MOLECULAR AND CELLULAR BIOLOGY, Sept. 2004, p. 7931–7940
0270-7306/04/$08.00?0 DOI: 10.1128/MCB.24.18.7931–7940.2004
Vol. 24, No. 18
Nuclear Receptors CAR and PXR Cross Talk with FOXO1 To
Regulate Genes That Encode Drug-Metabolizing
and Gluconeogenic Enzymes
Susumu Kodama, Chika Koike, Masahiko Negishi,* and Yukio Yamamoto
Pharmacogenetics Section, Laboratory of Reproductive and Developmental Toxicology, National Institute of
Environmental Health Sciences, National Institutes of Health, Research Triangle Park, North Carolina
Received 23 February 2004/Returned for modification 18 March 2004/Accepted 16 June 2004
The nuclear receptors CAR and PXR activate hepatic genes in response to therapeutic drugs and xenobi-
otics, leading to the induction of drug-metabolizing enzymes, such as cytochrome P450. Insulin inhibits the
ability of FOXO1 to express genes encoding gluconeogenic enzymes. Induction by drugs is known to be
decreased by insulin, whereas gluconeogenic activity is often repressed by treatment with certain drugs, such
as phenobarbital (PB). Performing cell-based transfection assays with drug-responsive and insulin-responsive
enhancers, glutathione S-transferase pull down, RNA interference (RNAi), and mouse primary hepatocytes, we
examined the molecular mechanism by which nuclear receptors and FOXO1 could coordinately regulate both
enzyme pathways. FOXO1 was found to be a coactivator to CAR- and PXR-mediated transcription. In contrast,
CAR and PXR, acting as corepressors, downregulated FOXO1-mediated transcription in the presence of their
activators, such as 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene (TCPOBOP) and pregnenolone 16?-carboni-
trile, respectively. A constitutively active mutant of the insulin-responsive protein kinase Akt, but not the
kinase-negative mutant, effectively blocked FOXO1 activity in cell-based assays. Thus, insulin could repress the
receptors by activating the Akt-FOXO1 signal, whereas drugs could interfere with FOXO1-mediated tran-
scription by activating CAR and/or PXR. Treatment with TCPOBOP or PB decreased the levels of phos-
phoenolpyruvate carboxykinase 1 mRNA in mice but not in Car?/?mice. We conclude that FOXO1 and the
nuclear receptors reciprocally coregulate their target genes, modulating both drug metabolism and
Liver plays a major role in the metabolism of therapeutic
drugs and environmental contaminants. It is endowed with a
mechanism to induce hepatic enzymes, leading to increased
detoxification and elimination of those xenobiotics. Drug in-
duction is generally regulated by transcriptional activation of
hepatic genes encoding drug-metabolizing enzymes, such as
cytochrome P450s (CYPs) and specific transferases. Acting as
the principal transcription factors which also form complexes
with RXR, the nuclear receptors CAR and PXR play a central
role in induction by binding to the phenobarbital (PB)- and
xenobiotic-responsive enhancer modules PBREM and XREM
(5, 6, 11, 12, 18, 46, 53), respectively, and activating transcrip-
tion of their target genes, such as CYP genes, in response to a
distinct but overlapping group of xenobiotics (23, 36, 50). How-
ever, induction is heavily influenced by endocrine conditions;
glucocorticoid hormone, for example, augments CYP induc-
tion by PB (31). In contrast, insulin is known to repress the
induction of drug-metabolizing activity by certain drugs and in
diabetic livers (44, 49, 56). Insulin treatment either eliminated
or significantly reduced PB induction of CYP2B in rat primary
hepatocytes (17, 42, 58). Hepatic CYP2B, CYP3A, and
CYP4A were increased in experimentally generated diabetic
rats and mice and were reduced to normal levels by insulin
treatment (37, 56). While insulin is involved in regulating drug
metabolism, drug treatment influences insulin-responsive glu-
cose metabolism. For instance, both PB and 1,4-bis[2-(3,5-
dichloropyridyloxy)]benzene (TCPOBOP) decreased hepatic
enzymes, such as phosphoenolpyruvate carboxykinase 1
(PEPCK1) and glucose-6-phosphatase (G6P) (1, 24, 50).
Chronic PB treatment decreased plasma glucose and improved
insulin sensitivity in diabetic patients (20). Here we investi-
gated the molecular mechanism of insulin repression by focus-
ing on whether CAR and/or PXR could be coregulated by
insulin response transcription factor FOXO1 and vice versa.
A forkhead transcription factor, FOXO1, is an activator of
gluconeogenic genes, such as PEPCK1, G6P, and insulin-like
growth factor-binding protein 1 (29, 32, 38, 57). Insulin inhibits
FOXO1 activity, leading to repression of these genes. In fact,
the insulin receptor knockout mouse partially restored serum
glucose to normal levels when it was crossed with a Foxo1?/?
mouse (28). These gluconeogenic genes contain an insulin
response sequence (IRS) to which FOXO1 can bind directly
and activate in the absence of insulin (7, 32, 38, 51, 57). Insulin
treatment triggers the phosphorylation of FOXO1 via a phos-
phatidylinositol 3-kinase (PI3K)-Akt pathway (7, 30, 35). Phos-
phorylation inactivates FOXO1 by decreasing the binding af-
finity of FOXO1 to IRS and/or exporting FOXO1 from the
nucleus (2, 25, 34, 59). FOXO1 has been implicated as a
coregulator for various nuclear receptors, coactivating or core-
pressing the estrogen, thyroid, and retinoid A receptors (39,
60), whereas glucocorticoid and progesterone receptors (60),
* Corresponding author. Mailing address: Pharmacogenetics Sec-
tion, Laboratory of Reproductive and Developmental Toxicology, Na-
tional Institute of Environmental Health Sciences, National Institutes
of Health, Research Triangle Park, NC 27709. Phone: (919) 541-2404.
Fax: (919) 541-0696. E-mail: email@example.com.
hepatic nuclear factor 4 (HNF4) (9), and peroxisome prolif-
erator-activated receptor ? (PPAR?) (4) are corepressed. In
addition, the androgen receptor prevents the binding of
FOXO1 to IRS, leading to a repression of FOXO1-IRS activity
(21). Our yeast two-hybrid screening identified FOXO1 as a
CAR-binding protein. These observations prompted us to in-
vestigate whether FOXO1 mediates the insulin-dependent re-
pression of CAR and/or PXR-mediated trans-activation of
drug-induced genes, such as CYP genes. Moreover, we also
investigated whether CAR and/or PXR receptors reciprocally
coregulate FOXO1-IRS activity.
By employing glutathione S-transferase (GST) pull-down,
mammalian two-hybrid, and transient luciferase reporter assays
and in vivo animal experiments, we demonstrate here that the
nuclear receptors and FOXO1 cross-talk to coregulate their re-
sponse elements. Drug and glucose metabolisms, which are two
major liver functions that can be regulated independently, ap-
peared to be coordinately regulated by this cross-talk.
MATERIALS AND METHODS
Materials. 5?-Androst-16-en-3?-ol (androstenol) was purchased from Ster-
aloids, Inc. Insulin, insulin-like growth factor 1 (IGF1), TCPOBOP, and preg-
nenolone 16?-carbonitrile (PCN) were purchased from Sigma-Aldrich. Restric-
tion endonucleases and DNA-modifying enzymes were purchased from New
England Biolabs, Inc. [35S]methionine was purchased from Amersham.
LY294002 was purchased from Calbiochem.
Plasmid constructions. The following plasmids were kindly provided: XREM-
3A4-Luc plasmid by Brian Goodwin at GlaxoSmithKline (5); rPKB?/pECE and
rPKB?K179 M/pECE by Ushio Kikkawa at Kobe University; myr-rPKB?/pcDL-
SRM by Kiyoshi Hidaka at the National Institute of Environmental Health
Sciences. In all plasmids, m, r, and h in front of the insert denote mouse, rat, and
human, respectively. The pGL3/1.8 kbp-2B6-Luc, pGL3/1.6 kbp-2B6-Luc (52),
pGEX/mCAR, pCMX/hRXR (11), pCR3/mCAR (47), and pBIND/mCAR (55)
plasmids were described previously. The CAR mutant T176V in pCR3 plasmid
and more details of its characterization will be published elsewhere (A. Ueda, K.
Matsui, Y. Yamamoto, L. Pedersen, T. Sueyoshi, and M. Negishi, unpublished
data). Full-length PXR (18) was amplified from a mouse liver cDNA library and
was cloned into pcDNA3-V5-His (Invitrogen) and pGEX-4T-1 (Stratagene)
vectors designated pcDNA3/mPXR and pGEX/mPXR, respectively. cDNA en-
coding residues 425 to 652 was amplified from a FOXO1 clone selected from the
yeast two-hybrid screening and was inserted into pACT (Promega) to generate
pACT/mFOXO1Ct. Using proper sets of primers, the first and second exons were
amplified from mouse genomic DNA using LA taq DNA polymerase with GC
buffer (TaKaRa Shuzo, Otsu, Japan) and were cloned into pCR2.1-TOPO (In-
vitrogen). Inserts were cut out by double digestion with BamHI and EcoRI from
the TOPO plasmids, ligated, and recloned into pcDNA3-V5-His to obtain the
full-length cDNA of FOXO1, designated pcDNA3/mFOXO1. Using pcDNA3/
mFOXO1 as a template, proper mutated primers, and the QuikChange mu-
tagenesis kit (Stratagene), Thr24, Ser253, and Ser316 were mutated to alanine to
generate pcDNA3/mFOXO13A. To clone the full-length FOXO1 cDNA into
pGEX-4T-1, the cDNA was amplified to include an additional XhoI site at the
5? end, digested with XhoI, and inserted at the SalI site of pGEX-4T-1 to create
pGEX/mFOXO1. For the various Akt expression vectors, the full-length cDNAs
of wild-type Akt (AktWt), kinase-negative Akt (AktKN), and constitutively acti-
vated Akt (AktCA) were amplified from rPKB?/pECE, rPKB?K179 M/pECE,
and myr-rPKB?/pcDL-SRM, respectively. These cDNAs were cloned into
pcDNA3-V5-His plasmids designated pcDNA3/AktWt, pcDNA3/AktKN, and
pcDNA3/AktCA, respectively. To construct the pGL3/IRS-tk-Luc plasmid, a
complementary pair of oligonucleotides containing three tandem copies of IRS
GTA-3? from the human IGFBP1 gene) (48) was synthesized. The underlined
sequence indicates an additional overhanging XhoI restriction site. Oligonucle-
otides were annealed and cloned into the XhoI-digested site of the pGL3-Basic
vector containing a 160-bp thymidine kinase (tk) promoter (47), and a plasmid
was selected containing six repeats of IRS. For amplification of FOXO1 from its
genomic clone, Pfu Turbo DNA polymerase (Stratagene) was used. Insert se-
quences of newly constructed vectors were verified by DNA sequencing.
Yeast two-hybrid screen. The ligand-binding domain (residues 87 to 358) of
the mouse CAR T176V mutant was cloned into the pGBKT7 plasmid (Clon-
tech). This plasmid was cotransfected with the mouse liver MATCHMAKER
cDNA library (Clontech) into the Y190 yeast strain. Positive yeast colonies were
selected in the presence of TCPOBOP (1 ?M), and the binding activity was
measured by ?-galactosidase assay according to the manufacturer’s instructions.
One-hundred five yeast colonies were isolated from 2.2 ? 106primary transfor-
mants. The plasmids encoding the activation domain fusion proteins were re-
covered from these yeast transformants, and the inserts were sequenced. Three
of these inserts were found to encode FOXO1 and were cloned as a fusion
protein in frame with the GAL4 activation domain.
GST pull-down assay. The GST fusion proteins GST-mCAR, GST-mPXR,
and GST-mFOXO1 were expressed in Escherichia coli strain BL21 cells and were
purified with glutathione-Sepharose. The cDNAs of FOXO1, CAR, and PXR in
pcDNA3-V5-His were in vitro translated in the presence of [35S]methionine
using the TNT-coupled reticulocyte lysate system (Promega) according to the
manufacturer’s instructions. GST pull-down assays were carried out as described
Cell culture and transfection. HepG2 cells were cultured in minimum essential
medium (MEM) supplemented with 10% fetal bovine serum, 2 mM L-glutamine,
100 U of penicillin per ml, and 100 ?g of streptomycin per ml in an atmosphere
of 5% CO2at 37°C. Primary hepatocytes were prepared from male Cr1:CD-
1(ICR)BR mice (Charles River Laboratories) as described previously (13).
HepG2 cells plated in 24 wells at 50 to 60% confluence were transfected with
lipofectamine 2000 (Invitrogen) according to the manufacturer’s instruction.
pRL-CMV (10 ng) (Promega) was used as an internal control for all transfection
assays. Mouse primary hepatocytes were cotransfected with pGL3/1.8 kbp-2B6-
Luc (10 ?g) and pRL-CMV (5 ?g) by the electroporation method (16). For
experiments with RNAi, SMARTpool (M-003006-01-05; Dharmacon) for human
FOXO1 or an unrelated RNA (5?-NNUGGUAGGCAGGAUGAGUAC-3?) as
nonspecific control was transfected into HepG2 cells using lipofectamine 2000.
After 24 h of transfection the cells were treated with drug for an additional 48 h
Animal treatment. Cr1:CD-1(ICR)BR male mice (7 to 8 weeks old) main-
tained on a 12-h light and 12-h dark cycle were randomly divided into two groups
that were intraperitoneally injected with TCPOBOP (0.3 mg/kg of body weight)
or dimethyl sulfoxide (DMSO). Twenty-four hours after injection, mice were fed
until sacrificed to prepare liver RNAs. The original background strain of Car?/?
mice used in a previous work (50) was changed to C3H/HeNCrlBR mice
(Charles River) by repeated backcrossing until microsatellite analysis showed
that mice contained 98% of the C3H markers. Car?/?and Car?/?male mice
were treated with a single injection of 90 mg of diethylnitrosamine/kg, followed
by chronic treatment with PB in drinking water (500 ppm) for 32 weeks. These
treatments were originally for investigating liver tumor promotion. Livers and
blood samples were collected and used for analyses.
Real-time PCR. TRIZOL reagent (Invitrogen) was used to extract total RNAs
from HepG2 cells, mouse livers, and primary mouse hepatocytes, and cDNAs
were prepared from total RNAs using Super-Script II reverse transcriptase
(Invitrogen). Real-time PCR was performed with an ABI prism 7700 sequence
detection system (Applied Biosystems) with the following primers and probes:
Hs00231106_m1 (Applied Biosystems) for the human FOXO1 gene; 5?-6FAM-
CTCTTCCA-3?, and 5?-CAGCAGGCGCAAGAACTGA-3? for CYP2B10 (14);
CATCCGCAAGCTGAAG-3?, and 5?-TTCGATCCTGGCCACATCTC-3? for
the mouse Pepck1 gene. The TaqMAN rodent GAPDH control reagent (Applied
Biosystems) was used as internal control.
Gel shift assay. CAR, RXR, PXR, and FOXO1 proteins were produced by in
vitro translation as described above. For probes, three sets of double-stranded
oligonucleotides containing NR1 sequence (47), wild-type IRS (5?-GATCGCT
AGATGCAAAACAACTTGTGACGATC-3?), and mutant IRS (5?-GATCGC
TAGATGCAACAGAACTTGTGACGATC-3?) were synthesized and end-la-
beled by using [32P]dATP and DNA polymerase Klenow fragment (New England
Biolab). Gel shift assays were performed as described previously (11). For com-
petition or antibody supershift assays, unlabeled oligonucleotide, rabbit immu-
noglobulin G (Santa Cruz), or FOXO1 antibody (Santa Cruz) was added 15 min
before adding radioactive oligonucleotide to start the binding reaction.
FOXO1 as a CAR-binding protein. Yeast two-hybrid screen-
ing of a mouse liver cDNA library was performed using the
7932 KODAMA ET AL.MOL. CELL. BIOL.
CAR ligand-binding domain as bait. Of the total 105 positive
clones isolated, 18 were RXR? and five were RXR?. Both
receptors are known to form heterodimers with CAR, indicat-
ing that the screening was effective. The forkhead transcription
factor FOXO1 was also one of the positive clones, and the
deduced amino acid sequences from three independent clones
contained 228 residues of the C-terminal region of FOXO1.
Subsequently, a mammalian two-hybrid assay was employed to
examine the functional interaction between CAR and FOXO1
in which pBind/mCAR (CAR fused to the GAL4 DNA-bind-
ing domain) and VP16AD/mFOXOCt(cloned VP16 activation
domain in front of FOXO1Ct) were cotransfected into HepG2
cells. Although CAR is constitutively active in HepG2 cells, its
fusion with GAL4DBD decreased constitutive activity and
conferred the capability of being activated by TCPOBOP (Fig.
1a). Cotransfection of VP16AD/mFOXOCtresulted in a 3.0-
fold coactivation in the presence of TCPOBOP, while no such
coactivation was observed in the presence of the CAR repres-
sor androstenol (Fig. 1a). VP16AD/mFOXOCtalone did not
enhance the reporter activity. We performed a GST pull-down
assay to investigate the direct binding of CAR with FOXO1 by
using an in vitro-translated full-length FOXO1 and a bacteri-
ally expressed recombinant GST-mCAR fusion protein or vice
versa. Binding of GST-CAR (or GST-FOXO1) to the
labeled FOXO1 (or CAR) was decreased in the presence of
androstenol and was recovered by the addition of TCPOBOP
(Fig. 1b). These results unequivocally showed that CAR binds
to FOXO1 in a ligand-dependent manner.
FOXO1 as a CAR coactivator. Given the fact that FOXO1 is
a CAR-binding protein, a series of transient transfection assays
were performed to investigate whether FOXO1 coregulated
the trans-activation of the CAR target gene CYP2B6. For this
purpose, the 1.8-kbp 5?-flanking DNA of the CYP2B6 gene
that contains a 51-bp CAR-binding site, called PBREM, was
placed in front of a Luc reporter gene. FOXO1 coactivated the
1.8 kbp-2B6-Luc reporter 2.5-fold when CAR was coexpressed
in the presence and absence of TCPOBOP, whereas FOXO1
alone could not activate the reporter (Fig. 2a). FOXO1 con-
tains three Akt phosphorylation sites (Thr-24, Ser-253, and
Ser-316). The FOXO13Amutant, in which all three sites were
replaced by alanine, should result in an increased retention
and activity of FOXO1 in the nucleus (7, 38). The nuclear-
retaining FOXO13Amutant was more effective in coactivating
the reporter as indicated by an approximately 4-fold increase
instead of the 2.5-fold increase for FOXO1 (Fig. 2a). To ex-
amine whether coactivation by FOXO1 required the CAR-
binding site, the 1.6-kbp 5?-flanking DNA of the CYP2B6 gene
that lacked the PBREM was used in the reporter assay.
FOXO1 did not coactivate the 1.6 kbp-2B6-Luc reporter (Fig.
2a). Moreover, similar to the 1.8 kbp-2B6-Luc reporter, the
PBREM-tk-Luc reporter gene was also coactivated by FOXO1
(data not shown). Thus, FOXO1 coactivated CAR-mediated
PBREM activity. Although the fact that coactivation by
FOXO1 and FOXO13Awas repressed by androstenol sug-
gested coactivation occurred when CAR was active, determin-
ing whether the coactivation required direct activation of CAR
by TCPOBOP was problematic because CAR was constitu-
tively active in transfected HepG2 cells. Therefore, the CAR
mutant T176V that possessed low constitutive activity and the
capability of being activated by TCPOBOP was also used for
these assays. In these experiments, coactivation by FOXO1
occurred only in the presence of TCPOBOP (Fig. 2a). Because
CAR activates XREM of the CYP3A gene that contains mul-
tiple receptor-binding sites (43), we also tested whether
FOXO1 could coactivate the XREM (Fig. 2b). CAR-mediated
XREM activity was upregulated by FOXO1 and FOXO13A
2.5- and 3.5-fold, respectively. Furthermore, this upregulation
occurred only in the presence of TCPOBOP when T176V was
used for assays. These results clearly indicated that FOXO1
acts as a coactivator to regulate CAR-mediated trans-activa-
tion of multiple targets.
Akt as a signal of FOXO1-CAR coactivation. FOXO1 is an
insulin response transcription factor that is regulated by the
PI3K-Akt pathway. In response to insulin, PI3K inactivates
FOXO1 by phosphorylation. If this inactivation was conserved
FIG. 1. Direct binding of CAR to FOXO1. (a) Mammalian two-
hybrid assay. The pG5-Luc reporter plasmid was cotransfected with
various combinations of pBIND, pACT, pBIND/mCAR, and pACT/
mFOXO1Ctinto HepG2 cells. The amount of each plasmid was 0.2 ?g,
and the total amount of plasmids was adjusted by pcDNA3-V5-His. At
24 h after transfection, cells were treated with 0.1% DMSO,
TCPOBOP (250 nM), or androstenol (10 ?M) for an additional 24 h.
Subsequently, cells were harvested and cell extracts were prepared for
dual-luciferase assays. Relative activities were calculated by taking the
activity obtained from the GAL4DBD- and VP16AD-transfected cells
in the presence of DMSO as one. Bars indicate means ? standard
deviation. (b) GST pull-down assay. In vitro-translated
FOXO1 and CAR were incubated with bacterially expressed GST-
mFOXO1 and GST-mCAR fusion proteins, respectively, in the pres-
ence of 0.1% DMSO, 1 ?M TCPOBOP (T), 10 ?M androstenol (A),
or both (T/A). GST was used as a negative control for binding. Sodium
dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiogra-
phy were done as described in Materials and Methods.
VOL. 24, 2004COREGULATION BY FOXO1 AND RECEPTORS CAR AND PXR 7933
in HepG2 cells, insulin treatment should decrease FOXO1-
CAR activity to a lesser extent than the FOXO13A-CAR ac-
tivity. Consistent with this hypothesis, insulin treatment de-
creased FOXO1 coactivation of CAR by about 40%, but
coactivation with FOXO13Awas decreased by only 14% (Fig.
3a). This differential effect of FOXO1 and FOXO13Awas
significantly pronounced in the presence of PI3K inhibitor
LY294002. In fact, insulin did not reduce FOXO1 activity in
LY294002-treated HepG2 cells (Fig. 3a). To provide further
evidence that supports the direct involvement of FOXO1 in
coactivating CAR, coexpression of Akt or its functional mu-
tants should affect the coactivation accordingly. To examine
this hypothesis, AktWtand its kinase-negative and constitu-
tively activated mutants (AktKNand AktCA, respectively) were
coexpressed with FOXO1, CAR, and the 1.8-kbp Luc reporter
in HepG2 cells. Following insulin treatment, FOXO1 activity
was decreased by 40% (Fig. 3b). Similar to this insulin re-
sponse, coexpression of Akt decreased FOXO1-dependent co-
activation in the presence or absence of TCPOBOP, whereas,
as expected from the lack of kinase activity, AktKNdid not
affect the coactivation of CAR. Coexpression of AktCAap-
peared to effectively inactivate FOXO1, reducing the coacti-
vation by over 80%. Neither FOXO1 coactivation of CAR nor
its repression by AktWtwas observed in the presence of an-
FIG. 2. Coactivation of CAR by FOXO1 in HepG2 cells. (a) Tran-
sient transfection assays were performed by cotransfecting pCR3/
mCAR, pcDNA/T176V, pcDNA3/mFOXO1, or pcDNA3/mFOXO13A
with 1.8 kbp-2B6-Luc or 1.6 kbp-2B6-Luc reporter plasmid into
HepG2 cells. (b) XREM-3A4-Luc plasmid was used as the reporter.
The amount of each plasmid was 0.2 ?g, and the total amount of
plasmid in an assay was adjusted by using an empty plasmid. Cells were
treated as described in the legend to Fig. 1. Relative activities were
calculated by taking the activity of the cells that were transfected by
only the reporter plasmid in the presence of DMSO as one. Bars
indicate means ? standard deviations.
FIG. 3. Regulation by rAkt of coactivation of CAR by FOXO1.
Either pCR3/mCAR or pCR3/T176V was cotransfected with pcDNA3/
mFOXO1, pcDNA3/rAktWt, pcDNA3/rAktKN, and/or pcDNA3/rAk-
tCAinto HepG2 cells. In addition, the 1.8 kbp-2B6-Luc reporter was
also cotransfected. Except for the cases of insulin (500 nM) and
LY294002 (20 ?M), the drug concentrations of all treatments and
cultures were the same as those described in the legend to Fig. 1. (a)
Differential effects of insulin on the wild-type and mutant FOXO1-
mediated CAR activities. The percentage of activity relative to the
CAR- and FOXO1/FOXO13A-transfected cells in the presence of
TCPOBOP is given as 100. Bars indicate means ? standard deviations.
(b) Differential effects of the wild-type and mutated Akts on the
FOXO1-mediated CAR activity. Relative activities were calculated by
taking the activity of the cells that were transfected by only the reporter
plasmid in the presence of DMSO as one.
7934 KODAMA ET AL.MOL. CELL. BIOL.
drostenol. When the corresponding experiments were per-
formed with the T176A mutant, AktWtand its mutants had an
effect only in the presence of TCPOBOP (Fig. 3b). These
results indicate that an insulin-PI3K-Akt pathway regulates
FOXO1-dependent coactivation of CAR.
Given the fact that AktCAdecreased the activity level of the
1.8 kbp-2B6-Luc reporter below the activity level observed
without exogenous FOXO1 (Fig. 3b), we examined whether
AktCAinactivated endogenous FOXO1 to decrease the re-
porter activity. For this experiment, HepG2 cells were treated
with insulin and IGF1, and subsequently CAR-mediated re-
porter activity was measured. Although IGF1 was slightly more
effective, both insulin and IGF1 reduced the activity by 15 to
25%, suggesting that AktCAinactivated endogenous FOXO1
in HepG2 cells (Fig. 4a). If, in fact, the endogenous FOXO1
was directly involved in the regulation of CAR activity via Akt,
the removal of Akt should eliminate the activity. To test this
hypothesis, RNAi of FOXO1 was cotransfected into HepG2
cells. Real-time PCR analysis indicated an approximately 50%
reduction of endogenous FOXO1 mRNA (Fig. 4b). This re-
duction of endogenous FOXO1 mRNA was correlated with
the repression of CAR activity (Fig. 4c). In fact, the reporter
activity derived from CAR with RNAi of FOXO1 was de-
creased at a level equivalent to that observed by cotransfection
of AktCA. These results suggested that endogenous FOXO1
could also be functional in regulating the CAR-mediated
transactivation in HepG2 cells.
Although insulin repression of PB induction was reported
for rat primary hepatocytes (17, 42), it has not been demon-
strated in mouse primary hepatocytes. Upon finding that
FOXO1 is a key coactivator of CAR in HepG2 cells, we ex-
amined whether insulin repressed the induction of the Cyp2b10
gene by TCPOBOP in mouse primary hepatocytes (Fig. 5a).
TCPOBOP induced CYP2B10 mRNA levels fivefold without
insulin in the culture medium. The induction was slightly aug-
mented by 10?12M insulin, followed by a gradual decrease in
mRNA levels as insulin concentration increased. The rate of
induction by TCPOBOP at 10?6M insulin was only threefold
compared to the maximum eightfold induction at 10?12M
insulin. Under the culture conditions used, insulin clearly re-
pressed the induction of the endogenous Cyp2b10 gene, sug-
gesting that, acting as a CAR coactivator in response to insulin,
FOXO1 can regulate Cyp2b10 induction in vivo in hepatocytes.
In attempting to correlate the induction of the Cyp2b10 gene
with CAR-mediated 1.8 kbp-2B6-Luc reporter activity, the re-
porter plasmid was transfected into mouse primary hepato-
cytes, followed by insulin treatment. While insulin at 10?6M
was somewhat less effective in repressing CAR activity, the
luciferase reporter activity was decreased as insulin concentra-
tion increased up to 10?8M (Fig. 5b). These results support
the hypothesis that insulin regulates CAR activity to repress
the Cyp2b10 gene in primary hepatocytes.
PXR as another FOXO1 target. PXR activates the same
enhancer sequences as CAR. Therefore, we examined whether
FOXO1 could regulate PXR activity in the same way it can
with coactivated CAR. A GST pull-down assay showed that in
vitro-translated FOXO1 bound to GST-mPXR, which was in-
creased by the specific PXR activator pregnenolone PCN (Fig.
6a). PCN activated the PXR-dependent XREM-3A4-Luc pro-
moter in transfection assays in the absence of FOXO1 10-fold
(Fig. 6b). In the presence of exogenous FOXO1 and
FOXO13A, the promoter activity was further enhanced 2.5-
and more than 5-fold, respectively (Fig. 6b). Similar to results
obtained with CAR, the coactivation by FOXO1 was repressed
by insulin treatment and coexpression of AktWt, whereas inac-
FIG. 4. Regulation of CAR activity by endogenous FOXO1. (a)
Transient transfection assays. The 1.8 kbp-2B6-Luc reporter and
pCR3/mCAR were cotransfected with or without pcDNA3/rAktCA
into HepG2 cells. Cells were cultured and treated with insulin, IGF1,
and/or TCPOBOP as described in the legend to Fig. 1. Relative fold
activities were calculated by taking the activity of the cells that were
transfected with the pCR3/mCAR and 1.8 kbp-2B6-Luc plasmids in
the presence of TCPOBOP as 100. Bars indicate means ? standard
deviations. (b) Real-time PCR. HepG2 cells were transfected with
FOXO1 RNAi (50 pmol) or control oligonucleotides (50 pmol) in
MEM supplemented with 10% fetal bovine serum for 48 h and MEM
containing 5% charcoal dextran-treated fetal calf serum for another
24 h. Total RNA was extracted for real-time PCR analysis with specific
probes for FOXO1 mRNA. (c) Transient transfection assay. pCR3/
mCAR and 1.8 kbp-2B6-Luc plasmids were cotransfected in the pres-
ence of FOXO1 RNAi or control RNAi. Cells were cultured and
treated as described above. Relative activities were calculated by tak-
ing the activity of the cells that were transfected with the pCR3/mCAR
and 1.8 kbp-2B6-Luc plasmids in the presence of TCPOBOP and
control RNAi as 100.
VOL. 24, 2004COREGULATION BY FOXO1 AND RECEPTORS CAR AND PXR7935
tive AktKNdid not affect the coactivation (Fig. 6c). The con-
stitutively active AktCAabolished the coactivation completely
and decreased the promoter activity to a level below that with-
out FOXO1. These results confirmed that FOXO1 also coac-
tivates PXR via an insulin-PI3K-Akt pathway in the same way
that CAR activity is regulated.
CAR and PXR as FOXO1 corepressors. Whether CAR
and/or PXR could repress FOXO1-mediated trans-activation
of the IRS was investigated, because other nuclear receptors,
such as ER? and PPAR?, are known to repress IRS activation.
When FOXO1 alone was cotransfected with the IRS-tk-Luc
reporter into HepG2 cells, it activated the IRS approximately
eightfold (Fig. 7a). This IRS activity was not affected by the
presence of either the CAR activator TCPOBOP or the re-
pressor androstenol. As expected, insulin treatment decreased
IRS activity by 50%. Coexpression of CAR repressed IRS
activity to a level similar to that obtained by insulin treatment.
Consistent with its constitutive active nature, CAR repressed
the IRS activity in the absence of TCPOBOP. However, the
FIG. 5. Repression of the Cyp2b10 gene by insulin in mouse pri-
mary hepatocytes. (a) Mouse primary hepatocytes were prepared and
treated with 0.1% DMSO or 250 nM TCPOBOP in the presence of
various concentrations of insulin ranging from 10?12to 10?6M. At 8 h
after treatment, total RNA was extracted and subjected to real-time
PCR analysis with specific probes for the Cyp2b10 gene. Relative levels
were expressed by taking the level in DMSO-treated hepatocytes with-
out insulin treatment as one. (b) Luciferase activity was measured 24 h
after electroporation with pGL3/1.8 kbp-2B6-Luc (10 ?g), pRL-CMV
(5 ?g), and drug treatment. Relative activities were calculated by
taking the level in DMSO-treated hepatocytes without insulin treat-
ment as one. Bars indicate means ? standard deviations.
FIG. 6. Regulation of PXR activity by FOXO1. (a) GST pull-down
assay. In vitro-translated35S-labeled FOXO1 and PXR were incubated
with bacterially expressed GST-mFOXO1 and GST-mPXR, respec-
tively, in the presence of 0.1% DMSO (D) or 50 ?M PCN (P). (b)
Transient transfection assay. XREM-3A4-Luc reporter was cotrans-
fected with pcDNA3/mFOXO1 (or pcDNA3/mFOXO13A) and/or
pcDNA3/mPXR into HepG2 cells. Cells were treated with drugs, and
the activity was calculated as described in the legend to Fig. 1. In the
case of PCN treatment, 10 ?M was used to activate PXR. (c) Transient
transfection assays. XREM-3A4-Luc reporter was cotransfected with
pcDNA3/mPXR and pcDNA3/mFOXO1 into HepG2 cells. In some
cases, cells were additionally transfected with pcDNA3/rAktWt,
pcDNA3/rAktKN, or pcDNA3/rAktCA. PCN (10 ?M) and insulin (500
nM) were used to treat cells. Relative activities were calculated by
taking the activity of the cells that were transfected by only the reporter
plasmid in the presence of DMSO as one. Bars indicate means ?
7936 KODAMA ET AL.MOL. CELL. BIOL.
fact that the activity was retained in the presence of androste-
nol suggested that the receptor must be in an active form to
repress the IRS. Additional coexpression of RXR with CAR
eliminated FOXO1-IRS activity in which CAR, not RXR, was
the primary responsible factor, because the activity was re-
tained when RXR alone was coexpressed. During overexpres-
sion of RXR, CAR has a stronger ability to repress the IRS.
Similar to CAR, when PXR was activated by PCN the receptor
could also repress FOXO1-IRS activity (Fig. 7b). RXR over-
expression also increased the capability of PXR to repress the
activity even in the absence of PCN. These results clearly
indicate that both CAR and PXR repress FOXO1-IRS activity
in cotransfected HepG2 cells. The repression was strictly acti-
vator dependent in the absence of coexpressed RXR, while it
lost some degree of dependency in the presence of RXR.
Subsequently we performed gel shift assays using the in
vitro-translated proteins and an oligonucleotide probe contain-
ing IRS (Fig. 8). One major and multiple minor band shifts
were formed with in vitro-translated FOXO1 (lane 1). Only the
major one was specifically supershifted by anti-FOXO1 anti-
body (lane 2 versus 3) and was competed with 250? cold
oligonucleotide to approximately 33% of the FOXO1-IRS
binding (lane 1) but not with mutated oligonucleotide (lanes 4
and 5 versus lanes 6 and 7), indicating that this represented a
FOXO1-IRS complex. Formation of the complex was also re-
duced in the presence of RXR (lane 8), CAR (lane 9), and
PXR (lane 11). When both CAR (or PXR) and RXR were
present in the assay (lanes 10 and 12), the complex formation
was further reduced to 48% of the amount shown in lane 1,
which was similar to the degree of reduction by the cold com-
petitor. Thus, both CAR and PXR became more effective in
reducing the FOXO1-IRS complex by heterodimerizing with
RXR. The degree of complex reduction by the receptors cor-
related with that of the inhibition by them of FOXO1-IRS
activity (Fig. 7). Thus, the nuclear receptors CAR and PXR
directly bind to FOXO1 and reduce formation of the FOXO1-
IRS complex, repressing its activity.
Apparent role of CAR in vivo. The results obtained from
various binding and cell-based transfection assays indicated
that activation of CAR may result in the repression of
FOXO1-regulated genes in liver in vivo. Animal experiments
were performed to investigate whether CAR could regulate
FIG. 7. Repression by CAR and PXR of FOXO1-IRS activity. The
IRS-tk-Luc reporter and pcDNA3/mFOXO1 were cotransfected with
pCR3/mCAR (a) or pcDNA3/mPXR (b) into HepG2 cells. In some
cases, pCMX/hRXR was additionally transfected. Relative activities
were calculated by taking the activity of the cells that were transfected
by only the reporter plasmid in the presence of DMSO as one. Bars
indicate means ? standard deviations.
FIG. 8. CAR and PXR inhibit FOXO1 binding to IRS. In vitro-
translated FOXO1, CAR, hRXR, and PXR were incubated with the
radiolabeled oligonucleotides for wild-type (w) and mutant (m) IRS.
For antibody supershift assays, unlabeled oligonucleotides, rabbit im-
munoglobulin G, and hFOXO1 antibody were cotreated (lane 2 and
3). Excess amounts of unlabeled wIRS and mIRS (25? and 250?)
were used for competitive assays (lanes 4 to 7). Gel shift assays were
performed as described in Materials and Methods. Band intensity of
the IRS-FOXO1 complex was measured with Gene Tools (Hitachi
Software Engineering Co., Ltd.).
VOL. 24, 2004 COREGULATION BY FOXO1 AND RECEPTORS CAR AND PXR7937
gluconeogenesis. Because CAR is a stress response transcrip-
tion factor, its functional role may depend on the liver’s phys-
iological state. Thus, to avoid the risk that these states could
confuse the interpretation of the functional role of CAR in
liver in vivo, two animal models were used for the investiga-
tions. First, Cr1:CD-1(ICR)BR male mice were treated with
TCPOBOP, resulting in the repression of PEPCK1 mRNA,
while CYP2B10 mRNA was induced (Fig. 9a), which was as-
sociated with a slight decrease of the serum glucose levels (168
? 33.2 and 135.7 ? 11.9 mg/dl in the DMSO and TCPOBOP
treatments, respectively). Second, Car?/?and Car?/?mice
were chronically treated with PB, in which PEPCK1 mRNA
was decreased in only the Car?/?mice (Fig. 9b), indicating
that CAR mediated the decrease. In addition, we examined if
this decrease altered serum glucose level in PB-treated Car?/?
and Car?/?mice, finding that the glucose level was only in-
creased in the Car?/?mice. Thus, the lack of CAR-mediated
repression appeared to prevent an increase in serum glucose
level in response to the PB treatment. These results indicate
that the CAR-FOXO1 pathway, in fact, can modulate serum
glucose levels by regulating gluconeogenic enzymes, such as
PEPCK1, and provide insight into understanding the multiple
functional roles of CAR in animals.
An increase in liver microsomal drug-metabolizing activity in
alloxan-treated diabetic rats was first reported in the early 1960s
when drugs such as barbiturates were also found to induce the
same activity (3, 33). It was later confirmed that the increased
drug metabolism activity was due to the induction of enzymes,
such as CYP2B and CYP3A. For example, P450b (presently
called CYP2B1) was increased 25- to 30-fold in the liver micro-
somes of alloxan-treated rats, and insulin treatment restored
P450b to normal levels (56). In the present study, we have dem-
onstrated that the regulatory mechanisms of the insulin response
and of the drug-induced transcription cross paths through direct
interaction between FOXO1 and the nuclear receptors CAR and
PXR. Accordingly, drug metabolism and gluconeogenesis could
be reciprocally coregulated by the transcription factors in re-
sponse to insulin and/or drugs (Fig. 10).
Our study first showed that FOXO1 binds directly to CAR.
While both FOXO1- and insulin-nonresponsive FOXO13Aco-
FIG. 9. CAR regulates the Pepck1 gene. Male mice were treated as
described in Materials and Methods. (a) Liver RNA samples were
individually prepared from three mice for each group of Cr1:CD-
1(ICR)BR mice and were individually subjected to real-time PCR
analysis with specific probes for PEPCK1 and CYP2B10 mRNAs,
respectively. Relative mRNA levels were expressed by taking those
with DMSO as one. Values are means ? standard errors of the means.
An asterisk indicates a P value of ?0.05 and a double asterisk indicates
a P value of ?0.0001 for PEPCK1 and CYP2B10 mRNAs, respectively,
compared with those of the DMSO-treated mice. (b) Liver RNAs were
prepared from 12 mice for each of the Car?/?and Car?/?groups and
were individually subjected to real-time PCR assay. Relative mRNA
levels were expressed by taking those with no PB treatment as one (?
standard errors of the mean). An asterisk indicates P ? 0.02 and a
double asterisk indicates P ? 0.0001 for PEPCK1 and CYP2B10 mR-
NAs, respectively, compared with those of the mice not treated with
PB. (c) Serum glucose levels of the Car?/?and Car?/?mice. Serum
was collected from 20 for each group of the Car?/?and Car?/?mice.
The glucose level was individually measured using the COBAS MIRA
Plus CC Analyzer (Roche Diagnostics) and was averaged. KO, knock-
FIG. 10. Schematic representation of cross-talk. Arrows indicate
activation and coactivation, while stop bars indicate repression and
7938 KODAMA ET AL.MOL. CELL. BIOL.
activated CAR activity, insulin effectively repressed FOXO1
coactivation but not FOXO13Acoactivation. AktCA, but not
AktKN, inhibited FOXO1-CAR activity. Finally, insulin re-
presses the induction of CYP2B10 mRNA in TCPOBOP-
treated mouse primary hepatocytes. CAR accumulates in the
nucleus after treatment with its activators, such as PB,
TCPOBOP, and estradiol (15, 16). FOXO1 is in a position to
coactivate CAR when the receptor accumulates in the nucleus.
By virtue of the PI3K-Akt pathway, insulin inactivates FOXO1
by exporting it from the nucleus to the cytoplasm (2, 34, 59).
When in the nucleus, FOXO1 upregulates CAR activity, aug-
menting the expression of CAR target genes, such as CYP2B
genes. Insulin, on the other hand, downregulates the receptor
by exporting its coactivator FOXO1 from the nucleus, which
could result in the repression of the CAR-regulated genes.
Thus, coregulation of CAR by FOXO1 provides insulin with a
regulatory mechanism to repress hepatic drug metabolism. In
addition to CAR, FOXO1 also activates PXR, suggesting that
insulin repression of drug metabolism via FOXO1 may be a
general pathway. A gel shift assay with the CAR-binding se-
quence NR1 as a probe showed that binding of CAR-RXR to
NR1 was decreased in the presence of FOXO1, suggesting that
CAR interacted with FOXO1 (data not shown). However, a
supershift band that represented a FOXO1-CAR-RXR com-
plex was not detected under the experimental conditions used.
While one can argue that the ternary complex is simply too
large in size to enter into a gel or is not stable enough to be
detected, failing to detect it would also pose the question of
whether FOXO1 forms a stable complex with CAR-RXR-NR1
to activate it. If so, how could FOXO1 activate CAR or RXR?
These molecular mechanisms remain elusive and will be an
interesting research objective for future investigations.
Our studies with CAR-null mice suggested that the receptor
regulated the PB-mediated decrease of PEPCK1 mRNA. Our
present investigations revealed the possible regulatory mecha-
nism for PB-induced repression of gluconeogenic genes (Fig.
10). Following activation by PB, CAR binds to FOXO1 and
prevents its binding to the IRS, which could result in transcrip-
tional repression of the genes that are regulated by IRS, such
as PEPCK1. In addition, future investigations may find the
other gluconeogenic enzymes, including G6P and also key en-
zymes such as pyruvate dehydrogenase kinase 4 in glycolysis, to
be regulated by CAR-FOXO1. This repression mechanism of
FOXO1 by CAR is consistent with previous findings that nu-
clear hormone receptors (e.g., estrogen and androgen recep-
tors, retinoid A receptor, and HNF4) bind to the DNA-binding
domain of FOXO1 (21, 39) and that FOXO1 binding to IRS is
decreased in the presence of the androgen receptor in a gel
shift assay (21). CAR and PXR have now extended the list of
nuclear receptors (e.g., ER, AR, GR, TR, RAR, HNF4, and
PPAR?) that interact with and coregulate FOXO1. In addition
to cell proliferation (22, 26, 40), glucose metabolism (29, 30),
and adipogenesis (27), drug metabolism becomes a novel tar-
get of coregulation by FOXO1 cross-talking with these nuclear
receptors. Because the constitutively active AktCAinhibited
CAR activity in HepG2 cells, endogenous FOXO1 could be
functional in coactivating the receptor. This functionality was
further substantiated by the concomitant reduction of endog-
enous FOXO1 by RNAi and the repression of CAR activity.
FOXO1 is the first factor to coregulate the nuclear receptors in
response to an endogenous signal among various coregulators,
such as SRC-1, NCoR, SMRT, and SHP, that are already
known to regulate CAR and PXR (8, 10, 54). PGC1 is another
signal-responsive coregulator that can activate both CAR and
PXR (41). Whether it regulates the receptor in a signal-depen-
dent manner has not been demonstrated.
While our present study concerned the mouse nuclear recep-
tors and FOXO1, we also examined and confirmed that FOXO1
coactivated human CAR while the human receptor repressed
FOXO1 (data not shown). It has been known for many years that
diabetes elevates the activity of hepatic drug metabolism in hu-
mans. Moreover, PB was clinically used in diabetic patients to
decrease their plasma glucose (20). The CAR and PXR response
elements, such as PBREM and XREM, are conserved in many
human genes, such as CYP2B6 (47), CYP3A4 (5), and UGT1A1
(45). Although a physiological implication of this cross-talk reg-
ulation remains only speculative at the moment, one possible
scenario may be to balance NADPH and NADH consumption.
Insulin acts to decrease NADPH consumption by drug metabo-
lism, while drugs act to repress gluconeogenesis to increase
NADPH supply for drug metabolism. NADPH is an essential
electron donor for cytochrome P450-dependent monooxygenase
activity. The concept of reciprocal coregulation by CAR or PXR
and FOXO1 may have clinical implications in the understanding
of disease prevention, drug-drug interactions, and the develop-
ment of better drug treatments.
1. Argaud, D., S. Halimi, F. Catelloni, and X. M. Leverve. 1991. Inhibition of
gluconeogenesis in isolated rat hepatocytes after chronic treatment with
phenobarbital. Biochem. J. 280:663–669.
2. Biggs, W. H., III, J. Meisenhelder, T. Hunter, W. K. Cavenee, and K. C.
Arden. 1999. Protein kinase B/Akt-mediated phosphorylation promotes nu-
clear exclusion of the winged helix transcription factor FKHR1. Proc. Natl.
Acad. Sci. USA 96:7421–7426.
3. Dixon, R. L., L. G. Hart, and J. R. Fouts. 1961. The metabolism of drugs by liver
microsomes from alloxan-diabetic rats. J. Pharmacol. Exp. Ther. 133:7–11.
4. Dowell, P., T. C. Otto, S. Adi, and M. D. Lane. 2003. Convergence of
proxisome proliferator-activated receptor-? and Foxo1 signaling pathway.
J. Biol. Chem. 278:45485–45491.
5. Goodwin, B., M. R. Redinbo, and S. A. Kliewer. 2002. Regulation of cyp3a
gene transcription by the pregnane x receptor. Annu. Rev. Pharmacol. Toxi-
6. Goodwin, B., E. Hodgson, and C. Liddle. 1999. The orphan human pregnane
X receptor mediates the transcriptional activation of CYP3A4 by rifampicin
through a distal enhancer module. Mol. Pharmacol. 56:1329–1339.
7. Guo, S., G. Rena, S. Cichy, X. He, P. Cohen, and T. Unterman. 1999.
Phosphorylation of serine 256 by protein kinase B disrupts transactivation by
FKHR and mediates effects of insulin on insulin-like growth factor-binding
protein-1 promoter activity through a conserved insulin response sequence.
J. Biol. Chem. 274:17184–17192.
8. Handschin, C., and U. A. Meyer. 2003. Induction of drug metabolism; the
role of nuclear receptors. Pharmacol. Rev. 55:649–673.
9. Hirota, K., H. Daitoku, H. Matsuzaki, N. Araya, K. Yamagata, S. Asada, T.
Sugaya, and A. Fukamizu. 2003. Hepatocyte nuclear factor-4 is a novel
downstream target of insulin via FKHR as a signal-regulated transcriptional
inhibitor. J. Biol. Chem. 278:13056–13060.
10. Honkakoski, P., and M. Negishi. 2000. Regulation of cytochrome P450
(CYP) genes by nuclear receptors. Biochem. J. 347:321–337.
11. Honkakoski, P., I. Zelko, T. Sueyoshi, and M. Negishi. 1998. The nuclear
orphan receptor CAR-retinoid X receptor heterodimer activates the pheno-
barbital-responsive enhancer module of the CYP2B gene. Mol. Cell. Biol.
12. Honkakoski, P., R. Moore, K. A. Washburn, and M. Negishi. 1998. Activa-
tion by diverse xenochemical of the 51-base pair phenobarbital-responsive
enhancer module in the CYP2B10 gene. Mol. Pharmacol. 53:597–601.
13. Honkakoski, P., R. Moore, J. Gynther, and M. Negishi. 1996. Characteriza-
tion of phenobarbital-inducible mouse cyp2b10 gene transcription in primary
hepatocytes. J. Biol. Chem. 271:9746–9753.
14. Kakizaki, S., Y. Yamamoto, A. Ueda, R. Moore, T. Sueyoshi, and M. Negishi.
2003. Phenobarbital induction of drug/steroid-metabolizing enzymes and
nuclear receptor CAR. Biochim. Biophys. Acta 1619:239–242.
VOL. 24, 2004COREGULATION BY FOXO1 AND RECEPTORS CAR AND PXR 7939
15. Kawamoto, T., S. Kakizaki, K. Yoshinari, and M. Negishi. 2000. Estrogen
activation of the nuclear orphan receptor CAR (constitutive active receptor)
in induction of the mouse Cyp2b10 gene. Mol. Endocrinol. 14:1897–1905.
16. Kawamoto, T., T. Sueyoshi, I. Zelko, R. Moore, K. Washburn, and M.
Negishi. 1999. Phenobarbital-responsive nuclear translocation of the recep-
tor CAR in induction of the CYP2B gene. Mol. Cell. Biol. 19:6318–6322.
17. Kawamura, A., Y. Yoshida, N. Kimura, H. Oda, and A. Kakinuma. 1999.
Phosphorylation/dephosphorylation steps are crucial for the induction of
CYP2B1 and CYP2B2 gene expression by phenobarbital. Biochem. Biophys.
Res. Commun. 264:530–536.
18. Kliewer, S. A., J. T. Moore, L. Wade, J. L. Staudinger, M. A. Watson, S. A.
Jones, D. D. McKee, B. B. Oliver, T. M. Willson, R. H. Zetterstrom, T.
Perlmann, and J. M. Lehman. 1998. An orphan nuclear receptor activated by
pregnanes defines a novel steroid signaling pathway. Cell 92:73–82.
19. Kobayashi, K., T. Sueyoshi, K. Inoue, R. Moore, and M. Negishi. 2003.
Cytoplasmic accumulation of the nuclear receptor CAR by a tetratricopep-
tide repeat protein in HepG2 cells. Mol. Pharmacol. 64:1069–1075.
20. Lahtela, J. T., A. J. Arranto, and E. A. Sotanaieni. 1985. Enzyme inducers
improve insulin sensitivity in no-insulin-dependent diabetic subjects. Diabe-
21. Li, P., H. Lee, S. Guo, T. G. Unterman, G. Jenster, and W. Bai. 2003.
AKT-independent protection of prostate cancer cells from apoptosis medi-
ated through complex formation between the androgen receptor and FKHR.
Mol. Cell. Biol. 23:104–118.
22. Machida, S., E. E. Spangenburg, and F. W. Booth. 2003. Forkhead transcription
factor FoxO1 transduces insulin-like growth factor’s signal to p27Kip1 in pri-
mary skeletal muscle satellite cells. J. Cell. Physiol. 196:523–531.
23. Maglich, J. M., C. M. Stoltz, B. Goodwin, D. Hawkins-Brown, J. T. Moore,
and S. A. Kliewer. 2002. Nuclear pregnane X receptor and constitutive
androstane receptor regulate overlapping but distinct sets of genes involved
in xenobiotic detoxification. Mol. Pharmacol. 62:638–646.
24. Manenti, G., T. A. Dragani, and G. Della Porta. 1987. Effects of phenobar-
bital and 1,4-bis[2-(3,5-dichloropyridyloxy)]benzene on differentiated func-
tions in mouse liver. Chem. Biol. Interact. 64:83–92.
25. Matsuzaki, H., H. Daitoku, M. Hatta, K. Tanaka, and A. Fukamizu. 2003.
Insulin-induced phosphorylation of FKHR (FOXO1) targets to proteosomal
degradation. Proc. Natl. Acad. Sci. USA 100:11285–11290.
26. Modur, V., R. Nagarajan, B. M. Evers, and J. Milbrandt. 2003. FOXO
proteins regulate tumor necrosis factor-related apoptosis inducing ligand
expression. Implications for PTEN mutation in prostate cancer. J. Biol.
27. Nakae, J., T. Kitamura, Y. Kitamura, W. H. Biggs III, K. C. Arden, and D.
Accili. 2003. The forkhead transcription factor Foxo1 regulates adipocyte
differentiation. Dev. Cell. 4:119–129.
28. Nakae, J., W. H. Biggs III, T. Kitamura, W. K. Cavenee, C. V. E. Wright,
K. C. Arden, and D. Accili. 2002. Regulation of insulin action and pancreatic
?-cell function by mutated alleles of the gene encoding forkhead transcrip-
tion factor FOXO1. Nat. Genet. 32:245–253.
29. Nakae, J., T. Kitamura, D. L. Silver, and D. Accili. 2001. The forkhead
transcription factor Foxo1 (Fkhr) confers insulin sensitivity onto glucose-6-
phosphatase expression. J. Clin. Investig. 108:1359–1367.
30. Nakae, J., B.-C. Park, and D. Accili. 1999. Insulin stimulates phosphorylation
of the forkhead transcription factor FKHR on serine 253 through a wort-
mannin-sensitive pathway. J. Biol. Chem. 274:15982–15985.
31. Pascussi, J. M., Z. Dvorak, S. Gerbal-Chaloin, E. Assenat, P. Maurel, and
M. J. Vilarem. 2003. Pathophysiological factors affecting CAR gene expres-
sion. Drug. Metab. Rev. 35:255–268.
32. Puigserver, P., J. Rhee, J. Donovan, C. J. Walkey, J. Cliff Yoon, F. Oriente,
Y. Kitamura, J. Altomonte, H. Dong, D. Accili, and B. M. Spiegelman. 2003.
Insulin-regulated hepatic gluconeogenesis through FOXO1-PGC-1? inter-
action. Nature 423:550–555.
33. Remmer, H., and H. J. Merker. 1963. Drug-induced changes in the liver
endoplasmic reticulum; association with drug-metabolizing enzymes. Science
34. Rena, G., A. R. Prescott, S. Guo, P. Cohen, and T. G. Unterman. 2001. Roles
of the forkhead in rhabdomyosarcoma (FKHR) phosphorylation sites in
regulating 14–3–3 binding, transactivation and nuclear targeting. Biochem. J.
35. Rena, G., S. Guo, S. C. Cichy, T. G. Unterman, and P. Cohen. 1999. Phos-
phorylation of the transcription factor forkhead family member FKHR by
protein kinase B. J. Biol. Chem. 274:17179–17183.
36. Rosenfield, J. M., R. Vargas, W. Xie, and R. M. Evans. 2003. Genetic
profiling defines the xenobiotic gene network controlled by the nuclear
receptor pregnane X receptor. Mol. Endocrinol. 17:1268–1282.
37. Sakuma, T., R. Honda, S. Magushi, H. Tamaki, and N. Nemoto. 2001.
Different expression of hepatic and renal cytochrome P450s between the
streptozotocin-induced diabetic mouse and rat. Xenobiotica 31:223–237.
38. Schmoll, D., K. S. Walker, D. R. Alessi, R. Grempler, A. Burchell, S. Guo, R.
Walther, and T. G. Unterman. 2000. Regulation of glucose-6-phosphatase
gene expression by protein kinase B? and the forkhead transcription factor
FKHR. Evidence for insulin response unit-dependent and -independent
effects of insulin on promoter activity. J. Biol. Chem. 275:36324–36333.
39. Schuur, E. R., A. V. Loktev, M. Sharma, Z. Sun, R. A. Roth, and R. J. Weigel.
2001. Ligand-dependent interaction of estrogen receptor-? with members of
the forkhead transcription factor family. J. Biol. Chem. 276:33554–33560.
40. Seoane, J., H. V. Le, L. Shen, S. A. Anderson, and J. Massague. 2004.
Integration of Smad and forkhead pathways in the control of neuroepithelial
and glioblastoma cell proliferation. Cell 117:211–223.
41. Shiraki, T., N. Sakai, E. Kanaya, and H. Jingami. 2003. Activation of orphan
nuclear constitutive androstane receptor requires subnuclear targeting by
peroxisome proliferator-activated receptor ? coactivator-1?. A possible link
between xenobiotic response and nutritional state. J. Biol. Chem. 278:11344–
42. Sidhu, J., and C. J. Omiecinski. 1999. Insulin-mediated modulation of cy-
tochrome P450 gene induction profiles in primary rat hepatocyte cultures.
J. Biochem. Mol. Toxicol. 13:1–9.
43. Smirlis, D., R. Muangmoonchai, M. Edwards, I. R. Phillips, and E. A.
Shephard. 2001. Orphan receptor promiscuity in the induction of cyto-
chromes p450 by xenobiotics. J. Biol. Chem. 276:12822–12826.
44. Sotaniemi, E. A., O. Pelkonen, A. J. Arranto, P. Tapanainen, A. Rautio, and
M. Pasanen. 2002. Diabetes and elimination of antipyrine in man: an analysis
of 298 patients classified by type of diabetes, age, sex, duration of disease and
liver involvement. Pharmacol. Toxicol. 90:155–160.
45. Sugatani, J., H. Kojima, A. Ueda, S. Kakizaki, K. Yoshinari, Q.-H. Gong,
I. S. Owens, M. Negishi, and T. Sueyoshi. 2002. The phenobarbital response
enhancer module in the human bilirubin UDP-glucuronosyltransferase
UGT1A1 gene and regulation by nuclear receptor CAR. Hepatology 33:
46. Sueyoshi, T., and M. Negishi. 2001. Phenobarbital response elements of
cytochrome P450 genes and nuclear receptors. Annu. Rev. Pharmacol. Toxi-
47. Sueyoshi, T., T. Kawamoto, I. Zelko, P. Honkakoski, and M. Negishi. 1999.
The repressed nuclear receptor CAR responds to phenobarbital in activating
the human CYP2B6 gene. J. Biol. Chem. 274:6043–6046.
48. Suwanickul, A., S. L. Morris, and D. R. Powell. 1993. Identification of an
insulin-responsive element in the promoter of the human gene for insulin-
like growth factor binding protein-1. J. Biol. Chem. 268:17063–17068.
49. Thummel, K. E., and J. B. Schenkman. 1989. Effects of testosterone and
growth hormone treatment on hepatic microsomal P450 expression in the
diabetic rat. Mol. Pharmacol. 37:119–129.
50. Ueda, A., H. K. Hamadeh, H. K. Webb, Y. Yamamoto, T. Sueyoshi, C. A.
Afshari, J. M. Lehmann, and M. Negishi. 2002. Diverse roles of the nuclear
orphan receptor CAR in regulating hepatic genes in response to phenobar-
bital. Mol. Pharmacol. 61:1–6.
51. Vander Kooi, B. T., R. S. Streeper, C. A. Svitek, J. K. Oeser, D. R. Powell,
and R. M. O’Brien. 2003. The three insulin response sequences in the
glucose-6-phosphatase catalytic subunit gene promoter are functionally dis-
tinct. J. Biol. Chem. 278:11782–11793.
52. Wang, H., S. Faucette, T. Sueyoshi, R. Moore, S. Ferguson, M. Negishi, and
E. L. LeCluyse. 2003. A novel distal enhancer module regulated by pregnane
X receptor/constitutive androstane receptor is essential for the maximal
induction of CYP2B6 gene expression. J. Biol. Chem. 278:14146–14152.
53. Wei, P., J. Zhang, M. Egan-Hafley, S. Liang, and D. D. Moore. 2000. The
nuclear receptor CAR mediates specific xenobiotic induction of drug me-
tabolism. Nature 407:920–923.
54. Xie, W., and R. M. Evans. 2000. Orphan nuclear receptors: the exotics of
xenobiotics. J. Biol. Chem. 276:37739–37742.
55. Yamamoto, Y., T. Kawamoto, and M. Negishi. 2003. The role of the nuclear
receptor CAR as a coordinate regulator of hepatic gene expression in de-
fense against chemical toxicity. Arch. Biochem. Biophys. 409:207–211.
56. Yamazoe, Y., N. Murayama, M. Shimada, K. Yamauchi, and R. Kato. 1989.
Cytochrome P450 in livers of diabetic rats; regulation by growth hormone
and insulin. Arch. Biochem. Biophys. 268:567–575.
57. Yeagley, D., S. Guo, T. Unterman, and P. G. Quinn. 2001. Gene- and
activation-specific mechanisms for insulin inhibition of basal and glucocor-
ticoid-induced insulin-like growth factor binding protein-1 and phosphoenol-
pyruvate carboxykinase transcription. Roles of forkhead and insulin re-
sponse sequences. J. Biol. Chem. 276:33705–33710.
58. Yoshida, Y., N. Kimura, H. Oda, and A. Kakinuma. 1996. Insulin suppresses
the induction of the CYP2B1 and CYP2B2 gene expression by phenobarbital
in adult rat cultured hepatocytes. Biochem. Biophys. Res. Commun. 229:
59. Zhang, X., L. Gan, H. Pan, S. Guo, X. He, S. T. Olson, A. Mesecar, S. Adam,
and T. G. Unterman. 2002. Phosphorylation of serine 256 suppresses trans-
activation by FKHR (FOXO1) by multiple mechanisms. Direct and indirect
effects on nuclear/cytoplasmic shuttling and DNA binding. J. Biol. Chem.
60. Zhao, H. H., R. E. Herrera, E. Coronado-Heinsohn, M. C. Yang, J. H.
Ludes-Meyers, K. J. Seybold-Tilson, Z. Nawaz, D. Yee, F. G. Barr, S. G.
Diab, P. H. Brown, S. A. W. Fuqua, and C. K. Osborne. 2001. Forkhead
homologue in rhabdomyosarcoma functions as a bifunctional nuclear recep-
tor-interacting protein with both coactivator and corepressor functions.
J. Biol. Chem. 276:27907–27912.
7940KODAMA ET AL.MOL. CELL. BIOL.