nucleic acids research, 2007, doi:10.1093/nar/gkm090

Nucleic Acids Research 2007 1ndash14doi101093nargkm090

Transactivation of a DR-1 PPRE by a humanconstitutive androstane receptor variant expressedfrom internal protein translation start sitesMatthew A Stoner1 Scott S Auerbach2 Stephanie M Zamule1

Stephen C Strom3 and Curtis J Omiecinski1

1Center for Molecular Toxicology amp Carcinogenesis The Pennsylvania State University University ParkPA 16802 USA 2Department of Pharmacology University of Washington Seattle WA 98195 USA and3Department of Pathology University of Pittsburgh Pittsburgh PA 15261 USA

Received December 18 2006 Revised January 10 2007 Accepted February 1 2007

ABSTRACT

Downstream in-frame start codons produce amino-terminal-truncated human constitutive androstanereceptor protein isoforms (NCARs) The NCARsare expressed in liver and in vitro cell systemsfollowing translation from in-frame methionine AUGstart codons at positions 76 80 125 128 168 and265 within the full-length CAR mRNA The resultingCAR proteins lack the N-terminal DNA-bindingdomain (DBD) of the receptor yielding NCARvariants with unique biological function Althoughthe NCARs maintain full retinoid X receptor alpha(RXRa) heterodimerization capacity the NCARsare inactive on classical CAR-inducible directrepeat (DR)-4 elements yet efficiently transactivatea DR-1 element derived from the endogenousPPAR-inducible acyl-CoA oxidase gene promoterRXRa heterodimerization with CAR1 CAR76 andCAR80 isoforms is necessary for the DR-1 PPREactivation a function that exhibits absolute depen-dence on both the respective RXRa DBD and CARactivation (AF)-2 domains but not the AF-1 or AF-2domain of RXRa nor CARrsquos DBD A new model ofCAR DBD-independent transactivation is proposedsuch that in the context of a DR-1 peroxisomeproliferator-activated response element only theRXRa portion of the CAR-RXRa heterodimer bindsdirectly to DNA with the AF-2 domain of tetheredCAR mediating transcriptional activation of thereceptor complex

INTRODUCTION

CAR (NR1I3) is a member of the nuclear receptorsuperfamily comprised of 48 members in humans (1)

The structural features of CAR include a clearly identifi-able DNA-binding domain (DBD) but the lack of aconventional amino-terminal AF-1 motif found in mostother nuclear receptors (2) While nuclear receptors aretypically activated by chemical ligands CAR maintainsa high constitutive activity limited by the cytosolicsubcellular localization of the unactivated protein (3)CAR appears to be retained in a cytoplasmic complexwith phosphatase 2A (PP2A) heat shock protein 90(Hsp90) (4) and a cytosolic CAR retention proteintermed CCRP (5) Ligand and non-ligand activatorssuch as 6-(4-chlorophenyl)imidazo[21-b]thiazole-5-carbaldehyde O-(34-dichlorobenzyl)oxime (CITCO) (6)and phenobarbital respectively (3) induce CARcytoplasmic-to-nuclear translocation through a pathwayinvolving an okadaic-acid-sensitive phosphatase (37)Early studies with CAR revealed that certain andros-

tane derivatives bind directly to the receptor and functionto inhibit its transactivation These repressive ligandswere termed lsquoinverse agonistsrsquo (8) Furthermore inverseagonist-bound CAR can be reactivated by exposingthe receptor complex to specific inducer chemicals(910) such as 14-bis[2-(35-dichloropyridyloxy)]benzene(TCPOBOP) a ligand activator of mouse CAR (11) andCITCO the human CAR-specific agonist ligand (6)CAR targets include genes encoding phase I and phase

II drug metabolizing functions as well as drug transportgenes [reviewed in (12)] The target gene promoters oftencontain consensus DNA response elements includingdirect repeat (DR)-2 (13) DR-3 DR-4 (14) DR-5 (2)and everted repeat (ER)-6 or ER-8 (15) motifs CAR maybind the DNA response elements as a monomer or as aheterodimer with retinoid X receptor alpha (RXR)(1516) In some cases at least two separate elementsmay cooperate in a single gene promoter to furtherenhance CARRXR-mediated transcriptional activation(1718) Like many nuclear receptors CAR-mediatedtransactivation also depends on its association with

To whom correspondence should be addressed Tel 814-863-1625 Fax 814-863-1696 Email cjo10psuedu

2007 The Author(s)

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (httpcreativecommonsorglicenses

by-nc20uk) which permits unrestricted non-commercial use distribution and reproduction in any medium provided the original work is properly cited

Nucleic Acids Research Advance Access published March 13 2007

nuclear co-regulator proteins including the co-activatorsGRIP-1 (19) SRC-1 (20) and PGC-1 (21)In addition to the reference form of human CAR

termed CAR1 we and others have characterized alter-natively spliced CAR transcripts that exist in humantissues (22ndash24) At least one of these CAR variantsCAR3 possesses the unique biological property of beinga ligand-activated receptor in contrast to the constitu-tively active nature of CAR1 (14) The results presentedin the current investigation extend the complexity ofCAR regulation by demonstrating that the receptorrsquosexpression is post-transcriptionally modulated throughthe use of internal translation start sites resulting in thegeneration of truncated CAR proteins with alteredbiological function Our data also suggest a new modelfor CARndashDNA interaction whereby RXRa binds directlyto a DR-1 peroxisome-proliferator-activated responseelement (PPRE) as a heterodimer with tethered CARsuch that only RXRa makes direct DNA contact withthe tethered CAR AF-2 allowing for transcriptionalactivation

EXPERIMENTAL PROCEDURES

Cell lines chemicals and biochemicals

COS-1 African green monkey kidney epithelial cell line(ATCC Manassas VA) was cultured in DulbeccorsquosModified Eagle Medium (Invitrogen Gaithersburg MD)plus 10 fetal bovine serum 15 gl sodium bicarbonateHEPES L-glutamine and penicillinndashstreptomycin5a-Androstan-3a-ol (androstanol) and 9-cis-retinoic acid(RA) (Sigma St Louis MO) CITCO (BioMol PlymouthMeeting PA) and Wyeth 14643 were dissolved indimethylsulfoxide (DMSO) (Sigma St Louis MO)Enriched primary human hepatocyte cultures plated oncollagen were obtained through the Liver TissueProcurement and Distribution System Pittsburghfunded by NIH Contract NO1-DK-9-2310 Cells wereplaced in fresh Williamrsquos E media containing 1penicillinndashstreptomycin 1 HEPES 20 mM glutamine25 nM dexamethasone 10 nM insulin 1 linoleic acidBSA 5 ngml selenious acid and 5 mgml transferrin andoverlayed mL BD MatrigelTMmL BD MatrigelTM

Basement Membrane Matrix (BD Biosciences San JoseCA) as described previously (25) Adenovirus expressinghuman CAR was produced using a commercially availablesystem AdEasy (Stratagene Cedar Creek TX) Primaryhepatocytes from one individual were infected withAdenovirus-CAR (Adv-CAR) for 24 h cells were scrapedin Laemmlirsquos loading buffer and a small aliquot wasresolved by SDS-PAGE transferred to PVDF membraneand CAR detected by immunoblotting Lentiviral expres-sion plasmids pCDH1-MCS1-EF1-copGFP and pSIH1-H1-copGFP and packaging plasmid mix pPACKH1were purchased from System Biosciences (Mountain ViewCA) Production of lentiviral particles and subsequentinfection of target cells were performed essentiallyaccording to the manufacturerrsquos protocols except thatLipofectamine 2000 (Invitrogen Carlsbad CA) wasused for transfection of HEK-293T packaging cells

Pseudoviral supernatant harvested from the packagingcells was either used directly for target cell infection or wasconcentrated prior to infection by 10 PEG-8000 (SigmaSt Louis MO) precipitation Primary human hepatocyteswere infected with equal volumes of appropriate lentiviralsupernatant in the presence of 6 mgml polybrene (SigmaSt Louis MO) and media was replaced the followingday

Oligonucleotides and plasmids

Oligonucleotides and PCR primers were synthesized byIntegrated DNA Technologies (IDT) (Coralville IA)PCR primers and EMSA oligonucleotides used are listedin Supplementary Data A consensus DR-4 probe used inEMSA competition experiments contained the NR1sequence from CYP2B6 PBREM and the sense strandcontained the classical DR-4 element AGTTCA(N)4AGTTCA Reporter plasmids contained four copies ofthe rat acyl-CoA-oxidase (rAox) PPRE (26) a PCRfragment from rat cellular retinol-binding protein IIpromoter (rCRBPII) (27) and three copies of an optimizedAGTTCA DR-4(14) Complementary oligonucleotideswere annealed phosphorylated and inserted in the SmaIsite of ptk-Luc derived from pGL3-Basic vector(Promega Madison WI) which was modified by insertionof the minimal thymidine kinase (tk) element frompBLCAT prAox-1198-463-tk-Luc contains a singlecopy of the PPRE (DR-1 tgacctttgtcct) in its endogenouscontext encompassed by the region between 1198 and463 bp upstream of the rAox gene transcription start siteand was amplified using rat genomic DNA as a templatePrimer sequences employed were as previously described(28) but with slight modifications the sense primer(50-GGT ACC GGT ACC CCA GTA GAA CCT TGTTCA GG-30) and antisense primer (50-GGT ACC GCTAGC CAG GGT CTC GGG CGG AGT GAA G-30)contained the underlined KpnI and NheI restrictionenzyme sites respectively The 754-bp amplicon was gel-purified restriction enzyme digested gel-purified andligated into the ptk-Luc vector (Promega Madison WI)Plasmids were prepared using QIAfilter Plasmid Maxi Kit(Qiagen Valencia CA) and sequenced

Coupled in vitro transcription translation

Rabbit reticulocyte lysate and wheat germ expressionsystems were used essentially according to the manufac-turerrsquos (Promega Madison WI) protocols DNAtemplates were in the form of plasmids containingCAR-coding sequence downstream of T7 promoteror PCR products of T7 promoter-tagged CAR fragmentsIn some experiments [35S]methionine (ICN) was incorpo-rated while in other experiments a polyclonal antibodyagainst human CAR (23) was used to perform westernblotting

Western-immunoblotting analysis

Cells were harvested directly in 1 Laemmlirsquos loading dyeand boiled for 5min Alternately cells were scraped inradioimmunoprecipitation assay (RIPA) buffer proteinconcentrations were measured by Bradford method

2 Nucleic Acids Research 2007

and then samples were mixed with Laemmlirsquos buffer(BioRad Hercules CA) For western blots examiningexpression of endogenous CAR variants in primaryhuman hepatocytes the cells from four different humandonors were cultured as described above harvested byscraping in Trizol reagent (Invitrogen Carlsbad CA) andprotein samples were isolated according to the extendedmanufacturerrsquos protocol One-hundred and fifty micro-grams of total protein were dissolved in loading buffer andwere resolved by SDS-PAGE and separated proteins weretransferred to PVDF membrane (BioRad Hercules CA)then probed with primary antibodies against humanCAR (generated in our laboratory) or RXRa (D-20)(Santa Cruz) Secondary antibody goatndashanti-rabbit-IgG-HRP (Santa Cruz Biotechnology Santa Cruz CA) wasused at 15000 dilution Blots were exposed to LumiLightchemiluminescent substrate (Roche Applied ScienceIndianapolis IN) and placed on autoradiography film(American X-Ray amp Medical Supply Sacramento CA)PVDF membrane was stained with Ponceau S in diluteacetic acid and briefly washed in de-ionized water untildistinct protein bands could be distinguished Pink-stainedblots were scanned and images were changed to grayscalefor publication

Transient transfections

p(NR1)5-Tk-Luc and pCMV-Renilla expression plasmid(Invitrogen Carlsbad CA) (for normalization of transfec-tion efficiency) and pCDNA31-CAR expression plasmidswere transiently transfected into COS-1 cells usingFugene6 (Roche Applied Science) for 12ndash24 hTransfection cocktails were prepared essentially accordingto the manufacturerrsquos recommendations except that1 phosphate-buffered saline (PBS) was substituted forserum-free medium in the transfection cocktail Cellswere harvested with 50 ml 1 passive lysis buffer(Promega Madison WI) and soluble protein wasextracted and assayed for luciferase activity usingStop-lsquoNrsquo-Glow system (Promega) and dual auto-injectluminometer (Turner Biosystems Sunnyvale CA) Assayreagents were diluted 11 with Tris-buffered saline and25 ml of each reagent was used to measure Firefly andRenilla luciferase activity in 40 ml of cell lysateExperiments were repeated at least two times andrepresentative results are shown as mean standarderror for three or four replicates per treatment group

Preparation of COS-1 cell nuclear extracts

Nuclear extracts were prepared as described (29) Brieflytransfected COS-1 cells were harvested in 1 Promegalysis buffer and allowed to swell on ice for 15mincentrifuged at 14000 g for 1min at room temperatureand a pellet containing crude nuclei was isolatedApproximately five pellet volumes of 1 lysis buffersupplemented with 500mM KCl (high-salt buffer) wereadded to the pellet and further incubated on ice for 30minto 1 h with frequent vortexing Nuclei in high-salt bufferwere centrifuged at 14000 g for 1min at room tempera-ture and aliquots of nuclear proteins in supernatant werestored at 808C for use in EMSA

Electrophoretic mobility shift assays (EMSAs)

A 20ndash30 ml EMSA reaction mixture contained75ndash150mM KCl (a balance of low-salt and high-saltbuffers used in the nuclear extraction protocol) 5 mg ofcrude nuclear protein 1 mg poly(dI-dC) (Roche MolecularBiochemicals Indianapolis IN) with or without unla-beled competitor oligonucleotide and 10ndash100-fmol-labeledprobe Oligonucleotides were end labeled by incubationwith T4-polynucleotide kinase (Promega Madison WI)and [g32P]-ATP (ICNMP Biomedicals Solon OH)Reactions were incubated at room temperature for15min and proteinndashDNA complexes were resolved by5 PAGE and specific DNAndashprotein complexes wereobserved as more slowly migrating complexes in the gelGels were dried under vacuum and exposed to X-ray film

CAR immunofluorescence microscopy

Cells were grown on Lab-Tek 4-well chamber slides(Nunc Naperville IL) and fixed in 208C methanolair-dried then permeabilized and rehydrated in PBSsupplemented with 03 Tween-20 (PBS-Tween) Cellswere blocked in goat serum washed and incubated withrabbit polyclonal CAR antibody (150 dilution) Primaryantibody was washed from cells and fluorescein isothio-cyanate (FITC)-conjugated goatndashanti-rabbit IgG(CALTAG Laboratories Burlingame CA) diluted 1200in antibody dilution buffer was placed on the cells for 2 hat room temperature Finally slides were extensivelywashed in PBS-Tween blotted dry covered with ProlongAnti-fade mounting solution with 406-diamidino-2-phenylindole (DAPI) DNA-labeling reagent (MolecularProbes Eugene OR) and sealed with a coverglassImmunofluorescence of CAR protein was observed witha Nikon inverted fluorescence microscope (Nikon USAMelville NY) Image capture was performed with SpotRTsoftware and digital camera (Diagnostic InstrumentsSterling Heights MI)

Real-time PCRmethods and data analysis

RNA was isolated from primary human hepatocytes usingTRIzol Reagent (Invitrogen Carlsbad CA) and con-verted to cDNA using High Capacity cDNA Archive Kit(Applied Biosystems Foster City CA) both according tothe manufacturerrsquos protocols Real-time PCR was per-formed using Assays-on-Demand Gene ExpressionProducts (Applied Biosystems Foster City CA) accord-ing to the manufacturerrsquos instructions with minormodifications and run on an ABI 7300 Real-time PCRSystem (Applied Biosystems) Real-time PCR data wereanalyzed using the CT method (30)

RESULTS

Initial identification of truncated CAR proteins

Primary human hepatocytes express multiple nuclearreceptors In Figure 1A we present results from westernimmunoblot assays performed on whole-cell extractsderived from primary hepatocytes of four individualhuman donors using rabbit polyclonal antibodies directed

Nucleic Acids Research 2007 3

against the carboxyl-terminus of CAR and RXRaMultiple protein bands were detected for CAR At leastthree immunoreactive CAR bands were observed OneCAR product migrated at the predicted molecular weightof the full-length protein (37 kDa) A second prominentimmunoreactive band displayed an apparent molecularweight in excess of 25 kDa and a third band migrated at20 kDa Similar results were obtained with proteinextracts analyzed from liver tissues of other indivi-duals (23) The CAR immunoreactive bands detectedco-migrated with positive control extracts from HEK-293T cells transfected with expression plasmids for CAR1and CAR76 that were run concurrently Known splicevariants of CAR include CAR2 containing a four aminoacid (4 aa) insertion (SPTV) between exons 6 and 7CAR3 which possesses a 5 amino acid (5 aa) insertion(APYLT) between exons 8 and 9 a transcript including

insertion of both SPTV and APYLT (DBL) and a variantpossessing a partial deletion of exon 7 (7) that results inthe loss of 39 aa from the ligand-binding domain of CAR(23) Each of these variant proteins is therefore predictedto migrate well above the 25 kDa range In COS-1 cellstransiently transfected with expression plasmids for CARand CAR splice variants multiple immunoreactive CARproducts of similar molecular mass were also observedin western immunoblot assays conducted using an anti-body directed against the carboxyl-terminus of CAR(Figure 1B)

We performed computer alignments using the ClustalWtool (httpwwwebiacukclustalw) (31) to demonstratethe conservation of human mouse and rat CAR mRNAsequences (Figure 2A) The numbers indicated above thesequences are relative to the predicted initial start oftranslation at codon 1 in the human sequence Numbers tothe right denote relative sequence within the coding exonsof CAR It should be noted that mouse and rat CARinclude 30 extra nucleotides coding for 10 additionalamino-terminal amino acids as compared with humanCAR such that codon 76 in human CAR equates tocodon 86 etc in mouserat CAR respectively Thealignments of (human) codons 76 and 80 were most highlyconserved across species while more variability existed for

A

B

Figure 1 Western immunoblot analyses (A) Expression of CAR-variant proteins and RXRa protein in human primary hepatocytesfrom four donors Lanes 1ndash4 150 mg whole-cell protein extract fromeach of four donors lanes 5 and 6 are positive control whole-cellprotein extracts from HEK-293T human embryonic kidney cell linestransfected with expression plasmids for CAR 1 or CAR 76 and treatedwith 1 mM MG-132 proteasome inhibitor to augment CAR expressionPonceau S staining was performed to indicate total protein loading ineach lane (B) CAR protein expression in COS-1 cells transientlytransfected with CAR and CAR splice variant expression plasmidsLane 1 empty vector transfection (V) lane 2 CAR wild-type (WT)lane 3 CAR4aa (SPTV amino acids inserted between exons 67)lane 4 CAR5aa (APYLT amino acids inserted between exons 78)CAR-DBL (four and five amino acid insertion) and CAR7 (partialdeletion of exon 7 coding for 39 amino acids) Embedded asterisksdenote unexpected lower molecular weight CAR products

A

B

Figure 2 ClustalW alignment of human mouse and rat CAR mRNAsequences (A) Numbers above sequences denote NetStart10-predictedin-frame internal protein translation start sites Numbers abovesequences represent codon numbering based on human sequenceNumbers listed to the side indicate relative position of nucleotideswithin coding sequence asterisks denote sequence conservation acrossspecies (B) CAR protein and mRNA sequences alignment PredictedCAR translation start sites are denoted by arrows at 1 76 80 125 128168 and 265 For full-length CAR protein translation begins from astart codon 1 in exon 2 and continues to codon 348 in exon 9


codons 125 and 128 The human CAR-coding sequencewas analyzed with the web tool NetStart 10 (httpwwwcbsdtudkservicesNetStart) (32) to predict internalprotein translation start sites on a 0ndash10 scale A call of 05or higher was assigned by the program as a high-probability translation start site Codon 125 is includedin the figure as well as it scored close to 05 within the04ndash05 range In Figure 2B a simple alignment of humanCAR protein and exons was performed and revealed thatall potential internal translation start sites remove theamino-terminal (N-terminal) DBD of the receptor andthat most fall in the exon 3ndash5 region Thereforesubsequent analyses were focused on the less severelydeleted variants arising from potential translation startsites at codons 76 80 125 128 and 168

Truncated CAR proteins transactivate a DR-1 PPRE

To quickly test whether predicted downstream andin-frame AUG start codons in CAR would drive trans-lation of shorter protein products the first AUG wasmutated to AAG to produce a lysine (K) instead ofmethionine (M) at position 1 and the derived constructwas termed CAR M1K Assuming CAR mRNA istranslated to protein by a lsquocap-dependentrsquo scanningribosome process there is an expectation that increasedtranslation would then occur from the use of downstreamAUG start sites in the CAR M1K template Resultspresented in Figure 3A clearly show that overexpression ofCAR1 by Adv-CAR-mediated infection of primarycultures of human hepatocytes or expression of CAR1and CAR M1K mutant in wheat germ lysates producedmultiple protein products as detected by immunoblottingwith an antibody designed against the C-terminus ofCAR In lanes 1 and 3 major CAR products wereobserved in close alignment with the 37 kDa and betweenthe 25 and 37 kDa markers respectively In lane 4introduction of the M1K mutation resulted in a dramaticincrease in the expression of truncated CAR protein likelyfrom initiation at one or both of the codons 76 and 80The very low level of apparent full-length CAR in theCAR M1K lane was not observed in subsequent experi-ments but in this overexposed film the band may havearisen from use of the less common internal translationstart at leucinecodon 8 or more likely from slight spill-over from the CAR1 lane during gel loading Once it wasestablished that CAR M1K expression resulted in expres-sion of truncated protein products the potential tran-scriptional activities of the variant proteins were tested intransient transfection assays using assorted previouslyidentified nuclear receptor consensus response elementsWe hypothesized that the truncated forms of CAR mayexert dominant negative or enhanced activity on specificelements Previous studies in this and other laboratoriesdemonstrated that CAR1 primarily activates reportergenes driven by DR-4 or DR-5 elements (214) andexhibits particularly strong activation of a DR-4 thatcontains the perfect half-site AGTTCA (15) In our initialtransient transfection assays we screened a number ofpotential response elements for activity of CAR1 andCAR M1K in COS-1 cells Surprisingly CAR1 and CAR

M1K similarly activated the special DR-1 element rAoxPPRE containing the sequence (TGACCT(N)1TGTCCT) as well as an element containing a series of DR-1motifs derived from the rCBPII promoter where CARM1K exhibited greater responsiveness than did CAR1(Figure 3B) These marked transcriptional activationresponses were dependent on the presence ofco-transfected RXRa and also exhibited apparent cellline dependency as they were retained in HepG2 andCOS-1 cells but not readily detected in HuH7 A549Hepa1c1c7 HEK-293A or MCF-7 cells (data not shown)

A

B

Figure 3 Western immunoblot assay (A) CAR full-length proteinand translational variants are expressed in Adv-CAR-infectedhuman hepatocytes and in in vitro wheat germ expression systemLane 1 human primary hepatocytes infected with adenovirusexpressing wild-type CAR mRNA (lanes 2ndash4 wheat germ lysates)lane 2 empty vector control lane 3 pCDNA31()CAR 1 lane 4pCDNA31()CAR M1Kmdashstart methionine mutated to lysineNumbers to the left of figure indicate predicted internal translationvariantsrsquo relative molecular weights (Mr) (B) CAR 1 and CAR M1Kdifferentially transactivate DR-4 and DR-1 elements COS-1 cells weretransiently transfected with 100 ng of reporters p(DR-4)3-tk-Lucp(rCRBPII)-tk-Luc or p(rAox-PPRE)4-tk-Luc 50 ng RXRa and CARexpression plasmids and 10 ng pRL-CMV


The basis for the cell specificity of the transcriptionalresponses is not yet understood however the cell selectivepresenceabsence of specific co-regulator proteins is likelyresponsible NCAR functional characterization studieswere continued in COS-1 cells since these cells weremarkedly activated in a highly reproducible manner andexhibited the highest transfection efficiencies of all the celllines tested

NCARs activate transcription in the absence of a DBD

Given the initial success with the CARM1Kmutation andthe expression of truncated CAR proteins we testedwhether expression of the lower molecular weight CARproducts could be ablated by mutating additional putativedownstream start codons at positions 76 and 80 Indeedexpression of CAR M1K in wheat germ and rabbit

reticulocyte lysate systems in the presence of[35S]-methionine (Figure 4A) (lanes 2 and 7) resulted inthe predominant production of a lower molecular weightspecies When CAR76 template was overexpressed(lanes 3 and 8) the product co-migrated with the proteinproduced by CAR M1K suggesting that codon 76 may bea biologically relevant internal translation start sitein CAR However expression of CAR 1M76K (lanes 4and 9) revealed that mutation of codon 76 alone in thecontext of the full-length template could not abolishthe lower molecular weight product Subsequent dualmutation of both codons 76 and 80 in the form of CAR1M76KM80K (lanes 5 and 10) completely eliminatedexpression of the truncated form of CAR supporting amodel of ribosome scanning from the 50- lsquocaprsquo structureof CAR mRNA resulting in the use of codon 80 as an

Figure 4 Truncated forms of CAR are expressed and activate transcription through a special DR-1 element (A) SDS-PAGE of coupled in vitrotranscriptiontranslation of [35S]-methionine-labeled CAR in wheat germ lysate (lanes 1ndash5) and rabbit reticulocyte lysate (lanes 6ndash10) Lanes 1 and 6pCDNA31()CAR 1 lanes 2 and 7 pCDNA31()CAR M1K lanes 3 and 8 pCDNA31()CAR 76 lanes 4 and 9 pCDNA31()CAR 1 M76Kand lanes 5 and 10 pCDNA31()CAR 1 M76KM80K (B) Differential activation of p(DR-4)3-tk-Luc and p(rAox-PPRE)4-tk-Luc by amino-terminally truncated forms of CAR (C) Activation of p(rAox-1198-463)-tk-Luc a genomic promoter fragment containing a single copy of thePPRE element by CAR 1 and CAR 76 COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ng pCDNA31() empty vectoror pCDNA31()CAR plasmids 50 ng pCDNA31(thorn)RXRa and 10 ng pRL-CMV (for transfection normalization) and assayed for luciferase activity24ndash48 h post-transfection


initiation site only when codon 76 was mutatedIn wheat germ lysate and in longer exposures of rabbitreticulocyte lysate even lower molecular weightCAR products were detected likely attributable tofurther alternative translation from codons 125 and 128(data not shown)

To test the transcriptional activity of the various CARconstructs COS-1 cells were transiently transfected withp(DR-4)3-tk-Luc and p(rAox-PPRE)4-tk-Luc CAR1 andtranslational variants of CAR1 including CAR M1KCAR76 CAR80 CAR125 CAR128 or CAR168 wereexamined all in the presence of co-transfected RXRaexpression plasmid (Figure 4B) In Figure 4C CAR1 andCAR76 activated transcription of a non-oligomerizedPPRE from the rAox gene promoter contained in a750-bp DNA fragment of rat genomic DNA and placedupstream of tk-Luc The magnitude of the response andpattern of activation closely resembled the activationobtained when using four copies of the PPRE upstream oftk-Luc To summarize the results obtained CAR1 wasactive on DR-4 and DR-1 elements but successivedeletion of CAR protein to position 80 redirected fullactivation to only the DR-1 element When CARexpression was forced from downstream codons at 125128 and 168 the yielded proteins were without demon-strable transactivation potential on either DR-4 or DR-1elements

Results detailed in Figure 5 demonstrate that CAR1transcriptional activation of the DR-4 element is RXRaindependent and that CAR76 and CAR80 were comple-tely inactive on this element regardless of RXRaco-transfection To ascertain the extent to which CARcould be deleted on its N-terminus and still retaintranscriptional activity on a DR-4 motif a number ofother deletion mutants of CAR were tested (Figure 5A)Deletion of 10 aa from the N-terminus of CAR(CAR11) resulted in complete loss of activity howeverco-transfection of RXRa restored transactivation ofCAR11 on the DR-4 element near to the levels seenwith CAR1 Further truncation of CAR (CAR14 toCAR125) produced constructs that were incapable ofactivating DR-4 response element-driven luciferase geneexpression CAR11 includes the first cysteine residuewithin the first zinc finger of the receptor predicted to bean important determinant for DNA binding Surprisinglywhen the same successively deleted CAR constructs wereassayed for activity on the DR-1 element in the presenceof excess RXRa all the constructsmdashthrough CAR103maintained full transcriptional activity while CAR113and CAR125 were inactive (Figure 5B) Recently thecrystal structure of CARRXRa was reported derivedfrom a minimal CARRXRa dimerization interface thatincluded amino acids 103ndash348 of CAR (33) Our resultssimilarly suggest that proper heterodimer formationbetween RXRa and CAR residues 103ndash348 is importantfor CAR transcriptional activity on a DR-1 element

CAR activity on DR-1 PPRE is RXR-dependent

Since the first 102 aa of CAR were dispensable fortranscriptional activity on a DR-1 we probed the

importance of other domains for this activity We deleteda portion of the CAR C-terminus to remove 8 aa thatcontained the receptorrsquos AF-2 domain The resultsindicated that even in the presence of RXRa CAR1iAF-2 was devoid of transcriptional activity on theDR-1 element (Figure 6A) Although CAR activity on theDR-1 did not require the first third of the protein it wasdependent on an intact AF-2 domain as well asco-transfected RXRa To further delineate the RXRa-dependent transactivation of the DR-1 by CAR and theNCARs a series of RXRa domain deletion constructswere co-transfected into COS-1 cells (Figure 6B) Deletionof RXRaDBD amino acids 135ndash161 resulted in a lossof CAR1 and NCAR transcriptional activity likelydue to the loss of RXRarsquos ability to directly bindDNA However deletion of the AF-1 and AF-2 functionsof RXRa alone or in combination had little effect onthe transactivation ability of CAR1 CAR76 or CAR80to activate the DR-1 element (Figure 6C) These resultsprovided additional support for the concept thatdirect binding of RXRa to DNA was a necessarycomponent for CARrsquos activity at the DR-1 element

A

B

Figure 5 Transient transfection assays (A) An intact DNA-bindingdomain is necessary for CAR transcriptional activity on a DR-4element COS-1 cells were transfected with 100 ng p(DR-4)3-tk-Luc50 ng empty vector or RXRa 50 ng empty vector or CAR constructsand 10 ng pRL-CMV (B) CAR amino acids 103ndash348 andco-transfected RXRa are required for full transcriptional activityon a DR-1 element COS-1 cells were transfected with 100 ngp(rAox-PPRE)4-tk-Luc 50 ng empty vector or RXRa 50 ng emptyvector or CAR constructs and 10 ng pRL-CMV


but that CARrsquos AF-2 domain provides the requisiteinteraction with co-regulatorsSince RXRa homodimers bind well to DR-1 elements

we also tested whether RXRa deletion mutants werethemselves transcriptionally active on p(rAox-PPRE)4-tk-Luc in the presence of the RXRa ligand RA (Figure 6D)The results obtained showed that wild-type and AF-1-deleted RXRa were fully activated by ligand while anydeletion of AF-2 or DBD yielded transcriptionallyinactive proteinsOur initial data indicated that CAR or NCAR inter-

action with RXRa was necessary for CAR-mediatedtransactivation through the DR-1 PPRE Co-transfectionexperiments were performed with two different point-mutated RXRa plasmids to demonstrate that

CARRXRa dimerization was absolutely required fortranscriptional activity (Figure 7A) Use of RXRa Y397Aa mutation in the human receptor corresponding to a well-characterized heterodimerization-deficient Y402A muta-tion in mouse RXRa (34) completely ablated CAR-mediated transactivation on the PPRE However thisheterodimerization-mutant RXRa was still fully func-tional as a homodimer after addition of RA Anadditional mutation of RXRa yielded L430F a mutantdefective in homodimerization (35) as demonstrated hereby the lack of induction of luciferase activity in thepresence of RA However the L430F mutant mediatednearly full induction of reporter activity when CAR1 wasco-transfected Further immunofluorescence microscopydemonstrated transfected CAR1 was expressed primarily

A B

C D

Figure 6 Transient transfection assays (A) CAR AF-2 is required for transactivation through a DR-1 element COS-1 cells were transfected with100 ng p(rAox-PPRE)4-tk-Luc 50 ng empty vector or RXRa 50 ng empty vector or CAR constructs and 10 ng pRL-CMV (B) CAR activation ofp(rAox-PPRE)4-tk-Luc is RXRa DNA-binding domain dependent COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ngpCDNA31() empty vector or pCDNA31()CAR plasmids 50 ng pCDNA31() pCDNA31(thorn)RXRa or pCDNA31(thorn)RXRaDBD and 10 ngpRL-CMV (for transfection normalization) and assayed for luciferase activity 24ndash48 h post-transfection (C) CAR transactivation of p(rAox-PPRE)4-tk-Luc is independent of RXRa AF-1 and RXRa AF-2 COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ngpCDNA31() empty vector or pCDNA31()CAR plasmids 50 ng pCDNA31() pCDNA31(thorn)RXRa pCDNA31(thorn)RXRaAF-1 orpCDNA31(thorn)RXRaAF-2 and 10 ng pRL-CMV (for transfection normalization) and assayed for luciferase activity 24ndash48 h post-transfection(D) RXRa deletion constructs DBD- and AF-2-dependently transactivate p(rAox-PPRE)4-tk-Luc in response to RXR ligand 9-cis-retinoic acid (RA)COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ng pCDNA31(thorn) or pCDNA31(thorn)RXRa or RXRa domain deletionvariants and 10 ng pRL-CMV (for transfection normalization) for 24 h then treated with dimethyl sulfoxide (DMSO) solvent control or 1 mM9-cis RA for 24 h and assayed for luciferase activity


in the nucleus of COS-1 cells either in the absence orpresence of co-transfected RXRa (Figure 7B) In contrastCAR76 and CAR80 expression was evenly distributedbetween the cytoplasmic and nuclear compartments inthe absence of co-expressed RXRa but became primar-ily nuclear in the presence of excess RXRa Whenthe RXRaY397A heterodimerization mutant wasco-expressed with CAR1 CAR76 or CAR80 the dis-tribution of the CAR proteins was similar to resultsobtained with co-transfected empty vector alone

CAR ligands modulate variant protein activity

CAR exhibits high constitutive activity in transfected cellspresumably because proteins involved in cytoplasmicsequestration of the protein in primary hepatocytes aremissing from cell lines (5) Consequently early studiesidentified so-called inverse agonists of CAR such asandrostanol which could repress the constitutive transac-tivation potential of CAR (8) When CAR1 was trans-fected with RXRa p(DR-4)3-tk-Luc activity was highlyinduced as predicted while CAR76 and PPARa were

inactive on this element (Figure 8A) A titration ofandrostanol from 05 to 8 micromolar (mM) indicatedthat a maximal inhibition of CAR1 constitutive activitywas achieved at 2 mM A similar androstanol treatmentwas repeated on cells transfected with CAR1 CAR76 andPPARa with p(rAox-PPRE)4-tk-Luc A 2 mM concentra-tion of reagent produced a maximal inhibition of CAR1and CAR76 activity and modestly raised PPARa activityat higher concentrations (Figure 8B) Although inhibitionof PPARa activation by androstanol was not anticipatedthe effect provided an internal control for the experimentsand indicated that androstanol was not simply decreasingluciferase activity through a non-specific mechanism oftranscriptional repression Since the truncated CAR76and CAR80 isoforms behaved like CAR1 in the presenceof androstanol it appears that ligand binding anddimerization properties of CAR are maintained in theleast truncated NCARs CAR1 76 and 80 activities thatwere decreased by androstanol treatment on the p(rAox-PPRE)4-tk-Luc construct were restored by co-treatmentwith the human CAR agonist ligand CITCO (Figure 8C)Titration of CITCO from 03 to 25mM revealed that03mM was sufficient to completely antagonize the effectof 2mM androstanol treatment and lower concentrationsof CITCO in the range of 3ndash30 nanomolar (nM)significantly activated androstanol-repressed CAR1CAR76 and CAR80 transcription on the DR-1 PPRE(data not shown)

No correlation between CARRXRaDNA binding andtransactivation

Nuclear extracts from COS-1 cells enriched for CAR1 orCAR76 expression (by neomycin selection of a CAR-expressing cell population) were used to perform gel shiftassays with a [32P]-labeled CYP2B6-derived DR-4element (NR1) or a [32P]rAoxPPRE (Figure 9A) Theseresults demonstrated that CAR1 and CAR76 bind to bothelements but that the interaction of CAR1 was the mostrobust Since CAR76 was not active on a DR-4 intransient transfection assays and CAR1 and 76 wereequally active on the rAox-PPRE DR-1 in the luciferaseassays no obvious correlation was apparent betweenDNA-binding affinity in the EMSAs and the respectivereporter gene activation levels When CAR 1 or PPARawere expressed with RXRa in COS-1 cells and nuclearextracts from those cells were incubated with [32P]NR1DR-4 or [32P]rAox-PPRE DR-1 CAR1 and PPARainteracted with both DR-4 and DR-1 elements morestrongly than did RXRa alone (Figure 9B) Again even inthe presence of excess RXRa no direct correlation wasmanifested between receptorndashDNA complex formationand transactivation of DR-4- or DR-1- driven luciferasereporter activity

Transcriptional activation in primary human hepatocytes

Although the rAoxPPRE promoter construct used in theCOS-1 transfection studies presented in Figure 4B wasderived from a natural mammalian DR-1 promoterelement and was activated by CAR76 we further testedthe relative biological activities associated with the CAR1

A

B

Figure 7 Transient transfection assays (A) Interplay of RXRahomodimerization or heterodimerization with CAR in transactivationof p(rAox-PPRE)4-tk-Luc in the absence or presence of RXRa ligand9-cis-RA COS-1 cells were transiently transfected with 100 ng ofreporter construct 50 ng pCDNA31() or pCDNA31()CAR 150 ng pCDNA31(thorn) or pCDNA31(thorn)RXRa or RXRa point mutants(Y397A heterodimerization mutant L430F homodimerization mutant)and 10 ng pRL-CMV (for transfection normalization) for 24 h thentreated with dimethyl sulfoxide (DMSO) solvent control or 100 nM9-cis-retinoic acid (RA) for 24 h and assayed for luciferase activity(B) RXRa heterodimerization-dependent nuclear localization ofCAR 1 76 80 in COS-1 cells COS-1 cells were transfected with200 ng pCDNA31() empty vector CAR 1 76 or 80 and 200 ngpCDNA31() RXRa RXRaY397A (heterodimerization mutant) for24ndash48 h Immunofluorescent detection of CAR was performed asdescribed in the Experimental procedures section


and CAR76 receptor isoforms in a more in vivo context Inthese latter assays we infected primary cultures of humanhepatocytes with CAR lentiviral vectors that stablyintegrate in the genome and direct cellular expression oftheir respective receptor proteins The results from twodifferent donors are presented in Figure 10 and recapi-tulate the data derived from the transfected COS-1 cellexperiments indicating that CAR76 expression in hepa-tocytes selectively activated expression of the endogenous

human PEPCK and HMGCS2 genes both of whichcontain DR-1 elements in their respective promoterregions

DISCUSSION

Previous studies of CAR variants demonstrated that notall of the detected protein products were likely attributedto alternatively spliced mRNA species (23) For examples

A B

C

Figure 8 Transient transfection assays (A) 5a-Androstan-3a-ol (androstanol) dosendashresponse to optimize downregulation of CAR 1-dependenttransactivation of p(DR-4)3-tk-Luc COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ng pCDNA31()pCDNA31()CAR 1 or CAR76 50 ng pCDNA31()PPARa and pCDNA31(thorn)RXRa and 10 ng pRL-CMV (for transfection normalization)for 24 h Cells were then treated with dimethyl sulfoxide (DMSO) solvent control or 05 1 2 4 or 8 mM androstanol for 24 h and assayed forluciferase activity (B) 5a-Androstan-3a-ol (androstanol) dosendashresponse to optimize downregulation of CAR 1 and NCAR-dependenttransactivation of p(rAox-PPRE)4-tk-Luc COS-1 cells were transiently transfected with 100 ng of reporter construct 50 ng pCDNA31()pCDNA31()CAR 1 or CAR76 50 ng pCDNA31()PPARa and pCDNA31(thorn)RXRa and 10 ng pRL-CMV (for transfection normalization) for24 h Cells were then treated with DMSO solvent control or 05 1 2 4 or 8mM androstanol for 24 h and assayed for luciferase activity (C) CITCOdosendashresponse to optimize re-activation of androstanol-antagonized CAR 1 and NCAR activity on p(rAox-PPRE)4-tk-Luc COS-1 cells weretransiently transfected with 100 ng of reporter construct 50 ng pCDNA31() pCDNA31()CAR 1 CAR 76 CAR 80 or pCDNA31()PPARaand pCDNA31(thorn)RXRa and 10 ng pRL-CMV (for transfection normalization) for 24 h Cells were then treated with DMSO solvent control 2 mMandrostanol or co-treated with androstanol and increasing CITCO (03 06 125 25 mM) for 24 h and assayed for luciferase activity


when COS-1 cells were transfected with plasmids expres-sing various CAR mRNA splice variants lower molecularweight bands were consistently observed in SDS-PAGEgels Even upon expressing a CAR splice variant posses-sing a partial deletion of exon 7 a proportionally lowermolecular weight band was detected From these observa-tions we hypothesized that isoforms of CAR may arisefrom the use of internal translation start sites In thepresent study we identify new protein translationalvariants of human CAR that differ from those reportedpreviously Specifically we demonstrate that downstreamin-frame start codons produce N-terminally truncatedCAR variants (NCAR) from wild-type as well asalternatively spliced mRNA transcripts We determinedthat CAR1 CAR76 and CAR80 proteins all functionallyactivate luciferase reporter activity when driven by a DR-1-type PPRE promoter element

Specifically since strong Kozak sequences (3637) andtranslational start sites were predicted from the use ofmethionines at positions 1 76 80 125 128 168 and 265we generated plasmid constructs with optimized Kozaksequences to enable overexpression of CAR proteins fromeach putative internal start site and expressed theseconstructs in transiently transfected COS-1 cells in rabbitreticulocyte lysates and in wheat germ in vitro translationsystems In each case protein bands were observed thatmigrated with the expected molecular weights of thetruncated CAR76 and CAR80 proteins and alsoco-migrated with the more rapidly migrating bandsdetectable upon co-expression of the full-length proteinencoded by the CAR1 template A methionine to lysinemutation at codon 1 (CAR M1K) abolished expression offull-length CAR but yielded prominent bands resultingfrom CAR76 and CAR80 initiation Similarly when thecodons for amino acids 76 and 80 were mutated from

methionine to lysine in CAR1 the CAR76 and CAR80bands were abolished with proteins predicted to initiate atamino acids 125 and 128 becoming readily detectableDramatic sequential mutations of upstream methioninesalways led to increases of protein expression from down-stream internal translation start sites supporting a modelof lsquocap-dependentrsquo translation of CAR mRNA whereasinternal ribosome entry site (IRES)-dependent levels ofprotein translation of NCAR would be expected toproceed independently of mutations of upstream initiatorcodons (38)Others have reported that full-length mouse CAR can

activate a DR-2-type PPRE but failed to show anysignificant transactivation on a DR-1 (13) Rather mouseCAR repressed ligand-dependent PPAR aRXRa activityon a DR-1 (13) A more recent study suggested that full-length CAR binds to the promoters of the CYP7A1 andPEPCK genes decreasing their expression by competingwith the transcription factor HNF4 for binding to DR-1elements (39) Perhaps the specific sequence context of aDR-1 element may influence whether CAR exerts apositive or negative effect on transcription activityThe reported discrepancies may also have resulted fromdifferences in the cell lines used or from the inclusionexclusion of RXRa co-transfection It is interestingthat another laboratory reported that alternative splicingof human CAR may result in an mRNA variantthat lacks the usual protein translational start sitein exon 2 instead enabling translation from an AUGstart codon in exon 1 (24) In this latter case the use of analternative exon 1 translation initiation codon predictsthe inclusion of unique N-terminal amino acids thatare in-frame with amino acids encoded by exon 3These investigators also predicted CAR translationfrom codons 125 and 128 rather than codons 76 and 80

A B

Figure 9 Electrophoretic mobility shift assays (A) CAR 1 and CAR 76 binding to [32P]NR1(DR-4) or [32P]rAox-PPRE COS-1 cells were transientlytransfected with pCDNA31() CAR 1 or CAR 76 and subjected to selection with 500 mgml G418 to enrich a population of cells overexpressingCAR 1 and CAR 76 Nuclear extracts were obtained from the CAR-expressing cells and used in EMSA as described in the Experimental proceduressection (B) CAR 1 and PPARa binding to [32P]NR1(DR-4) or [32P]rAox-PPRE COS-1 cells were transiently transfected with pCDNA31() CAR 1or pCDNA31()PPARa for 24 h nuclear extracts were obtained and EMSA performed as described in the Experimental procedures section


that we identify here as the actual start sites used togenerate NCARs possessing unique transcriptionalactivities (24)In our experiments full-length CAR and the NCARs

CAR76 and CAR80 exhibited constitutive activity on aspecial DR-1 PPRE and the receptor activities weredownregulated by treatment with the CAR inverse agonist5a-androstan-3a-ol (androstanol) Co-treatment withandrostanol and the human CAR-specific agonistCITCO completely restored CAR and NCAR activitiesto constitutive levels Transient co-expression of RXRavariants and CAR or NCARs revealed that receptorheterodimerization was the most important criterion forgene promoter activation In this respect the RXRa AF-1domain was not required for transactivation by theNCARs nor was RXRarsquos AF-2 domain When point-mutated RXRa expression constructs were co-transfectedwith the NCARs the RXRa heterodimer-defectivevariant Y397A completely ablated the ability of theNCARs to activate a PPRE response element yet the

Y397A RXRa could still homodimerize and function inreporter transactivation Conversely a mutation of RXRathat abolishes RXRa homodimerization L430F stillallowed for efficient heterodimer formation with CARand the NCARs reflected by subsequent reporter geneactivation but effectively inhibited RXRa homodimertransactivation in the presence of chemical agonist

Therefore our results demonstrate that the lack of aCAR-DNA-binding domain in a CAR-RXRa receptorheterodimer does not inactivate the dimerrsquos transactiva-tion function on a DR-1 or in a PPRE promoter contextIn fact examples of classical nuclear receptors have beenreported previously that retain transactivation ability ongene promoters without direct DNA-binding interactionsincluding ER (40) the retinoic acid receptor (RAR)(41) and the glucocorticoid receptor (42) In the lattercase transgenic mice possessing DNA-binding defectiveGRs were still viable and the modified receptors retainedfunctional ability to cross-talk with other transcrip-tion factors thereby participating in processes such

A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2

Empty CAR1 CAR76Empty CAR1 CAR76Empty CAR1 CAR76Empty CAR1 CAR76


Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B

CAR Expression CYP2B6 Expression PEPCK Expression HMGCS2 Expression

Figure 10 Lentiviral infections and real-time PCR analysis (A) GFP expression in primary human hepatocytes infected with GFP-expressinglentivirus Pseudoviral particles were produced using pSIH1-H1-copGFP lentiviral vector virus was concentrated and infections were conducted asdescribed in the Experimental procedures section Images depict hepatocytes at 5 days post-infection in phase-contrast (100) and fluorescencemicroscopy (100) (B) Real-time PCR analysis of CAR CYP2B6 PEPCK and HMGCS2 expression levels in primary human hepatocytes from twodifferent donors infected with CAR1- and CAR76-expressing lentivirus Pseudoviral particles expressing CAR1 and CAR76 in pCDH1-MCS1-EF1-copGFP lentiviral vectors were produced and infections were performed as described in the Experimental procedures section Four days post-infection hepatocytes were harvested and RNA was isolated converted to cDNA subjected to real-time PCR and the data analyzed as described inthe Experimental procedures section Each bar depicts the average relative fold expression of two PCR reactions from a single hepatocyte donor


as transrepression For CAR and perhaps for othernuclear receptors as well we predict that truncated formsof the receptor may exhibit full or partial activationfunction or even repression and that the spectrum ofthese activities may manifest on an altered repertoire ofresponse elements Specifically we propose a new modelof CAR DBD-independent transactivation whereby onlythe RXRa component of the heterodimer binds directly toa DR-1-containing PPRE tethered to CAR or NCARs(CAR76 or CAR80) with functional contributionsfrom the CAR AF-2 domain providing the basis forco-activator recruitment and subsequent gene transactiva-tion function This type of heterodimerndashDNA interactionis in striking contrast to that reported previously forthe APYLT-CAR splice variant a variant that possesses a5 aa insertion in the receptorrsquos dimerization interface (14)For the latter receptor the RXRa DBD rather thanthe APYLT-CAR DBD is expendable for transcriptionalactivation implying a model invoking direct contact ofonly the CAR portion of the dimer with DNA notthe RXRa component (14) Taken together along withthe recently identified phosphorylation modificationsthat impact the ability of CAR to translocate (43) it isbecoming clear that the post-transcriptional regulationof CAR expression is an important means of generating adiversity of receptor modalities that likely function toenhance its complex role as a receptor integrator ofxenobiotic and endobiotic sensing in mammalian cells

ACKNOWLEDGEMENTS

The authors wish to thank Ms Denise Weyant for DNAsequencing Mr Michael Packard for assistance withfluorescence microscopy Mr Shengzhong Su for assis-tance with plasmid generation Ms Mary Johnson forassistance with human primary hepatocyte cell culture andadenovirus infections and Dr Jerry Thompson for helpfuldiscussions Funding to pay the Open Access publicationcharge was provided by a grant from The NationalInstitute of General Medical Sciences USPHS GM066411(CJO)

Conflict of interest statement None declared

REFERENCES

1 ChawlaA RepaJJ EvansRM and MangelsdorfDJ (2001)Nuclear receptors and lipid physiology opening the X-files Science294 1866ndash1870

2 BaesM GulickT ChoiHS MartinoliMG SimhaD andMooreDD (1994) A new orphan member of the nuclear hormonereceptor superfamily that interacts with a subset of retinoic acidresponse elements Mol Cell Biol 14 1544ndash1552

3 KawamotoT SueyoshiT ZelkoI MooreR WashburnK andNegishiM (1999) Phenobarbital-responsive nuclear translocation ofthe receptor CAR in induction of the CYP2B gene Mol Cell Biol19 6318ndash6322

4 YoshinariK KobayashiK MooreR KawamotoT andNegishiM (2003) Identification of the nuclear receptorCARHSP90 complex in mouse liver and recruitment of proteinphosphatase 2A in response to phenobarbital FEBS Lett 54817ndash20

5 KobayashiK SueyoshiT InoueK MooreR and NegishiM(2003) Cytoplasmic accumulation of the nuclear receptor CAR

by a tetratricopeptide repeat protein in HepG2 cells MolPharmacol 64 1069ndash1075

6 MaglichJM ParksDJ MooreLB CollinsJL GoodwinBBillinAN StoltzCA KliewerSA LambertMH et al (2003)Identification of a novel human constitutive androstane receptor(CAR) agonist and its use in the identification of CAR target genesJ Biol Chem 278 17277ndash17283

7 SidhuJS and OmiecinskiCJ (1997) An okadaic acid-sensitivepathway involved in the phenobarbital-mediated induction ofCYP2B gene expression in primary rat hepatocyte culturesJ Pharmacol Exp Ther 282 1122ndash1129

8 FormanBM TzameliI ChoiHS ChenL SimhaD SeolWEvansRM and MooreDD (1998) Androstane metabolites bindto and deactivate the nuclear receptor CAR-beta Nature 395612ndash615

9 SueyoshiT KawamotoT ZelkoI HonkakoskiP andNegishiM (1999) The repressed nuclear receptor CAR respondsto phenobarbital in activating the human CYP2B6 geneJ Biol Chem 274 6043ndash6046

10 HonkakoskiP JaaskelainenI KortelahtiM and UrttiA (2001)A novel drug-regulated gene expression system based on the nuclearreceptor constitutive androstane receptor (CAR) Pharm Res 18146ndash150

11 TzameliI PissiosP SchuetzEG and MooreDD (2000) Thexenobiotic compound 14-bis[2-(35-dichloropyridyloxy)]benzene isan agonist ligand for the nuclear receptor CAR Mol Cell Biol 202951ndash2958

12 XuC LiCY and KongAN (2005) Induction of phase I II andIII drug metabolismtransport by xenobiotics Arch Pharm Res28 249ndash268

13 KassamA WinrowCJ Fernandez-RachubinskiF CaponeJPand RachubinskiRA (2000) The peroxisome proliferator responseelement of the gene encoding the peroxisomal beta-oxidationenzyme enoyl-CoA hydratase3-hydroxyacyl-CoA dehydrogenase isa target for constitutive androstane receptor beta9-cis-retinoic acidreceptor-mediated transactivation J Biol Chem 275 4345ndash4350

14 AuerbachSS StonerMA SuS and OmiecinskiCJ (2005)Retinoid X receptor-a-dependent transactivation by a naturallyoccurring structural variant of human constitutive androstanereceptor (NR1I3) Mol Pharmacol 68 1239ndash1253

15 FrankC GonzalezMM OinonenC DunlopTW andCarlbergC (2003) Characterization of DNA complexes formed bythe nuclear receptor constitutive androstane receptor J BiolChem 278 43299ndash43310

16 BurkO ArnoldKA GeickA TegudeH and EichelbaumM(2005) A role for constitutive androstane receptor in the regulationof human intestinal MDR1 expression Biol Chem 386 503ndash513

17 TironaRG LeeW LeakeBF LanLB ClineCB LambaVParvizF DuncanSA InoueY et al (2003) The orphan nuclearreceptor HNF4 alpha determines PXR- and CAR-mediatedxenobiotic induction of CYP3A4 Nat Med 9 220ndash224

18 ChenY KisslingG NegishiM and GoldsteinJA (2005) Thenuclear receptors constitutive androstane receptor and pregnane Xreceptor cross-talk with hepatic nuclear factor 4alpha to synergis-tically activate the human CYP2C9 promoter J Pharmacol ExpTher 314 1125ndash1133

19 MinGS KemperJK and KemperB (2002) Glucocorticoidreceptor-interacting protein 1 mediates ligand-independent nucleartranslocation and activation of constitutive androstane receptor invivo J Biol Chem 277 26356ndash26363

20 MuangmoonchaiR SmirlisD WongSC EdwardsMPhillipsIR and ShephardEA (2001) Xenobiotic induction ofcytochrome P450 2B1 (CYP2B1) is mediated by the orphan nuclearreceptor constitutive androstane receptor (CAR) and requiressteroid co-activator 1 (SRC-1) and the transcription factor Sp1Biochem J 355 71ndash78

21 ShirakiT SakaiN KanayaE and JingamiH (2003) Activationof orphan nuclear constitutive androstane receptor requires sub-nuclear targeting by peroxisome proliferator-activated receptorgamma coactivator-1 alpha ndash Possible link between xenobioticresponse and nutritional state J Biol Chem 278 11344ndash11350

22 ArnoldKA EichelbaumM and BurkO (2004) Alternativesplicing affects the function and tissue-specific expression of thehuman constitutive androstane receptor Nucl Recept 2 1


23 AuerbachSS RamsdenR StonerMA VerlindeC HassettCand OmiecinskiCJ (2003) Alternatively spliced isoforms of thehuman constitutive androstane receptor Nucleic Acids Res 313194ndash3207

24 LambaJK LambaV YasudaK LinYS AssemMThompsonE StromS and SchuetzE (2004) Expression ofconstitutive androstane receptor splice variants in human tissuesand their functional consequences J Pharmacol Exp Ther 311811ndash821

25 SidhuJS LiuF and OmiecinskiCJ (2004) Phenobarbitalresponsiveness as a uniquely sensitive indicator of hepatocytedifferentiation status requirement of dexamethasone and extra-cellular matrix in establishing the functional integrity of culturedprimary rat hepatocytes Exp Cell Res 292 252ndash264

26 VaranasiU ChuR HuangQ CastellonR YeldandiAV andReddyJK (1996) Identification of a peroxisome proliferator-responsive element upstream of the human peroxisomal fatty acylcoenzyme A oxidase gene J Biol Chem 271 2147ndash2155

27 NakshatriH and ChambonP (1994) The directly repeatedRG(GT)TCA motifs of the rat and mouse cellular retinol-bindingprotein II genes are promiscuous binding sites for RAR RXRHNF-4 and ARP-1 homo- and heterodimers J Biol Chem 269890ndash902

28 RenB ThelenA and JumpDB (1996) Peroxisome proliferator-activated receptor alpha inhibits hepatic S14 gene transcriptionEvidence against the peroxisome proliferator-activated receptoralpha as the mediator of polyunsaturated fatty acid regulation ofs14 gene transcription J Biol Chem 271 17167ndash17173

29 StonerM SavilleB WormkeM DeanD BurghardtR andSafeS (2002) Hypoxia induces proteasome-dependent degradationof estrogen receptor alpha in ZR-75 breast cancer cellsMol Endocrinol 16 2231ndash2242

30 LivakKJ and SchmittgenTD (2001) Analysis of relative geneexpression data using real-time quantitative PCR and the2(-Delta Delta C(T)) method Methods 25 402ndash408

31 ChennaR SugawaraH KoikeT LopezR GibsonTJHigginsDG and ThompsonJD (2003) Multiple sequencealignment with the Clustal series of programs Nucleic Acids Res31 3497ndash3500

32 PedersenAG and NielsenH (1997) Neural network prediction oftranslation initiation sites in eukaryotes perspectives for EST andgenome analysis Proc Int Conf Intell Syst Mol Biol 5 226ndash233

33 XuRX LambertMH WiselyBB WarrenEN WeinertEEWaittGM WilliamsJD CollinsJL MooreLB et al (2004)A structural basis for constitutive activity in the human CARRXRalpha heterodimer Mol Cell 16 919ndash928

34 Vivat-HannahV BourguetW GottardisM and GronemeyerH(2003) Separation of retinoid X receptor homo- and heterodimer-ization functions Mol Cell Biol 23 7678ndash7688

35 WanH DawsonMI HongWK and LotanR (1998)Overexpressed activated retinoid X receptors can mediate growthinhibitory effects of retinoids in human carcinoma cells J BiolChem 273 26915ndash26922

36 KozakM (1987) At least six nucleotides preceding the AUGinitiator codon enhance translation in mammalian cells J MolBiol 196 947ndash950

37 KozakM (1987) An analysis of 50-noncoding sequences from 699vertebrate messenger RNAs Nucleic Acids Res 15 8125ndash8148

38 SvitkinYV HerdyB Costa-MattioliM GingrasACRaughtB and SonenbergN (2005) Eukaryotic translationinitiation factor 4E availability controls the switch betweencap-dependent and internal ribosomal entry site-mediatedtranslation Mol Cell Biol 25 10556ndash10565

39 MiaoJ FangS BaeY and KemperJK (2006) Functionalinhibitory cross-talk between constitutive androstane receptor andhepatic nuclear factor-4 in hepatic lipidglucose metabolism ismediated by competition for binding to the DR1 motif and to thecommon coactivators GRIP-1 and PGC-1alpha J Biol Chem281 14537ndash14546

40 PorterW SavilleB HoivikD and SafeS (1997) Functionalsynergy between the transcription factor Sp1 and the estrogenreceptor Mol Endocrinol 11 1569ndash1580

41 ShimadaJ SuzukiY KimSJ WangPC MatsumuraM andKojimaS (2001) Transactivation via RARRXR-Sp1 interactioncharacterization of binding between Sp1 and GC box motifMol Endocrinol 15 1677ndash1692

42 ReichardtHM KaestnerKH TuckermannJ KretzOWesselyO BockR GassP SchmidW HerrlichP et al (1998)DNA binding of the glucocorticoid receptor is not essential forsurvival Cell 93 531ndash541

43 HosseinpourF MooreR NegishiM and SueyoshiT (2006)Serine-202 regulates the nuclear translocation of constitutiveactiveandrostane receptor CAR Mol Pharmacol 691095ndash1102


nuclear co-regulator proteins including the co-activatorsGRIP-1 (19) SRC-1 (20) and PGC-1 (21)In addition to the reference form of human CAR

termed CAR1 we and others have characterized alter-natively spliced CAR transcripts that exist in humantissues (22ndash24) At least one of these CAR variantsCAR3 possesses the unique biological property of beinga ligand-activated receptor in contrast to the constitu-tively active nature of CAR1 (14) The results presentedin the current investigation extend the complexity ofCAR regulation by demonstrating that the receptorrsquosexpression is post-transcriptionally modulated throughthe use of internal translation start sites resulting in thegeneration of truncated CAR proteins with alteredbiological function Our data also suggest a new modelfor CARndashDNA interaction whereby RXRa binds directlyto a DR-1 peroxisome-proliferator-activated responseelement (PPRE) as a heterodimer with tethered CARsuch that only RXRa makes direct DNA contact withthe tethered CAR AF-2 allowing for transcriptionalactivation

EXPERIMENTAL PROCEDURES

Cell lines chemicals and biochemicals

COS-1 African green monkey kidney epithelial cell line(ATCC Manassas VA) was cultured in DulbeccorsquosModified Eagle Medium (Invitrogen Gaithersburg MD)plus 10 fetal bovine serum 15 gl sodium bicarbonateHEPES L-glutamine and penicillinndashstreptomycin5a-Androstan-3a-ol (androstanol) and 9-cis-retinoic acid(RA) (Sigma St Louis MO) CITCO (BioMol PlymouthMeeting PA) and Wyeth 14643 were dissolved indimethylsulfoxide (DMSO) (Sigma St Louis MO)Enriched primary human hepatocyte cultures plated oncollagen were obtained through the Liver TissueProcurement and Distribution System Pittsburghfunded by NIH Contract NO1-DK-9-2310 Cells wereplaced in fresh Williamrsquos E media containing 1penicillinndashstreptomycin 1 HEPES 20 mM glutamine25 nM dexamethasone 10 nM insulin 1 linoleic acidBSA 5 ngml selenious acid and 5 mgml transferrin andoverlayed mL BD MatrigelTMmL BD MatrigelTM

Basement Membrane Matrix (BD Biosciences San JoseCA) as described previously (25) Adenovirus expressinghuman CAR was produced using a commercially availablesystem AdEasy (Stratagene Cedar Creek TX) Primaryhepatocytes from one individual were infected withAdenovirus-CAR (Adv-CAR) for 24 h cells were scrapedin Laemmlirsquos loading buffer and a small aliquot wasresolved by SDS-PAGE transferred to PVDF membraneand CAR detected by immunoblotting Lentiviral expres-sion plasmids pCDH1-MCS1-EF1-copGFP and pSIH1-H1-copGFP and packaging plasmid mix pPACKH1were purchased from System Biosciences (Mountain ViewCA) Production of lentiviral particles and subsequentinfection of target cells were performed essentiallyaccording to the manufacturerrsquos protocols except thatLipofectamine 2000 (Invitrogen Carlsbad CA) wasused for transfection of HEK-293T packaging cells

Pseudoviral supernatant harvested from the packagingcells was either used directly for target cell infection or wasconcentrated prior to infection by 10 PEG-8000 (SigmaSt Louis MO) precipitation Primary human hepatocyteswere infected with equal volumes of appropriate lentiviralsupernatant in the presence of 6 mgml polybrene (SigmaSt Louis MO) and media was replaced the followingday

Oligonucleotides and plasmids

Oligonucleotides and PCR primers were synthesized byIntegrated DNA Technologies (IDT) (Coralville IA)PCR primers and EMSA oligonucleotides used are listedin Supplementary Data A consensus DR-4 probe used inEMSA competition experiments contained the NR1sequence from CYP2B6 PBREM and the sense strandcontained the classical DR-4 element AGTTCA(N)4AGTTCA Reporter plasmids contained four copies ofthe rat acyl-CoA-oxidase (rAox) PPRE (26) a PCRfragment from rat cellular retinol-binding protein IIpromoter (rCRBPII) (27) and three copies of an optimizedAGTTCA DR-4(14) Complementary oligonucleotideswere annealed phosphorylated and inserted in the SmaIsite of ptk-Luc derived from pGL3-Basic vector(Promega Madison WI) which was modified by insertionof the minimal thymidine kinase (tk) element frompBLCAT prAox-1198-463-tk-Luc contains a singlecopy of the PPRE (DR-1 tgacctttgtcct) in its endogenouscontext encompassed by the region between 1198 and463 bp upstream of the rAox gene transcription start siteand was amplified using rat genomic DNA as a templatePrimer sequences employed were as previously described(28) but with slight modifications the sense primer(50-GGT ACC GGT ACC CCA GTA GAA CCT TGTTCA GG-30) and antisense primer (50-GGT ACC GCTAGC CAG GGT CTC GGG CGG AGT GAA G-30)contained the underlined KpnI and NheI restrictionenzyme sites respectively The 754-bp amplicon was gel-purified restriction enzyme digested gel-purified andligated into the ptk-Luc vector (Promega Madison WI)Plasmids were prepared using QIAfilter Plasmid Maxi Kit(Qiagen Valencia CA) and sequenced

Coupled in vitro transcription translation

Rabbit reticulocyte lysate and wheat germ expressionsystems were used essentially according to the manufac-turerrsquos (Promega Madison WI) protocols DNAtemplates were in the form of plasmids containingCAR-coding sequence downstream of T7 promoteror PCR products of T7 promoter-tagged CAR fragmentsIn some experiments [35S]methionine (ICN) was incorpo-rated while in other experiments a polyclonal antibodyagainst human CAR (23) was used to perform westernblotting

Western-immunoblotting analysis

Cells were harvested directly in 1 Laemmlirsquos loading dyeand boiled for 5min Alternately cells were scraped inradioimmunoprecipitation assay (RIPA) buffer proteinconcentrations were measured by Bradford method













RESULTS







A

B


A

B







A

B















A

B







A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES


























































RESULTS







A

B


A

B







A

B















A

B







A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES


















































A

B


A

B







A

B















A

B







A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES



















































A

B















A

B







A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES



























































A

B







A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES



















































A B

C D











A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES























































A

B





DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES

















































DISCUSSION


A B

C







A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES



















































A B







A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES



















































A

B

0

5

10

15

20

0

1

2

3

4

5

0

05

1

15

2



Rel

ativ

e F

old

Exp

ress

ion

0

05

1

15

2

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e F

old

Exp

ress

ion

Rel

ativ

e fo

ld e

xpre

ssio

n

CAR Expression

0

50

100

150

200

CYP2B6 Expression

0

5

10

15

20

25

30

PEPCK Expression

0

05

1

15

2

25

0

05

1

15

2

25

HMGCS2 Expression

HepatocyteDonor A

HepatocyteDonor B





ACKNOWLEDGEMENTS



REFERENCES
















































ACKNOWLEDGEMENTS



REFERENCES















































nucleic acids research, 2007, doi:10.1093/nar/gkm090

Documents