Antiepileptic drugs affect neuronal androgen signaling via a …jpet.aspetjournals.org/content/jpet/early/2007/05/15/jpet.107.120303... · strual disorders, polycystic ovaries, infertility,

JPET #120303

1

Antiepileptic drugs affect neuronal androgen signaling

via a Cytochrome P450 dependent pathway

Marcel Gehlhaus, Nina Schmitt, Benedikt Volk and Ralf P. Meyer

Pathologisches Institut, Abt. Neuropathologie, Neurozentrum,

Universitätsklinik Freiburg, D-79106 Freiburg, Germany

JPET Fast Forward. Published on May 15, 2007 as DOI:10.1124/jpet.107.120303

Copyright 2007 by the American Society for Pharmacology and Experimental Therapeutics.

This article has not been copyedited and formatted. The final version may differ from this version.JPET Fast Forward. Published on May 15, 2007 as DOI: 10.1124/jpet.107.120303

at ASPE

T Journals on June 30, 2020

jpet.aspetjournals.orgD

ownloaded from

http://jpet.aspetjournals.org/

JPET #120303

2

Running title: Brain cytochrome P450 affect endocrine signaling

Address correspondence to:

Ralf P. Meyer, PhD.

Pathologisches Institut, Abt. Neuropathologie,

Neurozentrum, Universität Freiburg,

Breisacherstraße 64, D-79106 Freiburg,

Tel:+49-761-270-5078, Fax:+49-761-270-5050

E-Mail: [email protected]

Number of text pages: 28

Number of tables: 0

Number of figures: 7

Number of references: 37

Number of words in the Abstract: 249

Number of words in the Introduction: 665

Number of words in the Discussion: 1497

Recommended section assignment: Neuropharmacology

Abbreviations

AED- Antiepileptic drugs; ARE- androgen responsive element; CAT- Chloramphenicol-

Acetyltransferase; DPH- diphenyl-hydantoin, phenytoin; FL3a11- Full-length Cytochrome P450 3a11;

GFP- green fluorescent protein; MMTV- mouse mammary tumor virus; NFpan- Neurofilament pan;

P450, CYP- cytochrome P450; PWO- Pyrococcus woesei; siRNA- short interfering RNA; TK-

Thymidin Kinase.


at ASPE



ownloaded from


JPET #120303

3

Abstract

Recent data imply an important role for brain P450 in endocrine signaling. In epileptic patients, treat-

ment with P450 inducers led to reproductive disorders; in mouse hippocampus, phenytoin treatment

caused concomitant upregulation of CYP3A11 and androgen receptor (AR). In the present study, we

established specific in-vitro models to examine if CYP3A isforms are causative for enhanced AR ex-

pression and activation. Murine Hepa1c1c7- and neuronal-type rat PC-12 cells were used to investi-

gate P450 regulation and its effects on AR after phenytoin and phenobarbital administration. In both

cell lines, treatment with AED led to concomitant upregulation of CYP3A (CYP3A11 in Hepa1c1c7,

CYP3A2 in PC-12) and AR mRNA and protein. Inhibiton of CYP3A expression and activity by the

CYP3A-inhibitor ketoconazole or by CYP3A11 specific siRNA molecules reduced AR expression to

basal levels. The initial upregulation of AR signal transduction, measured by an androgen responsive

element CAT reporter gene assay, was completely reversed after specific inhibition of CYP3A11.

Withdrawal of the CYP3A11 substrate testosterone prevented AR activation, while AR mRNA ex-

pression remained upregulated. Additionally, recombinant CYP3A11 was expressed heterologously in

PC-12 cells, thereby eliminating any direct drug influence on AR. Again, the initial upregulation of

AR mRNA and activity was reduced to basal levels after silencing of CYP3A11. In conclusion, we

show here, that CYP3A2 and CYP3A11 are crucial mediators of AR expression and signaling after

AED application. These findings point to an important and novel function of P450 in regulation of

steroid hormones and their receptors in endocrine tissues like liver and brain.


at ASPE



ownloaded from


JPET #120303

4

Introduction

Cytochrome P450 is the collective term for a superfamily of heme-containing proteins that play an

important role in the oxidative phase I metabolism of numerous endogenous (e.g. steroids, bile acids)

and exogenous compounds (drugs, food) (Handschin and Meyer, 2003;Nelson et al., 2004). Although

the main site for P450 dependent drug metabolism is the liver, P450 isoforms also occur in defined

neuronal or glial cell populations of the brain (Volk et al., 1995;Strobel et al., 2001). The function of

P450 in brain is poorly understood and object of strong discussion. The need for understanding brain

P450 function is gaining more and more in importance, since by far the most of the neuroactive drugs

used in therapy of neurological diseases like epilepsy, anxiety disorders or depression interact with

specific P450 isoforms either as substrate or as inducer (Flockhart DA, P450-Drug interaction table.

Available from: http://medicine.iupui.edu/flockhart). These compounds have a lipophilic structure

which facilitates passing the blood brain barrier and penetration into the central nervous system

(Meyer et al., 2004). The cerebral isoforms CYP2B1, CYP2B10, CYP2C29, CYP3A11 and CYP3A13

have been found to be inducible by antiepileptic drugs (AED) like phenytoin or phenobarbital in rats

and mice (Volk et al., 1995;Thuerl et al., 1997;Meyer et al., 2001;Hagemeyer et al., 2003). Therefore,

it is a major attempt in clinic and research to understand the underlying molecular mechanisms of drug

therapy within the central nervous system in order to improve present therapeutic strategies.

Several recent reports indicate that function and regulation of brain P450 is different from liver. It has

been shown that enzymatic activity of brain P450 is dependent on the availability of the prosthetic

group heme (Meyer et al., 2002;Meyer et al., 2005) and that glial and neuronal P450 have different

function (Meyer et al., 2004). Some studies indicate that brain P450 may act in mediation of steroid

hormone signaling (Wang et al., 2000;Hagemeyer et al., 2003;Meyer et al., 2006). We hypothesize,

that brain P450 has functions beyond drug metabolism. In a previous study, we could demonstrate

concomitant upregulation of CYP3A11 and androgen receptor (AR) as well as enhanced testosterone

metabolism and upregulation of CYP19 (aromatase) in hippocampus of phenytoin treated mice (Meyer

et al., 2006). This indicates a connection between P450-inducing AED, decreased sex hormone levels

and altered androgen signaling. These observations are further supported by clinical data from epilep-


at ASPE



ownloaded from


JPET #120303

5

tic patients. Treatment of epilepsy with P450-inducing AED is frequently associated with endocrine

and reproductive disorders in both men and women. These side effects include bad mood states, men-

strual disorders, polycystic ovaries, infertility, impotence, and obesity. They are thought to originate

from reduced testosterone levels, most likely by its enhanced catabolism within the brain (Isojarvi et

al., 2005). Taken together, both the clinical observations and the data from mouse hippocampus imply

a crucial function of P450 in AR regulation within the central nervous system.

In order to clarify the interplay between drug application, subsequent P450 induction, and steroid

hormone signaling in brain, we established an in vitro assay to obtain more specific and detailed in-

formation about function and regulation of P450. Initially, we used mouse hepatic cell line Hepa1c1c7

to investigate common principles of P450 regulation and P450 dependent effects on steroid hormone

receptors. As a neuronal-type system, we selected rat pheochromocytoma PC-12 cells. This cell line is

well established as a neuronal model, since a neuronal phenotype can be induced by application of

nerve growth factor (NGF) (Khvotchev et al., 2003).

In the present study, we demonstrate that in both the hepatic and the neuronal-type cell line, androgen

receptor upregulation as well as enhanced AR-dependent signal transduction observed due to AED

application are mediated by CYP3A2 or CYP3A11. This was verified by the use of CYP3A11-specific

short interfering RNA (siRNA) molecules and the CYP3A inhibitor ketoconazole. We conclude, that

CYP3A isoforms themselves are main regulators of AR expression and downstream signaling. This is

a novel function for CYP3A which demonstrates the relevance of understanding the action of P450 in

steroid hormone sensitive areas, especially in the brain.


at ASPE



ownloaded from


JPET #120303

6

Methods

Chemicals. The antiepileptic drugs phenytoin and phenobarbital, the CYP3A inhibitor ketoconazole,

and the steroid hormone testosterone were purchased from Sigma-Aldrich (Deisenhofen, Germany).

The polyclonal anti-rabbit androgen receptor antibodies were from Santa Cruz Biotechnology (N-20,

sc-816, Santa Cruz, CA, USA). The CYP3A1 antibody (polyclonal anti rat CYP3A1, CR3310) was

purchased from Biotrend (Köln, Germany), monoclonal mouse anti-NFpan from Zymed Laboratories

Inc (San Francisco, USA). Gel purification- and DNA preparation kits, the customized siRNA nucleo-

tides, and the RNeasy Kit were from Qiagen (Hilden, Germany). Reverse Transcription was performed

using the Avian Myeloblastosis Virus Reverse Transcriptase (Promega, Mannheim, Germany). Q-PCR

ROX mastermix and the Q-PCR plastic ware were purchased from ABGene (Hamburg, Germany). All

PCR and Q-PCR primers were synthesized by in-house core facility (Gabor Igloi, University Freiburg,

Germany). The Pyrococcus woesei (PWO) proofreading DNA polymerase, the PCR nucleotide mix,

the CAT-Elisa- and the cell cytotoxicity kits were from Roche Diagnostics (Mannheim, Germany).

For all transfection experiments, we used the Nucleofector Cell Suspension R (Hepa1c1c7) and V

(PC-12) (Amaxa biosystems, Köln, Germany).

The pcDNA1.1Amp expression vector was purchased from Invitrogen BV (Groningen, The Nether-

lands), the MMTV-ARE-CAT and ARE-TK-CAT reporter constructs were a generous gift from R.

Schüle, University Hospital Freiburg, Germany, and described previously (Schule et al., 1990). All

other reagents were from commercial sources at the highest purity available.

Cloning of CYP3A11 full length and CYP3A11-GFP-fusion constructs. Specific primers for gen-

eration of mouse CYP3A11 full length cDNA were designed using the DNA-Star software package

(GATC Biotech, Konstanz, Germany), based on NCBI nucleotide database entry NM_007818 as a

template. The CYP311 forward primer (sense, including NotI site) corresponded to nucleotides 1-21

(5`-CAGCGGCCGCGAGGGAAGCATTGAGGAGGAT-3`), the reverse primer (antisense, including

XbaI site) corresponded to nucleotides 1697-1720 (5`-

CATCTAGAACCAGGCATCAAAACCAATCTATT-3`). Additionally, for nested PCR, the forward

primer corresponded to nucleotides 291-310 (5`-ATGGGGGTTGTTTGATGG TC-3`) and the reverse

primer to nucleotides 841-859 (5`-GGCGGCTTTCCTTCATTCT-3`). Total RNA from mouse hippo-


at ASPE



ownloaded from


JPET #120303

7

campus was isolated using RNeasy Kit (Qiagen, Hilden, Germany). One microgram total RNA was

reverse transcribed at 42°C for 1h. The full length cDNA was amplified by long-template PCR. Sepa-

rated PCR products were extracted from the agarose gel and purified with QIAGEN Qiaquick Gel

Extraction Kit according to the manufacturer’s protocol. A 1730bp NotI- XbaI fragment of the full

length CYP3A11 cDNA was cloned in sense orientation into pcDNA1.1_Amp expression vector

(Invitrogen BV, Groningen, Netherlands) using Rapid Ligation Kit 1 (Roche).

The CYP3A11-GFP fusion construct was used as a positive control for the RNAi- experiments. A full

length CYP3A11 cDNA, missing the stop codon, was generated from mouse brain cDNA and ligated in

frame into the pEGFP-N1 vector (BD Biosciences, Heidelberg, Germany). Hence, we obtained a

CYP3A11-GFP fusion protein with the P450 fused at the N-Terminus of GFP. The CYP3A11 forward

primer (sense strand), containing an additional Xho I cutting site, corresponded to nucleotides 4 to 21 (5’-

CAC CTCGAG GGAAGCATTGAGGAGGAT -3’), and the CYP3A11 reverse primer, with a Sac II cut-

ting site, to nucleotides 1572 to 1590 (5’- CAC CCGCGG TGCAGTCATAACTGGAGCA -3’) (based on

CYP3A11 GenBank Acc-No: nm_007818). Long-template PCR was performed using PWO-proofreading

polymerase. The Full Length CYP3A11 and CYP3A11-GFP plasmid were transformed into Escherichia

coli strain XL-1 blue and prepared using Qiagen Plasmid Midi kit (Qiagen, Hilden Germany). Finally, both

plasmids were checked by nested PCR and cycle sequencing using the MEGABACE 500 system (Amer-

sham) with the ABI BigDye 2.0 Mix and documented accurate sequences.

Cell culturing conditions. The mouse Hepatoma Hepa1c1c7 cells were grown as a monolayer in

Dulbecco’s modified Eagle medium (with GlutaMAX, supplemented with 4500 mg/ml glucose) con-

taining 10% (v/v) fetal bovine serum and 1% penicillin-streptomycin solution (Invitrogen, Karlsruhe,

Germany). Rat pheochromocytoma cells PC-12 (DSZM, Germany) were grown in RPMI 1640 Glu-

tamax (Invitrogen) supplemented with 10% horse serum (Invitrogen)/ 5% fetal bovine serum (Invitro-

gen) and 1% penicillin-streptomycin (Invitrogen) on collagen typ IV-coated (25 µg/ml) cell culture

flasks (Falcon). Cultures were maintained in a humidified incubator at 37°C with 5% CO2. Cells were

sub-cultured every 3-5 days with Accutase© (PAA Laboratories, Coelbe, Germany). Prior to experi-

ments, PC-12 cells were supplemented with 0.1 µg/ml Nerve Growth Factor (NGF, Sigma-Aldrich)

for neuronal differentiation.


at ASPE



ownloaded from


JPET #120303

8

Treatment of cell lines with antiepileptic drugs. The antiepileptic drugs phenytoin and phenobarbi-

tal were dissolved in DMSO and added diluted to a final concentration of 10 µM and 100 µM for 24 or

48h. Ketoconazole was used at 10 µM concentration. These drug concentrations were proved to be

non-toxic for the cells by a cytotoxicity assay (Cytotoxicity Detection Kit, Roche Diagnostics, Mann-

heim, Germany), based on lactate dehydrogenase release, as follows: 2.5 x 104 cells per well in a 96-

well plate were treated with 100 nM, 1 µM, 10 µM, 100 µM and 1 mM of the respective drug. After

48 h of treatment, cell viability was measured spectrophotometrically at 490 nm by examining the

formation of formazan resulting from INT bioreduction (data not shown). The steroid hormone testos-

terone, dissolved in DMSO, was added diluted to a final concentration of 100 nM in order to provide

sufficient concentrations of P450 substrate. By omitting testosterone, its influence on AR transcription

and activation was illustrated. All drug and hormone concentrations used have been validated previ-

ously to resemble therapeutic conditions (Thuerl et al., 1997;Tauer et al., 1998).

Immunoblot against CYP3A11 and AR. 106 cells from Hepa1c1c7 or PC-12 cells were homoge-

nized by sonication (homogenate). Subfractionation was carried out by differential centrifugation as

described earlier (Meyer et al., 2002). In order to achieve microsomes (mc) the cell homogenates were

centrifuged at 9000g. The 9000g supernatant (9S) was directly applied to ultracentrifugation at

150,000g for 30min at 4°C in a Beckman benchtop ultracentrifuge (Beckman, Krefeld, Germany). The

resulting pellet was designated as microsomes (mc). The 9000g pellet (9P) contains the cell nuclei and

was used to investigate AR localization in Hepa1c1c7 cells. Denatured protein was resolved on 9%

sodium dodecylsulfate – polyacrylamid gels. As a reference, mouse liver microsomes were used. Pro-

teins were transferred to polyvinylidene difluoride membranes (Immobilon P; Millipore, Schwalbach,

Germany) in buffer (125 mM Tris, 960 mM glycine). Incubations were performed with a CYP3A

polyclonal antibody (dilution 1: 2000) followed by exposure to horseradish peroxidase-conjugated IgG

goat anti-rabbit (Jackson Immunoresearch Lab. Inc.) at a dilution of 1: 10000. The specificity of the

antibodies for the antigen was proved previously (Hagemeyer et al., 2003). AR detection was per-

formed with an AR polyclonal antibody (dilution 1: 500) followed by exposure to the horseradish

peroxidase-conjugated IgG at a dilution of 1: 5000. The immunopositive bands were visualized with

enhanced chemoluminescence (ECL, Amersham Biosciences, Freiburg, Germany). The intensity of


at ASPE



ownloaded from


JPET #120303

9

the immunostained bands was evaluated by scanning densitometry using Tina 2.10 software (Raytest,

Straubenhardt, Germany).

Fluorescence double-labeling immunocytochemistry. Cells were grown on Thermanox-slides

(Nunc, Wiesbaden, Germany) and fixed in paraformaldehyde (PFA) (4% v/v) for 30 min at room tem-

perature. Unspecific binding sites were blocked by PBS / 0.3% Triton X100 (v/v) / 1 % BSA. The

cells were incubated in a cocktail of primary antibodies (rabbit anti-AR, dilution 1 : 200/ mouse anti-

NFpan, dilution 1 : 6) in PBS / 0.3% Triton X100 (v/v) / 1 % BSA at 4° C overnight. Afterwards, the

binding sites of the primary antibodies were labeled by fluorescence coupled secondary antibodies

applied for 2 h at 4°C (GAR-RRX, Jackson Immunoresearch Laboratories; GAM-Alexa 488, Molecu-

lar Probes). Samples were washed again and cover-slipped in Elvanol [20% polyvinyl alcohol (Cal-

biochem, Darmstadt, Germany), 1% DABCO (Aldrich, Steinheim, Germany) in 0.2 M Tris-HCl, pH

8.5]. Controls were performed by omitting one or both primary antibodies. Thus, there was specific

staining for only one of the antigens or, in the latter case, there was no specific staining at all. Fluores-

cent signals of 2 µm thick optical sections were examined by using a confocal laser scanning micro-

scope (Leica TCS NT, Heidelberg, Germany).

Immunocytochemistry. 2.5 x 105 cells used for immunocytochemistry were grown on Thermanox

coverslips (Nunc, Wiesbaden, Germany). 48 h after drug application, cells were fixed for 30 min in

4% paraformaldehyde in sodium phosphate buffer (pH 7.4). For immunostaining, we followed the

instructions for the Zymed kit (Invitrogen, Heidelberg, Germany). The antibody against AR was used

at a dilution of 1:400. All samples were incubated overnight at 4°C. The specificity of the antibody has

been demonstrated previously (Meyer et al., 2006). Negative controls were treated with only the sec-

ondary antibodies and proved to exert no immunosignal. Chromogen staining using 3,3’-

diaminobenzidine (DAB) was performed for 15 min at room temperature. Samples were washed again

and cover-slipped in Elvanol as described previously.

Transfection of cell lines with ARE-CAT constructs and full length CYP3A11. An ARE-CAT

reporter construct was used to report androgen receptor activation. In case of Hepa1c1c7, 106 cells

were pretreated for 48 h with 10 µM of the designated antiepileptic drug and 100nM testosterone on

6well dishes. Afterwards, 2 µg of MMTV-ARE-CAT construct was transfected into the cells by elec-


at ASPE



ownloaded from


JPET #120303

10

troporation using a specific Nucleofector ® solution (Cell Line Nucleofector® Kit R, VCA-1001) as

described below. The transfection efficiency in Hepa 1c1c7 cells achieved 75-80 %.

PC-12 cells were transfected either with the Full Length CYP3A11 plasmid (FL3a11) alone for quanti-

tative real time RT-PCR investigations or in combination with an ARE-CAT construct for CAT-

reporter gene assay (CAT-ELISA). Each transfection was performed using 106 cells, a specific Nu-

cleofector® solution (Cell Line Nucleofector® Kit V, VCA-1003), 0.5-1µg FL3a11 or 1 µg empty

pcDNA1.1Amp vector as a negative control. In case of the transactivation assay, we added addition-

ally 2µg ARE-TK-CAT reporter construct. Transfection efficiency for PC-12 was about 35-40%,

which is in accordance with the manufacturer’s cell line specific protocol (Amaxa). After electropora-

tion, PC-12 cells were plated onto collagen coated 6well dishes supplemented with fresh media and

analyzed after 5-8h by real time RT-PCR or after 24-48h by CAT-ELISA, respectively.

RNA-interference. Three different duplex siRNAs with TT-ends, located on different exons, against

both CYP3A11 and GAPDH were designed using a commercial siRNA design tool algorithm (Am-

bion, Austin, TX). GAPDH siRNA was used as a positive control. The selected siRNA duplexes were

cyp3a11_01 (5´ AAGGGCAGCATTGATCCTTAT 3´) (Exon 7), cyp3a11_02 (5´

AATTCGACATGGAGTGCTATA 3´) (Exon 2), and cyp3a11_03 (5´ AACAACCCAGAG-

GATCCTTTT 3´) (Exon 3) for CYP3A11, GAPDH_01 (5´ AACTACATGGTCTACATGTTC 3´),

GAPDH_02 (5 ́ AACCACGAGAAATATGACAAC 3´) and GAPDH_03 (5´ AAGTATGAT-

GACATCAAGAAG 3´) for GAPDH and as non silencing control a scrambled anti-sense siRNA (5´

AATTCTCCGAACGTGTCACGT 3´). All siRNA nucleotides were compared with mouse and rat

genome database (NCBI nblast) and demonstrated exclusive specificity against the intended target

genes. Each siRNA was used in a final concentration of 50 nM. Initially, CYP3A11 siRNA efficiency

was tested by cotransfection with 1µg CYP3A11-GFP fusion plasmid in Hepa1c1c7. As expected, the

silencing targeted against CYP3A11 resulted in a nearly complete loss of fluorescence of the

CYP3A11-GFP fusion protein (data not shown).

In case of Hepa1c1c7, 106 cells were treated for 48 h with 100 µM of the designated antiepileptic drug

before being transfected with 50 nM CYP3A11-, GAPDH- or non silencing- siRNA by electroporation

technique using the Nucleofector device (Nucleofector program T-20). After additional 48 h, mRNA

expression of AR, CYP3A11 and GAPDH was analyzed by quantitative real time RT-PCR, while the


at ASPE



ownloaded from


JPET #120303

11

resulting AR-dependent signaling was checked by CAT-reporter gene assay (see below). In case of PC-

12, the CYP3A11 siRNA nucleotides were cotransfected with those vectors used in either the

CYP3A11 overexpression or the transactivation assays.

RNA isolation and real-time PCR. Total RNA was isolated using RNeasy Kit (Qiagen, Hilden,

Germany). Two microgram total RNA was reverse-transcribed as described above. Real-time PCR

was performed on an ABI PRISM™ 7700 (Taqman) using the sequence detector software, version 1.9

(Applied Biosystems, Darmstadt, Germany). Taqman primers and probes for cyp3a11 (NM 007818),

CYP3A2 (U 09742), mouse AR (GenBank Acc: X53779), rat AR (NM 012052.1), and GAPDH (NM

008084) were designed with the help of the Primer Express software, version 1.5 (Applied Biosys-

tems). Taqman probes coupled to a 5`- fluorophore (FAM) and a 3`-quencher (TAMRA) were manu-

factured by Applied Biosystems. 15 min of initial denaturation at 95 °C were followed by a 40 cycle

two step PCR with each 30 sec annealing / elongation at 60 °C and 30 sec denaturation at 95 °C.

GAPDH was chosen as housekeeping gene, because neither testosterone nor the antiepileptic drugs

used are known to regulate GAPDH in brain (Joe and Ramirez, 2001). Primer and probe concentra-

tions were optimized as follows: mAR forward: 5`-TTAACGTCCTGGAAGCCATTG -3` (900 nM);

mAR reverse: 5`-CAAAGGAATCTGGTTGGTTGTTG-3` (900 nM); rAR forward: 5`-

CTTTCTTAATGT-CCTGGAAGCCAT-3`(900 nM), rAR reverse: 5`-

AAGGAATCAGGCTGGTTGT TGT-3`(900 nM) and AR-probe: 5`-FAM-

CCAGGAGTGGTGTGTGCCGGACAT-TAMRA-3` (100 nM) (located on r+mAR gene, exon 4).

mCyp3a11 forward: 5`-CCCAAAGGGTCAACAGTGATG-3` (900 nM); mCyp3a11 reverse: 5`-

CTTCAGGCTCTGACCAGTGCT-3` (900 nM); and mCyp3a11 probe: 5`-FAM-

TTCCATCTTATGCTCTTCACCATGACCA-TAMRA-3` (100 nM) (located on cyp3a11 gene, exon

6). rCyp3a2 forward: 5’-AAACCACCAGCAGCACACTCT-3’ (900 nM); rCyp3a2 reverse: 5’-

TCCTCCTGCAGTTTCTTCTGAAT-3’ (900 nM); and rCyp3a2 probe: 5’-TCTTGTATTTC-

CTGGCCACTCACCCTGA-3’ (100 nM) (located on cyp3a2 gene, exon 6). Levels of the GAPDH

housekeeping gene were determined for internal normalization using GAPDH forward: 5`-

CCAGAACATCATCCCTGCATC-3`; GAPDH reverse: 5`-GGTCCTCAGTGTAGCCCAAGAT-3`;

and GAPDH probe: 5`-FAM-CCGCCTGGAGAAACCTGCCAAGTATG-TAMRA-3` (located on


at ASPE



ownloaded from


JPET #120303

12

GAPDH gene, exons 5-6). Results were demonstrated as ratio of the quantity of the target gene versus

expression of the GAPDH housekeeping gene and finally normalized against the values of vehicle or

untreated controls. All primers and probes were compared with the mouse and rat genome database

(NCBI nblast) and demonstrated exclusive specificity against the intended target genes. Additionally,

all primers were checked in a standard RT-PCR reaction with cDNA of both cell lines prior to Taqman

real-time PCR, and demonstrated single bands at the expected nucleotide size in ethidiumbromide

staining (data not shown). This further allowed exclusion of genomic DNA contamination, as the

GAPDH primer set enables exon spanning.

Endogenous AR expression in both cell lines was evidenced by non quantitative RT-PCR using mouse

AR (mAR) and rat AR (rAR) selective Taqman primer pairs. The PCR products obtained had the ex-

pected size at 71bp/ 74bp length and were separated by electrophoresis using 3.5% MetaPhor© high

resolution agarose (Cambrex Bio Science). GeneRuler 100bp DNA ladder (Fermentas) and DNA

Molecular Marker X (Roche diagnostics, Mannheim, Germany) were used as markers.

CAT reporter gene assay (CAT-ELISA). Androgen receptor dependent CAT expression in cell lys-

ates was measured using the CAT-ELISA kit (Roche diagnostics, Mannheim, Germany) according to

the manufacturer’s instruction manual. Total protein content was determined using Bradford protein

assay for normalization of specific CAT expression. Experiments were confirmed two times at least,

each conducted in three sample wells of 75 µg (Hepa1c1c7) or 100-150 µg (PC-12) total protein, re-

spectively, supplemented with the provided substrate enhancer in an Elisa Reader at 410 and 490 nm.

Resulting CAT levels were normalized against the untreated cells transfected with non silencing

siRNA or against vector controls, respectively.

Statistical analysis. Each experiment was confirmed two times at least and measured in triplicate. The

interassay coefficients of variation are indicated in the respective figure legends and are generally less

than 10 %. Data were demonstrated as means ± SD of a representative experiment for each section.

For statistical analysis we used unpaired two-group t-test (two-tailed t-test) or, in case of multiple

comparisons, an ANOVA with a post-hoc correction (Bonferroni). Data processing was performed

using MS-Excel 2003 and SPSS 11.0.1 software (SAS Institute, Cary, NC, USA). In case of transient

transfections, data were normalized against the transfection efficiencies. The accepted significance

level was p < 0.05, at least.


at ASPE



ownloaded from


JPET #120303

13

Results

Basal characterization of Hepa1c1c7 and PC-12 cells. We initially proved endogenous expression

of the target genes AR and CYP3A11 in both cell lines, in order to estimate the changes evoked by

AED treatment or CYP3A11 overexpression, respectively. In murine hepatoma cell line Hepa1c1c7,

we found basal expression of AR mRNA in untreated cells using quantitative real time PCR (Taqman)

(Fig. 1a). In addition, CYP3A11 was expressed to considerable amount, as shown by immunoblot

(Fig. 1b). PC-12 cells originate from rat pheochromocytoma cells. In these cells, CYP3A2 is the main

CYP3A isoform expressed (data not shown). As expected, we could not find any murine specific

CYP3A11 expression in these cells (data not shown). Therefore, PC-12 cells revealed to be highly

suitable for heterologous overexpression of murine CYP3A11 and studying its effects. On the other

hand, we could detect AR mRNA as a slight but distinct band at 75bp in PC-12 cells, demonstrating

weak endogenous AR expression (Fig. 2a). A neuronal phenotype was induced in PC-12 due to appli-

cation of Nerve Growth Factor (NGF) (Tischler and Greene, 1975), validated by expression of NF-

pan. As displayed by fluorescence double-labeling immunocytochemistry, NF-pan is clearly co-

expressed with endogenous AR in NGF treated PC-12 cells (Fig. 2b).

Induction model: Treatment of hepatic and neuronal cells with P450-inducing drugs

We applied the commonly used CYP3A-inducers phenytoin and phenobarbital as equivalents in both

cell lines (Volk et al., 1995).

Regulation of CYP3A11 and AR mRNA expression in Hepa1c1c7 cells. Hepa1c1c7 cells were gener-

ally used as a model for P450 investigations since these cells are known to be very good inducible by

AED (Karenlampi et al., 1989). We therefore used this cell line as a model to investigate general rela-

tions between AR, P450 and AED treatment.

Application of phenytoin (DPH) to Hepa1c1c7 cells resulted in an orchestrated, congruent upregula-

tion of CYP3A11 and AR mRNA, as analyzed by quantitative real time PCR (Fig. 3 a, b). We found

7-fold induction of CYP3A11 and a 12-fold of AR. According to this observation, we analyzed if spe-

cific silencing of CYP3A11 also affects AR mRNA regulation. After application of specific CYP3A11


at ASPE



ownloaded from


JPET #120303

14

siRNA nucleotides, both target genes were downregulated (Fig. 3 a, b). We found an effective silenc-

ing of CYP3A11 (80%) as well as AR downregulation (75%) in phenytoin treated cells using specific

CYP3A11 siRNA.

Activation of AR signaling by P450- inducing drugs in Hepa1c1c7 cells. After having proved the cor-

relation between AED dependent CYP3A11 induction and AR mRNA upregulation, we next analyzed

downstream effects of AR activation. Ligand activated androgen receptors bind to specific androgen

responsive elements at the 5`flanking region of AR target genes, leading to initiation of gene transcrip-

tion. Using a MMTV-ARE-CAT reporter, we correlated AED dependent androgen receptor activation

with CAT expression, which was determined by CAT ELISA.

After treating the cells with 100µM phenytoin in presence of non silencing siRNA, we found CAT

expression upregulated to about 40% as compared to untreated Hepa1c1c7 cells, indicating enhanced

receptor activation (Fig. 3c). Transfection of the cells with CYP3A11 siRNA reversed this activation

completely. We found CAT expression of the phenytoin treated cells in the same range as the un-

treated non silencing controls. Untreated cells also revealed significant reduction of receptor activation

to about 40% as a result of CYP3A11 silencing (Fig. 3c).

Regulation of AR protein expression in Hepa1c1c7-cells after treatment with the CYP3A-inhibitor

ketoconazole. AR protein expression was found increased about 3.5-fold after treatment with pheno-

barbital, as compared with the untreated cells (Fig. 4a). In combination with CYP3A-inhibitor keto-

conazole, this AED-dependent AR protein upregulation was completely reversed to basal levels, as

evaluated by immunoblot (Fig 4a). These findings were corroborated by immunocytochemistry, as the

cells treated with phenytoin and phenobarbital clearly showed higher AR-staining intensity (Fig. 4b).

Regulation of CYP3A2, AR mRNA and AR protein expression in PC-12 cells. Next, we transferred this

induction model to neuronal cells. Therefore, PC-12 cells were treated with phenytoin and phenobarbi-

tal. Furthermore, in order to emphasize the role of CYP3A-isoforms in regulating AR expression,

these inducers were used in combination with the chemical CYP3A-inhibitor ketoconazole.

Application of phenytoin and phenobarbital to PC-12 cells resulted in a significant induction of en-

dogenous CYP3A2, which was initially shown to be the most prominent CYP3A-isofom in PC-12

cells. Phenobarbital increased CYP3A2 mRNA expression about 25-fold and Phenytoin 12-fold, as

analyzed by quantitative real time PCR (Fig. 5 a). Furthermore, Phenobarbital treatment doubled AR


at ASPE



ownloaded from


JPET #120303

15

mRNA expression compared to untreated PC-12 cells. This effect was completely reversed in case of

combination of inducer with the inhibitor ketoconazole (Fig. 5b). Finally, similar effects were ob-

served for AR protein expression. Phenytoin treatment increased AR protein levels 2.5-fold and phe-

nobarbital even 3.5-fold, as compared to untreated PC-12 cells. In both cases, AR protein expression

decreased to basal levels, if inducers were used in combination with ketoconazole (Fig. 5c).

Overexpression model: Heterologous expression CYP3A11 in neuronal-type PC-12 cells

Regulation of AR mRNA expression by CYP3A11. As mentioned above, rat PC-12 cells do not express

murine CYP3A11 endogenously, but exhibit weak basal AR expression (Fig. 2). Therefore, PC-12

cells revealed to be a good neuronal-type model to investigate whether the observed effects on AR can

be related to the drug or to CYP3A11. To test our hypothesis of CYP3A11 involvement, we tran-

siently transfected a FL3a11 plasmid into PC-12 and performed similar experiments as done with

Hepa1c1c7.

We observed upregulation of both CYP3A11 and AR mRNA (Fig. 6a, b). Overexpression of 1µg

FL3a11 resulted in about 4-fold enhanced AR mRNA expression compared with the vector controls in

presence of the non silencing siRNA (Fig. 6b). Next, we analyzed if specific silencing of CYP3A11

also affects AR mRNA regulation in PC-12. Co-transfection of 1µg FL3a11 with the specific

CYP3A11 siRNA resulted in downregulation of both target genes. CYP3A11 mRNA expression was

inhibited to approximately 75% compared to the non silencing control samples (Fig.6a). The expres-

sion of AR mRNA was found decreased almost to the same level as the vector controls after applica-

tion of CYP3A11 siRNA (Fig. 6b).

Regulation of AR signaling by CYP3A11. We investigated whether CYP3A11 overexpression leads to

increased receptor activation and signaling. Hence, we established a transactivation assay including

both the FL3a11 overexpression vector and an ARE-CAT reporter.

We found about 3.5-fold enhanced androgen receptor activity due to CYP3A11 overexpression in

presence of non silencing siRNA, as compared with the vector only samples. Addition of specific

CYP3A11 siRNA to this system resulted in decreased expression of CAT to levels of the vector con-

trol (Fig. 6c). These results of the CAT- ELISA assay are concordant with our findings achieved by

quantitative real time PCR.


at ASPE



ownloaded from


JPET #120303

16

Influence of testosterone withdrawal on AR expression and activation

In order to clarify the role of testosterone itself on AR expression and activation both methodical pro-

cedures, quantitative real time PCR and reportergene assay, were performed in Hepa1c1c7 cells as

mentioned before but by omitting the steroid hormone. In absence of testosterone, application of

100µM phenytoin (DPH) again resulted in congruent upregulation of CYP3A11 (80%) and AR

mRNA (50%) (Fig. 7a), but failed to enhance AR activation as measured by CAT- ELISA (Fig. 7b).


at ASPE



ownloaded from


JPET #120303

17

Discussion

The present study demonstrates that the P450 isoforms CYP3A11 (mouse) and CYP3A2 (rat) are cru-

cial for alterations in AR-dependent endocrine signaling in hepatic and neuronal cells after treatment

with P450 inducing antiepileptic drugs like phenytoin and phenobarbital. These endocrine alterations

are not due to effects of the drug itself or its putative metabolites acting as potential AR-ligands. AR

expression and signaling were clearly reduced when CYP3A2 / 11 expression have been inhibited

either by the use of specific siRNA molecules, or by the common chemical CYP3A-inhibitor keto-

conazole. We established two model systems to validate our concept. In the induction model using

Hepa1c1c7- or PC-12 cells, the P450 isoforms CYP3A11 or CYP3A2 have been induced by AED

application. In the overexpression model, recombinant CYP3A11 was applied to PC-12 cells. Our

results indicate that regulation of AR-dependent pathways is a major function of P450 especially in

brain. This is important since there is ongoing discussion on the function of brain P450 in drug re-

sponse.

The use of cell line models facilitates investigating the coherence between drug application, P450

function and AR regulation at the molecular level. We initially started with drug-dependent induction

of CYP3A11 in Hepa1c1c7. This cell line is known to be very good inducible by AED (Karenlampi et

al., 1989) and expresses CYP3A11 as well as AR constitutively. We decided to use this cell line to

investigate common principles of P450 regulation and its effects on steroid hormone receptor regula-

tion. Using this so called “induction model”, we demonstrate, that CYP3A11, after induction by AED,

modulates AR expression and signaling in a congruent manner. If CYP3A11 expression is silenced, or

if it is chemically inhibited by ketoconazole, AR expression and signaling decrease. These findings

point to a general function of P450 in regulation of steroid hormones and their receptors in endocrine

tissues.

Brain P450 is regulated differently from hepatic isoforms. This has to be taken in consideration when

applying this model to brain. Both in vitro and mouse in vivo models have revealed that the availabil-

ity of heme is a limiting step in brain P450 function (Meyer et al., 2002;Meyer et al., 2005). This indi-

cates a fundamental difference between brain P450 and that in liver, where a constitutive heme pool is


at ASPE



ownloaded from


JPET #120303

18

provided (Giger and Meyer, 1983). Nevertheless, previous studies in rodents encourage the concept

that brain P450s have a prominent function in steroid hormone regulation (Rosenbrock et al., 1999;

Wang et al., 2000; Meyer et al., 2006).

In order to validate this concept on molecular levels, we established a neuronal cell line model using

PC-12 cells. These cells exert a neuronal phenotype by application of NGF, and are therefore com-

monly used as a neuronal model (Khvotchev et al., 2003). This was corroborated by detection of the

neurofilament subtypes (60, 160 and 200 kDa, called NFpan) in our cell line system. Furthermore, we

could demonstrate constitutive expression of AR mRNA and protein in PC-12 by RT-PCR using spe-

cific rAR-primers and by coexpression of AR protein with NFpan. According to several other reports,

constitutive AR expression in PC-12 is very weak and hardly to detect (Lustig et al., 1994;Nguyen et

al., 2005). Nevertheless, if P450 was either induced by AED (CYP3A2) or expressed heterologously

(CYP3A11), AR was expressed at evident levels. We conclude that significant AR expression in PC-

12 cells is dependent on sufficient P450 presence. However, compared to Hepa1c1c7, higher expres-

sion levels of CYP3A2 or 11 are obviously necessary in PC-12 cells to activate AR. This is corrobo-

rated by the findings that chemical inhibition of CYP3A2 by ketoconazole or silencing of recombinant

CYP3A11 both led to downregulation of AR expression and signaling in PC-12 cells. Ketoconazole is

originally known as potent inhibitor of CYP3A activity, while recent reports also claim transcriptional

inhibition potency (Huang et al., 2007).

The PC-12 assay has a further advantage. It allows heterologous expression of murine CYP3A11 on a

rat background. Thereby, we can mimic P450 induction in a drug-independent model excluding direct

drug-mediated effects on AR regulation. This approach clearly demonstrates that AR expression and

signaling are regulated in the presence of CYP3A11 and not by a possible interaction of the drug with

the receptor. These findings are substantiated by other studies investigating AED action on reproduc-

tive function. An in-vivo study observed phenobarbital dependent effects on reproductive function of

rats accompanied with altered sex hormone levels (Gupta et al., 1980). Other studies investigating

direct interactions of AED with AR revealed that the P450 inducer carbamazepine and the non inducer

valproate did not show any direct AR agonist activity (Death et al., 2005). Actually, valproate exerted

an antagonistic activity on AR that is consistent with endocrine side-effects of valproate observed by

others (Nelson-DeGrave et al., 2004).


at ASPE



ownloaded from


JPET #120303

19

We hypothesize that the signal transduction between CYP3A and AR after drug treatment proceeds by

a mechanism leading to a compensatory activation of AR in case of enhanced P450-dependent andro-

gen metabolism. We could demonstrate that withdrawal of the CYP3A11 substrate testosterone in

phenytoin-treated cells prevents AR activation, while AR expression remains upregulated, although

less prominent than in presence of testosterone. Studies from brain and thyroid gland demonstrate that

testosterone is important for AR autoregulation (Lu et al., 1999; Banu et al., 2002). We assume that

testosterone has to be present as substrate or ligand in order to enhance and activate the P450-AR sys-

tem. This allows deeper mechanistic insight in the role of CYP3A11 in AR activation. It seems that

CYP3A11 itself has influence on AR transcription, while AR activation revealed to be dependent on

the availability of both CYP3A11 and its substrate testosterone. Previous studies demonstrate that

conversion of testosterone to its hydroxylated metabolites is substantially increased in brain as a con-

sequence of phenytoin administration (Rosenbrock et al., 1999; Meyer et al., 2006). Additionally, this

CYP3A11- dependent depletion of testosterone in mouse hippocampus is accompanied by increased

AR expression (Meyer et al., 2006). The hormone-sensitive cells might get a signal to compensate

reduced hormone levels by receptor upregulation in order to maintain overall hormonal influence on

cellular procedures. This ligand metabolism based hypothesis is corroborated by several in-vitro- and

in-vivo studies, which have revealed similar effects about androgen dependent AR regulation

(Krongrad et al., 1991; Quarmby et al., 1990). In addition, also estrogen receptor α was upregulated in

compensatory manner due to estrogen depletion (Kim et al., 2004; Agarwal et al., 2000).

A putative enzyme-receptor crosstalk appears plausible to link drug mediated CYP3A11 induction

with AR signaling. It is known from drug inducible liver CYP3A-isoforms that enzyme-receptor

crosstalk leads to diversed signaling (Handschin and Meyer, 2003). So, after phenobarbital induction,

the CYP3A regulating nuclear receptor CAR was shown to be related to AMPK- activation (Rencurel

et al., 2006). AMPK, as the main enzyme regulating lipolysis, also exerts crosstalk with PPARα, an-

other nuclear receptor active in lipid and energy homeostasis. Therefore, several CYP3A- inducing

drugs are effective in regulation of cellular energy turnover. We could demonstrate that CYP3A2 and

CYP3A11 are crucial and effective mediators of AR signaling. This highlights a novel function for

P450s and it expands our knowledge on P450 function especially in steroid hormone sensitive areas.


at ASPE



ownloaded from


JPET #120303

20

The role of CYP3A11 in steroid hormone receptor regulation after drug induction needs further inves-

tigation, which would go far beyond the scope of the present study.

Our findings may have strong impact in the pathology of neurological diseases. Clinical and animal

studies imply that P450s regulating steroid metabolism can influence neurological functions, since

neuroactive steroids are associated with memory, behavior, mood, neuroprotection, aging, and neuro-

transmission. In addition, steroid hormones were attributed to play an important role in brain devel-

opment, function, and plasticity (see review: Melcangi and Panzica, 2006). Therefore it appears con-

clusive, that changes in hormonal signal transduction lead to severe reproductive and endocrine as

well as mental disorders. This is substantiated by recent clinical data implying complex interactions

between P450-inducing AED and occurring reproductive and endocrine disorders (Isojarvi et al.,

2005;Herzog and Fowler, 2005). Epileptic patients receiving the classical P450 inducers demonstrate

unusual high occurrence of these disorders as compared with those treated with the non inducing

AEDs.

The relation between P450 induction, altered steroid hormone metabolism and AR-signaling may be

of particular interest for research, clinic and drug design, since 70-80% of all clinically used drugs

interact with P450 (Ingelman-Sundberg, 2004). This may influence drug design or drug use, as such

relations can lead to adverse drug effects with dramatic costs (100 billion US$ in the US) (Ingelman-

Sundberg, 2004), or to neuroprotective effects (Nguyen, Yao, and Pike, 2005). The present study

clearly demonstrates that the role of P450 in brain, or in general, is not restricted to hormonal clear-

ance and degradation. The enhanced expression of P450s after drug treatment obviously leads to acti-

vation of AR and to altered steroid hormone signaling with more or less severe consequences for the

patient.

In conclusion, our study demonstrates an essential role of P450s as endocrine modulators due to drug

dependent induction. Our data contribute to understand overall AED action and the role of P450 in

brain, thereby helping to improve drug design.


at ASPE



ownloaded from


JPET #120303

21

Acknowledgements

We thank Margarethe Ditter and Monika Hock, Dept. Neuropathology, for their extraordinary techni-

cal assistance and support, Georgios Pantazis for his support with the statistical analysis, and Dr. Rolf

Knoth, Dept. Neuropathology, for the confocal laser scanning microscopy and his scientific support.


at ASPE



ownloaded from


JPET #120303

22

References

Agarwal VR, Sinton CM, Liang C, Fisher C, German DC and Simpson ER (2000) Upregulation of estrogen

receptors in the forebrain of aromatase knockout (ArKO) mice. Mol Cell Endocrinol 162:9-16.

Banu SK, Govindarajulu P and Aruldhas MM (2002) Testosterone and estradiol up-regulate androgen and estro-

gen receptors in immature and adult rat thyroid glands in vivo. Steroids 67:1007-1014.

Death AK, McGrath KC and Handelsman DJ (2005) Valproate is an anti-androgen and anti-progestin. Steroids

70:946-953.

Giger U and Meyer UA (1983) Effect of succinylacetone on heme and cytochrome P450 synthesis in hepatocyte

culture. FEBS Lett 153:335-338.

Gupta C, Sonawane BR, Yaffe SJ and Shapiro BH (1980) Phenobarbital exposure in utero: alterations in female

reproductive function in rats. Science 208:508-510.

Hagemeyer CE, Rosenbrock H, Ditter M, Knoth R and Volk B (2003) Predominantly neuronal expression of

cytochrome P450 isoforms CYP3A11 and CYP3A13 in mouse brain. Neuroscience 117:521-529.

Handschin C and Meyer UA (2003) Induction of drug metabolism: the role of nuclear receptors. Pharmacol Rev

55:649-673.

Herzog AG and Fowler KM (2005) Sexual hormones and epilepsy: threat and opportunities. Curr Opin Neurol

18:167-172.

Huang H, Wang H, Sinz M, Zoeckler M, Staudinger J, Redinbo MR, Teotico DG, Locker J, Kalpana GV and

Mani S (2007) Inhibition of drug metabolism by blocking the activation of nuclear receptors by ketoconazole.

Oncogene 26:258-268.

Ingelman-Sundberg M (2004) Human drug metabolising cytochrome P450 enzymes: properties and polymor-

phisms. Naunyn Schmiedebergs Arch.Pharmacol. 369:89-104.

Isojarvi JI, Tauboll E and Herzog AG (2005) Effect of antiepileptic drugs on reproductive endocrine function in

individuals with epilepsy. CNS Drugs 19:207-223.

Joe I and Ramirez VD (2001) Binding of estrogen and progesterone-BSA conjugates to glyceraldehyde-3-

phosphate dehydrogenase (GAPDH) and the effects of the free steroids on GAPDH enzyme activity: physiologi-

cal implications. Steroids 66:529-538.


at ASPE



ownloaded from


JPET #120303

23

Karenlampi SO, Tuomi K, Korkalainen M and Raunio H (1989) Induction of cytochrome P450IA1 in mouse

hepatoma cells by several chemicals. Phenobarbital and TCDD induce the same form of cytochrome P450. Bio-

chem Pharmacol 38:1517-1525.

Khvotchev MV, Ren M, Takamori S, Jahn R and Sudhof TC (2003) Divergent functions of neuronal Rab11b in

Ca2+-regulated versus constitutive exocytosis. J Neurosci 23:10531-10539.

Kim SW, Kim NN, Jeong SJ, Munarriz R, Goldstein I and Traish AM (2004) Modulation of rat vaginal blood

flow and estrogen receptor by estradiol. J Urol 172:1538-1543.

Krongrad A, Wilson CM, Wilson JD, Allman DR and McPhaul MJ (1991) Androgen increases androgen recep-

tor protein while decreasing receptor mRNA in LNCaP cells. Mol Cell Endocrinol 76:79-88.

Lu S, Simon NG, Wang Y and Hu S (1999) Neural androgen receptor regulation: effects of androgen and

antiandrogen. J Neurobiol 41:505-512.

Lustig RH, Hua P, Smith LS, Wang C and Chang C (1994) An in vitro model for the effects of androgen on

neurons employing androgen receptor-transfected PC12 cells. Mol Cell Neurosci 5:587-596.

Melcangi RC and Panzica GC (2006) Neuroactive steroids: old players in a new game. Neuroscience 138:733-

739.

Meyer RP, Hagemeyer CE, Knoth R, Kaufmann MR and Volk B (2006) Anti-epileptic drug phenytoin enhances

androgen metabolism and androgen receptor expression in murine hippocampus. J Neurochem 96:460-472.

Meyer RP, Knoth R, Schiltz E and Volk B (2001) Possible function of astrocyte cytochrome P450 in control of

xenobiotic phenytoin in the brain: in vitro studies on murine astrocyte primary cultures. Exp Neurol 167:376-

384.

Meyer RP, Lindberg RL, Hoffmann F and Meyer UA (2005) Cytosolic persistence of mouse brain CYP1A1 in

chronic heme deficiency. Biol Chem 386:1157-1164.

Meyer RP, Podvinec M and Meyer UA (2002) Cytochrome P450 CYP1A1 accumulates in the cytosol of kidney

and brain and is activated by heme. Mol Pharmacol 62:1061-1067.

Meyer RP, Volk B and Knoth R (2004) Function of Astrocyte Cytochrome P450 in Control of Xenobiotic Me-

tabolism, in The Role of Glia in Neurotoxicity (Aschner M and Costa LG eds) pp 61-72, CRC Press, Boca Raton.

Nelson-DeGrave VL, Wickenheisser JK, Cockrell JE, Wood JR, Legro RS, Strauss JF, III and McAllister JM

(2004) Valproate potentiates androgen biosynthesis in human ovarian theca cells. Endocrinology 145:799-808.


at ASPE



ownloaded from


JPET #120303

24

Nelson DR, Zeldin DC, Hoffman SM, Maltais LJ, Wain HM and Nebert DW (2004) Comparison of cytochrome

P450 (CYP) genes from the mouse and human genomes, including nomenclature recommendations for genes,

pseudogenes and alternative-splice variants. Pharmacogenetics 14:1-18.

Nguyen TV, Yao M and Pike CJ (2005) Androgens activate mitogen-activated protein kinase signaling: role in

neuroprotection. J Neurochem 94:1639-1651.

Quarmby VE, Yarbrough WG, Lubahn DB, French FS and Wilson EM (1990) Autologous down-regulation of

androgen receptor messenger ribonucleic acid. Mol Endocrinol 4:22-28.

Rencurel F, Foretz M, Kaufmann MR, Stroka D, Looser R, Leclerc I, da S, X, Rutter GA, Viollet B and Meyer

UA (2006) Stimulation of AMP-Activated Protein Kinase Is Essential for the Induction of Drug Metabolizing

Enzymes by Phenobarbital in Human and Mouse Liver. Mol Pharmacol 70:1925-1934.

Rosenbrock H, Hagemeyer CE, Singec I, Knoth R and Volk B (1999) Testosterone metabolism in rat brain is

differentially enhanced by phenytoin-inducible cytochrome P450 isoforms. J Neuroendocrinol 11:597-604.

Schule R, Rangarajan P, Kliewer S, Ransone LJ, Bolado J, Yang N, Verma IM and Evans RM (1990) Functional

antagonism between oncoprotein c-Jun and the glucocorticoid receptor. Cell 62:1217-1226.

Strobel HW, Thompson CM and Antonovic L (2001) Cytochromes P450 in brain: function and significance.

Curr Drug Metab 2:199-214.

Tauer U, Knoth R and Volk B (1998) Phenytoin alters Purkinje cell axon morphology and targeting in vitro.

Acta Neuropathol 95:583-591.

Thuerl C, Otten U, Knoth R, Meyer RP and Volk B (1997) Possible role of cytochrome P450 in inactivation of

testosterone in immortalized hippocampal neurons. Brain Res 762:47-55.

Tischler AS and Greene LA (1975) Nerve growth factor-induced process formation by cultured rat pheochromo-

cytoma cells. Nature 258:341-342.

Volk B, Meyer RP, von Lintig F, Ibach B and Knoth R (1995) Localization and characterization of cytochrome

P450 in the brain. In vivo and in vitro investigations on phenytoin- and phenobarbital-inducible isoforms. Toxi-

col Lett 82-83:655-662.

Wang H, Napoli KL and Strobel HW (2000) Cytochrome P450 3A9 catalyzes the metabolism of progesterone

and other steroid hormones. Mol Cell Biochem 213:127-135.


at ASPE



ownloaded from


JPET #120303

25

Footnotes

This study was supported by the Deutsche Forschungsgemeinschaft (DFG Me 1544/4-1) and the For-

schungskommission of the University Hospital Freiburg (MEY 375/05, MEY 465/06).

Parts of this study were presented at the 15th International Symposium on Microsomes and Drug Oxi-

dations (2004) and the 16th International Symposium on Microsomes and Drug Oxidations (2006).

Reprint requests to Ralf P. Meyer, Ph.D., Pathologisches Institut, Abt. Neuropathologie,

Neurozentrum, Universität Freiburg, Breisacherstraße 64, D-79106 Freiburg, Tel:+49-761-270-5078,

Fax:+49-761-270-5050, E-Mail: [email protected]


at ASPE



ownloaded from


JPET #120303

26

Legends for Figures

Figure 1. Endogenous expression of AR and CYP3A in Hepa1c1c7 cells. (a) Agarose gel electro-

phoresis of RT–PCR products using mAR Taqman primers demonstrate the presence of mAR mRNA

in Hepa1c1c7. No expression was found in a non template control (NTC). (b) Immunoblot analysis of

microsomes (mc) and 9000g supernatant (9S) was done using the polyclonal anti rat CYP3A1/2 anti-

body. Murine liver microsomes were used as positive control.

Figure 2. Endogenous expression of AR and NFpan in PC-12. (a) Agarose gel electrophoresis of

RT–PCR products using rAR Taqman primers demonstrate the presence of rAR mRNA in PC-12. (b)

High-power confocal images of double-labeled PC-12 illustrating coexpression of neurofilament pan

(NFpan, top) and androgen receptor (AR, middle). Coexpression of the antigens in the overlay channel

(bottom).

Figure 3. Influence of phenytoin (DPH) treatment on CYP3A11 and AR regulation and activity

in Hepa1c1c7-cells. Regulation of (a) cyp3a11 and (b) AR mRNA after phenytoin (DPH) treatment

and additional silencing with siRNA against cyp3a11, measured by quantitative real time RT-PCR

(Taqman). Data were presented as ratio of expression quantities of the cyp3a11 or AR mRNAs vs. that

of GAPDH (means ± SD of a representative experiment of three experiments, each target gene tran-

script was measured in triplicate) and finally normalized against untreated non silencing control

(scrambled anti-sense siRNA), respectively. The interassay coefficient of variation was 10 %. The p-

values are indicated by asterisks, ***p < 0.001, **p < 0.01, and n.s. = not significant.

Regulation of (c) AR activity after phenytoin (DPH) treatment and additional silencing with siRNA

against cyp3a11 determined by a reporter gene assay using an MMTV-ARE-CAT reporter. Data were

presented as expression of CAT after 100µM phenytoin application and transfection of cyp3a11

siRNA or non silencing siRNA (scrambled anti-sense siRNA), respectively. CAT expression was de-

termined using CAT- ELISA and normalized vs. non silencing control transfected untreated cells.

Data are presented as means ± SD of a representative experiment of three experiments performed in

triplicates, all differences were ***p < 0.001. The interassay coefficient of variation was 3.5 %.


at ASPE



ownloaded from


JPET #120303

27

Figure 4. Androgen receptor protein expression in Hepa1c1c7 cells after treatment with P450-

inducing drugs. Expression of (a) AR protein after phenobarbital treatment with or without the in-

hibitor of CYP3A-activity ketoconazole (immunoblot). Data were presented as ratio of expression

quantities (optical density) of AR protein in 9000xg pellets resembling nuclei (9P) vs. that of GAPDH

of one representative experiment and normalized against untreated cells. Expression and localization

of (b) AR protein after AED treatment using immunocytochemistry. Phenytoin and Phenobarbital

treated Hepa1c1c7 cells demonstrated higher overall immunoreactivity as compared to untreated cells

(Scale bar = 25µm).

Figure 5. Regulation of CYP3A2 mRNA, AR mRNA and AR protein in PC-12 cells after treat-

ment with P450- inducers phenytoin and phenobarbital with or without CYP3A-activity inhibi-

tor ketoconazole. Expression of (a) CYP3A2 and (b) AR mRNA detected by quantitative real time RT-

PCR (Taqman). Data were presented as ratio of expression quantities of the (a) CYP3A2 or (b) AR

mRNA vs. that of GAPDH (means ± SD of a representative experiment of three experiments, each

target gene transcript was measured in triplicate). The p-values are indicated by asterisks, *p < 0.05,

***p < 0.001. The interassay coefficients of variation were less than 1 % (a), and 8.5 % (b).

Expression of AR protein (c) detected by immunoblot. Data were presented as ratio of expression quan-

tities (optical density) of AR protein vs. that of GAPDH in cell homogenates of one representative

experiment and normalized against untreated cells.

Figure 6. Regulation of cyp3a11 mRNA, AR mRNA and AR activation due to CYP3A11 overex-

pression and additional silencing with cyp3a11 siRNA in PC-12

Expression of CYP3A11 (a) and AR (b) mRNA detected by quantitative real time RT-PCR (Taqman).

Data were presented as ratio of expression quantities of the (a) cyp3a11 or (b) AR mRNA vs. that of

GAPDH (means ± SD of a representative experiment of six experiments, each target gene transcript

was measured in triplicate). Finally, (b) AR mRNA expression was normalized against vector/ non

silencing (scrambled anti-sense siRNA) transfected control (pcDNA/ non sil.ctrl.).

(c) Androgen receptor activation due to CYP3A11 overexpression and additional silencing with

cyp3a11 siRNA using an ARE-TK-CAT reporter. Data were presented as expression of CAT due to


at ASPE



ownloaded from


JPET #120303

28

CYP3A11 overexpression and co-transfection of cyp3a11 siRNA or non silencing siRNA (scrambled

anti-sense siRNA), respectively. CAT expression was determined using a CAT- ELISA and normal-

ized vs. vector/ non silencing transfected control (pcDNA/ non sil.ctrl). (Means ± SD of a representa-

tive experiment of four experiments performed in triplicates).

The interassay coefficients of variation were less than 8 %. The p-values are indicated by asterisks,

**p < 0.01, ***p < 0.001.

Figure 7. Influence of testosterone withdrawal on AR expression and activation in Hepa1c1c7

cells. (a) Expression of cyp3a11 and AR mRNA after 100µM phenytoin (DPH) treatment in absence of

testosterone. Data were presented as ratio of expression quantities of the cyp3a11 or AR mRNAs vs.

that of GAPDH measured by quantitative real time RT-PCR (means ± SD of a representative experi-

ment of three experiments, each target gene transcript was measured in triplicate) and finally normal-

ized against untreated cells. (b) AR activation determined by MMTV-ARE-CAT reporter gene assay.

Data were presented as expression of CAT after treatment with 100µM phenytoin (DPH). CAT ex-

pression was determined using CAT- ELISA and normalized vs. untreated cells (Means ± SD of a

representative experiment of four experiments performed in triplicates). The interassay coefficients of

variation were less than 8 % (a), and 1 % (b), respectively. The p-values are indicated by asterisks,

***p < 0.001, n.s. = not significant.


at ASPE



ownloaded from



at ASPE



ownloaded from


This article has not been copyedited and form

atted. The final version m

ay differ from this version.

JPET

Fast Forward. Published on M

ay 15, 2007 as DO

I: 10.1124/jpet.107.120303 at ASPET Journals on June 30, 2020 jpet.aspetjournals.org Downloaded from



at ASPE



ownloaded from



at ASPE



ownloaded from



at ASPE



ownloaded from



at ASPE



ownloaded from



at ASPE



ownloaded from


Documents

Antiepileptic drugs affect neuronal androgen signaling via a …jpet.aspetjournals.org/content/jpet/early/2007/05/15/jpet.107.120303... · strual disorders, polycystic ovaries, infertility,