydroxylase. Horm. Res. 64, 280-286.

Tintos, A., Gesto, M., Miguez, J.M., Soengas, J.L., 2008. beta-Naphthoflavone and benzo (a)pyrene treatment affect liver intermediary metabolism and plasma cortisol levels in rainbow trout Oncorhynchus mykiss. Ecotoxicol. Environ. Saf. 69, 180-186.

Tsou, T.C., Tsai, F.Y., Hsieh, Y.W., Li, L.A., Yeh, S.C., Chang, L.W., 2005. Arsenite induces endothelial cytotoxicity by down-regulation of vascular endothelial nitric oxide synthase. Toxicol. Appl. Pharmacol. 208, 277-284.

Ueng, Y.F., Kuwabara, T., Chun, Y.J., Guengerich, F.P., 1997. Cooperativity in oxidations catalyzed by cytochrome P450 3A4. Biochemistry 36, 370-381.

Walle, U.K., Walle, T., 2007. Bioavailable flavonoids: cytochrome P450-mediated me- tabolism of methoxyflavones. Drug Metab. Dispos. 35, 1985-1989.

Wang, T.T., 2002. Beta-naphthoflavone, an inducer of xenobiotic metabolizing en- zymes, inhibits firefly luciferase activity. Anal. Biochem. 304, 122-126.

Wang, X.L., Bassett, M., Zhang, Y., Yin, S., Clyne, C., White, P.C., Rainey, W.E., 2000. Transcriptional regulation of human 11beta-hydroxylase (hCYP11B1). Endocrinology 141, 3587-3594.