The aryl hydrocarbon receptor (AhR) tyrosine 9, a residue that is essential for AhR DNA binding activity, is not a phosphoresidue but augments AhR phosphorylation.

We delineate a mechanism by which dioxin (2,3,7,8-tetrachlorodibenzo-p-dioxin or TCDD)-mediated formation of the aryl hydrocarbon receptor (AhR) DNA binding complex is disrupted by a single mutation at the conserved AhR tyrosine 9. Replacement of tyrosine 9 with the structurally conservative phenylalanine (AhRY9F) abolished binding to dioxin response element (DRE) D, E, and A and abrogated DRE-driven gene induction mediated by the AhR with no effect on TCDD binding, TCDD-induced nuclear localization, or ARNT heterodimerization. The speculated role for phosphorylation at tyrosine 9 was also examined. Anti-phosphotyrosine immunoblotting could not detect a major difference between the AhRY9F mutant and wild-type AhR, but a basic isoelectric point shift was detected by two-dimensional gel electrophoresis of AhRY9F. However, an antibody raised to recognize only phosphorylated tyrosine 9 (anti-AhRpY9) confirmed that AhR tyrosine 9 is not a phosphorylated residue required for DRE binding. Kinase assays using synthetic peptides corresponding to the wild-type and mutant AhR residues 1-23 demonstrated that a tyrosine at position 9 is important for substrate recognition at serine(s)/threonine(s) within this sequence by purified protein kinase C (PKC). Also, compared with AhRY9F, immunopurified full-length wild-type receptor was more rapidly phosphorylated by PKC. Furthermore, co-treatment of AhR-deficient cells that expressed AhRY9F and a DRE-driven luciferase construct with phorbol 12-myristate 13-acetate and TCDD resulted in a 30% increase in luciferase activity compared with AhRY9F treated with TCDD alone. Overall, AhR tyrosine 9, which is not a phosphorylated residue itself but is required for DNA binding, appears to play a crucial role in AhR activity by permitting proper phosphorylation of the AhR.

PAS) transcription factor family, which includes Per, AhR nuclear translocator (ARNT or hypoxia-inducible factor 1-␤), Sim, and HIF 1-␣ (1)(2)(3)(4). This protein is believed to mediate the biological and toxic effects of a class of environmental pollutants, best exemplified by 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD or dioxin) (5). TCDD binding initiates the process of transformation, which includes dissociation of one or several of the chaperone-related proteins from the unliganded AhR in the cytoplasm, translocation of the receptor-ligand complex into the nucleus, and dimerization with ARNT, to form the active TCDD-AhR-ARNT transcription factor complex (6,7). The endogenous ligand for the AhR remains unknown.
The active AhR-ARNT complex specifically recognizes a consensus sequence, termed a dioxin response element (DRE, 5Ј-CACGCNA-3Ј), located in the upstream regulatory region of AhR-responsive genes, such as Cyp1A1, leading to transcription initiation. The basic region of the AhR contains multiple essential structural components for direct contact to DNA. It has two basic clusters separated by about 20 amino acids, both of which appear to be engaged in direct protein-DNA contact. Mutation of certain positively charged residues in either cluster (e.g. arginine 14 in the more N-terminal basic region or arginine 39 in the nominal basic region) to alanine or lysine has been found to abolish the formation of the ternary complex between the AhR-ARNT dimer and the DRE, suggesting that it is these basic residues that directly contact the DRE (8 -11).
Previous investigations had determined that when N-terminal deletion analyses were performed to map the DNA binding domain of an AhR mutant lacking the C-terminal half and the minimal ligand binding domain (AhRC⌬516), the first 9 amino acids, including tyrosine 9, were dispensable for constitutive DNA binding activity (11). In contrast, N-terminal deletion analyses of the full-length mouse AhR indicated that the first 9 amino acids are required for TCDD-induced DNA binding of the AhR but not required for heterodimerization with ARNT (10). Also, amino acid scanning mutations to alanine, serine, or tryptophan showed that tyrosine 9 is solely responsible for the loss of DNA binding and transcriptional activity of these Nterminal deletion mutants (9,10). Besides arginines 14 and 39, AhR tyrosine 9 was the only residue in the N terminus whose mutation resulted in dramatic loss of both DNA binding and transcriptional activity (9,10). These data clearly demonstrate that AhR tyrosine 9 is a critical residue required for full-length AhR activity. However, the mechanism by which AhR tyrosine 9 plays a role in AhR activity remains controversial. While this residue is not expected to directly contact DNA, several mechanisms including phosphorylation of this residue have been speculated (9,12,13). Here, we examine three known or speculated aspects of AhR activation in which tyrosine 9 may potentially play a role. We analyze the impact of a tyrosine 9 mutation on various known steps of the AhR mechanism of action, including ligand binding, nuclear localization, and ARNT heterodimerization, as well as the ability to bind to various DRE sequences. We also test the hypothesis that tyrosine 9 is required for ligand-dependent DNA binding (i.e. for full-length AhR) (11,14) and the hypothesis that tyrosine 9 is required for N-terminal cleavage of the AhR prior to DNA binding (15). Finally, we address the hypothesis that phosphorylation of, or mediated by, tyrosine 9 is critical for normal AhR activation (10,13).
While AhR tyrosine 9 does not appear to be essential for ligand-elicited nuclear localization or ARNT dimerization, it is necessary for DNA binding of the full-length protein. Although phosphorylation of the AhR tyrosine 9 is not required for DNA binding, mutating tyrosine 9 alters the isoelectric points of the AhR charged forms suggesting that tyrosine 9 plays a role in post-translational modification of other AhR residue(s). We demonstrate that protein kinase C (PKC) phosphorylates the wild-type (WT) AhR, and that mutating AhR tyrosine 9 decreases PKC-elicited phosphorylation of the AhR. Furthermore, we demonstrate that the decreased transcriptional activity of the AhRY9F mutant can be partially overcome upon co-treatment with TCDD and phorbol 12-myristate 13-acetate (PMA). These data demonstrate for the first time that, while tyrosine 9 itself is not phosphorylated, it can play a crucial role in phosphorylation of the AhR and AhR-mediated gene transcription.

EXPERIMENTAL PROCEDURES
Generation of AhR Constructs-Site-directed mutagenesis and construction of pcDNA3/␤AHR-HIS and pcDNA3/␤AHRY9F-HIS were performed as described previously (9,10,12,13). For generation of the pHM6/AhRs that code for N-terminally HA-tagged and C-terminally 6ϫ histidine-tagged AhRs, the wild-type and mutant AhR coding sequences were amplified from pcDNA3/␤AHR or pcDNA3/␤AHR mutant DNA by Pfu Turbo DNA polymerase using the forward primer, 5Ј-CGA CCC AAG CTT GCT AGC AAT GTC TAG CGG CGC CAA C-3Ј and the reverse primer, 5Ј-GGT TAC CCG CGG CTC GAG GGA TCC CAC TCT GCA CCT TGC TTA G-3Ј. HindIII and KspI were used to digest the PCR product and pHM6 to allow subcloning of the amplified AhR coding sequences into pHM6 (Roche Applied Science). For generation of the AhR346 truncation mutants, the wild-type and mutant AhR coding sequence was amplified from pcDNA3/␤AHR and pcDNA3/␤AHRY9F by Pfu Turbo DNA polymerase using T7 forward primer and the reverse primer, 5Ј-CCG CTC GAG CGG TCA GTG ATG GTG ATG GTG ATG CCG GAA AAC TGT CAT GCC-3Ј, containing coding sequences for 6 consecutive histidines, a stop codon, and a XhoI site. The PCR products were digested with HindIII and XhoI and subcloned into the similarly restricted pcDNA3/Neo to generate pcDNA3/␤AHR346-HIS and pcDNA3/␤AHR346Y9F-HIS.
Cell Culture-Wild-type (Hepa1c1c7) and AhR-deficient mouse hepatoma cells (BP r c1TAO) were kind gifts from Dr. J. Whitlock, Jr. (Stanford University). These cells were incubated at 37°C in a humidified incubator and grown in ␣-MEM supplemented with 10% fetal bovine serum. COS-7 cells were obtained from ATCC and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and were incubated at 37°C in a humidified incubator.
Subcellular Localization of AhR-GFP Fusion Proteins-The GFP coding sequence was amplified from pEGFP-C1 (Clontech, Palo Alto, CA) by Pfu Turbo DNA polymerase using the forward primer, 5Ј-ATA AGA ATG CGG CCG CTA AAA TGG TGA GCA AGG GC-3Ј containing the NotI site and the reverse primer, 5Ј-GGT ATG GCT GAT TAT GAT CAG-3Ј. The PCR products were digested with XhoI and NotI and subcloned into the similarly restricted pcDNA3/␤AHRs to generate pcDNA3/␤AHR495-GFP. These vectors were transiently transfected into Hepa1c1c7 cells on coverslips. After treatment with either Me 2 SO-or TCDD-containing medium for an hour, the cells were fixed with 4% paraformaldehyde for 30 min at room temperature, mounted on slides, and examined under a fluorescence microscope.
In Vitro Translation of Wild-type and Mutant AhRs and ARNT, Electrophoretic Mobility Shift Assay (EMSA), and Supershift EMSA-EMSAs were performed essentially as described previously (13,16). Equal amounts of in vitro translated AhRs and ARNT were incubated with either 0.1% Me 2 SO or 10 nM TCDD (Cambridge Isotopes, Cambridge, MA) for 90 min at room temperature and then mixed with 50,000 cpm of 32 P-labeled DRE at 0.085 M NaCl and 10 mM DTT and analyzed in a 4% nondenaturing gel. The DNA binding form of the AhR-ARNT complex was visualized by PhosphorImager (Molecular Dynamics, Sunnyvale, CA). For DRE synthesis, one strand of each DRE (17) was labeled at the 5Ј-end using T4 polynucleotide kinase prior to annealing with the unlabelled complementary strand; DRE D (AhRE3), 5Ј-GAT CCG GCT CTT CT CACGCAACTC CGA GCT CA-3Ј; DRE E (AhRE2), 5Ј-CCC AGT GCT GT CACGCTAGCT GGG GGA GGG GAA-3Ј; DRE A (AhRE5), 5Ј-TGC GCT TCT CACGCGAGCT TGG-3Ј. The duplex oligonucleotide contains a single AhR binding sequence (underlined). Supershift assays were performed as described for EMSAs with the exception of the use of the HA-tagged AhR expression vectors and the inclusion of antibody at the incubation step just prior to DRE addition. The antibodies that recognize the hemagglutinin epitope (HA.11, Covance, Berkeley, CA), the AhR (Rpt-9), a nonspecific rabbit anti-mouse IgG (HϩL) (Jackson Immunoresearch Laboratories), or the anti-AhR phosphotyrosine 9 antibody were used for supershift assays.
In Vitro Protein-Protein Interaction Study-Interaction of AhR with ARNT was investigated by using radiolabeled ARNT expressed in the TNT® (Promega) system as described above with the exception of the inclusion of 15 Ci of [ 35 S]methionine in the reaction mixture. 35 Slabeled ARNT was incubated with the similarly expressed but unlabeled histidine-tagged AhRs in the presence or absence of 10 nM TCDD (in 0.1% Me 2 SO) for 90 min at room temperature. AhR-bound ARNT was precipitated with 70 l of Ni-NTA-agarose for 30 min at 4°C in the His tag lysis buffer with 0.3 M NaCl. The pellet was washed twice with His tag lysis buffer containing 40 mM imidazole and boiled with SDS-PAGE buffer. The co-precipitated ARNT was analyzed by SDS-PAGE and visualized by autoradiography.
Co-transfection Assay of AhR-deficient Cells-To determine AhRWT and AhRY9F transactivation activity, 2 ϫ 10 5 AhR-deficient cells (BP r c1TAO or COS-7) were co-transfected with 0.45 g of AhR expression vectors (pcDNA3/␤AHRs), 0.45 g of DRE-reporter gene (p2DLuc), and 0.1 g of normalization vector (pRSVLacZ) using LipofectAMINE (Invitrogen) or GenePORTER (Gene Therapy Systems, San Diego, CA) in 6-well plates as described previously (12). After the recovery from transfection, the cells were treated with either 0.1% Me 2 SO or 4 nM TCDD (in 0.1% Me 2 SO) for 22 h and harvested for the reporter gene assay. DRE-driven luciferase activity was measured with the luciferase assay kit (Promega), and ␤-galactosidase activity was determined with Galacto-Light Plus (Tropix, Bedford, MA) according to the manufacturer using a Turner Model TD-20e Luminometer (Turner Designs, Sunnyvale, CA).
Immunoblot Analysis of Partially Purified Full-length AhRs with Anti-phosphotyrosine Antibody-The pcDNA3/␤AHR-HIS and pcDNA3/␤AHRY9F-HIS vectors were transiently transfected into the AhR-deficient TAO cells using GenePORTER according to the manufacturer and partially purified exactly as previously described using Ni-NTA-agarose (12). After separation by SDS-PAGE, the proteins were transferred to a PVDF membrane, and the membrane was probed with PY-20 monoclonal anti-phosphotyrosine antibodies (Transduction Laboratory, Lexington, KY) in 3% BSA-TBST. The membrane was stripped and reprobed with anti-AhR antibody (Rpt.1; ascites prepared using hybridoma cells that were a kind gift from Dr. G. Perdew (Pennsylvania State University)). The primary antibody was located with HRP-conjugated goat anti-mouse IgG and visualized with LumiGLO chemiluminescent substrate (KPL, Gaithersburg, MD).
Two-dimensional Gel Electrophoresis of Wild-type AhR and AhRY9F-Two-dimensional gel electrophoresis was performed utilizing radiolabeled AhRs separately expressed in the TNT® system as described above with the exception of inclusion of 15 Ci of [ 35 S]methionine in the reaction mixture, methods similar to those previously described (13). 50 l of the TNT® lysate containing 35 S-labeled AhR was diluted with 150 l of HEDG buffer (25 mM Hepes, 1.5 mM Na 2 EDTA, 1 mM DTT, 10% (v/v) glycerol; adjusted to pH 7.6) containing protease inhibitors. 50 l of this mixture was added to 30 l of HEDG containing 1.5 g/l carrier protein and added to 215 l of rehydration buffer (6.4 M urea, 2.6 M thiourea, 4% CHAPS, 39 mM DTT, 0.2% Bio-Lytes 3/10 ampholytes, 0.001% bromphenol blue). The Bio-Rad Protean IEF Cell and reagents were used for isoelectric focusing (IEF) and the Criterion Precast Gel System was used for SDS-PAGE (Bio-Rad). ReadyStrip IPG 11-cm strips (pH range 3-10) were actively rehydrated (application of 50 V) at 16°C covered with mineral oil for 26 h. For IEF, a slow ramping protocol was followed: a conditioning step of 250 V for 15 min, followed by voltage ramping up to 8000 V over 2.5 h, and final focusing at 8000 V for 80,000 VH. Maximum current limit was set at 50 A per gel, with total VH not exceeding 100,000 VH. Following IEF, the IPG strips were equilibrated for 30 min in DTT equilibration buffer (5 M urea, 0.8 M thiourea, 4% SDS, 0.375 M Tris-HCl pH 8.8, 20% glycerol, and 130 mM DTT), followed by 30 min in iodoacetamide equilibration buffer (5 M urea, 0.8 M thiourea, 4% SDS, 0.375 M Tris-HCl pH 8.8, 20% glycerol, and 135 mM iodoacetamide). IPG strips were loaded to the well of a 7.5% Tris-HCl precast gel and run at a constant voltage of 200 V for 1 h. Proteins from the 2nd dimension gels were transferred to Sequi-Blot PVDF membranes (Bio-Rad) and 35 S-labeled AhRs were visualized with a PhosphorImager (Amersham Biosciences).
Anti-AhR Phosphotyrosine 9 Production and Immunodetection of Phosphorylated AhR Tyrosine 9 -Synthesis of peptides, BSA conjugation, and generation of the phosphopeptide antibody was carried out by Alpha Diagnostic International (San Antonio, TX) as previously described (18 -20). A BLAST search was performed to optimize selection of sequence that would be most likely antigenic and exhibit minimal cross-reactivity with nonspecific proteins. The hydrophilic region, residues 4 -16 (GANIT(pY)ASRKRRK) with a cysteine on the C terminus for keyhole limpet hemocyanin conjugation, was used as the phosphotyrosine 9 peptide antigen purified to Ͼ90% for presentation to rabbits. For affinity purification, a non-phosphopeptide control of the same sequence without a phosphorylated tyrosine 9 was used as an additional affinity support to remove nonspecific antibodies prior to use of the phosphotyrosine 9 peptide affinity column for capture of the antibodies of interest. Both the control and phosphopeptide were used to test antibody cross-reactivity with non-phosphorylated peptides by ELISA. BSA was conjugated to the phosphopeptide (BSA-pY9) using the glutaraldehyde conjugation method to allow optimization for immunodetection of phosphorylated tyrosine 9. To determine the specificity of the anti-AhRpY9 antibody, the BSA-pY9 conjugate was incubated with a mixture of alkaline phosphatase and potato acid phosphatase to remove the phosphate group from tyrosine 9 to serve as a negative control. For detection of AhR from transfected cells, AhR-deficient cells were transfected with 8 g of an AhR expression vector (pHM6/AHR) using GenePORTER in 100-mm plates 24 h after plating 2 ϫ 10 6 cells in each plate. AhR-deficient cells from confluent plates were not transfected as a negative control. After 48 h of recovery, cells were treated for 1 h with either 0.1% Me 2 SO or 10 nM TCDD. Cells were washed with Hank's Buffer, and total protein was collected by triturating the cells in 150 l of high salt HEG buffer (25 mM Hepes, 1.5 mM Na 2 EDTA, 10% (v/v) glycerol, 0.45 M NaCl, adjusted to pH 7.6) containing protease inhibitors (Complete Mini Tablets, Roche Applied Science). The AhR was immunoprecipitated from the lysate preparation with an anti-HA antibody. The pelleted protein was loaded on a 7.5% polyacrylamide gel and separated by SDS-PAGE. Protein was transferred to a PVDF membrane and analyzed by immunoblot analysis. The primary antibodies were located with HRP-conjugated goat anti-rabbit IgG and visualized with LumiGLO chemiluminescent substrate. To determine if the anti-AhRpY9 antibody was capable of immunoprecipitating the AhR, wildtype and AhRY9F HA-tagged receptors were translated in a total volume of 50 l containing TNT®-coupled rabbit reticulocyte lysate in the presence of T7 RNA polymerase for 90 min at 30°C as described above with the exception of inclusion of 15 Ci of [ 35 S]methionine in the reaction mixture. Equal amounts of in vitro translated AhRs and nonlabeled ARNT were incubated with either 0.1% Me 2 SO or 10 nM TCDD (Cambridge Isotopes, Andover, MA) for 90 min at room temperature. Aliquots of the mixture were immunoprecipitated using either excess anti-HA antibody or excess anti-AhRpY9 antibodies. Pelleted protein was solubilized in SDS-loading buffer and separated on a 7.5% polyacrylamide gel by SDS-PAGE. [ 35 S]AhRs were visualized with a PhosphorImager.
PKC Activity Assays-For the peptide substrate PKC assays, peptides corresponding to the N terminus of the mouse AhR (residues 1-23) were synthesized by Alpha Diagnostic International. These peptides included the wild-type (AhR-(1-23)WT) sequence (MSSGANITYASRK-RRKPVQKTVK), the Y9F (AhR-(1-23)Y9F) mutant sequence (MSSG-ANITFASRKRRKPVQKTVK), and a sequence containing a phosphorylated tyrosine 9 (AhR-(1-23)pY9) (MSSGANIT(pY)ASRKRRKPVQKT-VK). A control peptide was synthesized corresponding to mouse AhR residues 383-406 (EEGREHLQKRSTSLPFMFATGEAVK). The specific PKC substrate peptide control (QKRPSQRSKYL) was purchased from Upstate Cell Signaling Solutions (Lake Placid, NY). All AhR peptides were conjugated with biotin on the C terminus and Ͼ96% pure as determined by mass spectrometry and high performance liquid chromatography. The aqueous peptides were monitored by amino acid analysis at the University of Rochester Proteomics Core Sequencing Facility to ensure that the appropriate amounts of peptides were used for each experiment. The PKC assays (using an active mixture of PKC ␣, ␤, and ␥) with the required reagents purchased from Upstate Cell Signaling Solutions were carried out as described by the manufacturer. For the PKC-elicited phosphorylation of full-length AhR, AhRs were expressed in the TNT system as described above, immunoprecipitated using either Rpt-9 or the anti-HA antibody and protein A/G-agarose plus (Santa Cruz Biotechnology). The pelleted immunocomplexes were washed extensively with IP wash buffer (40 mM Tris, 150 mM NaCl, and 1% Triton X-100, pH 7.4), and then either incubated for 15 min at 30°C with alkaline phosphatase or buffer and washed extensively in IP wash buffer. The washed immunocomplexes were then incubated for the indicated times at 30°C with an active mixture of PKC and [ 32 P]ATP, as described by the manufacturer. Pelleted protein was solubilized in SDS-loading buffer at the time point indicated and loaded to a 7.5% denaturing gel for SDS-PAGE. Protein was then transferred to PVDF membrane and detected using a PhosphorImager. The amount of AhR was analyzed by immunoblot analysis using anti-AhR antibodies or anti-HA antibodies as indicated. The primary antibodies were located with HRP-conjugated IgG and visualized with LumiGLO chemiluminescent substrate.
PMA Effect on Transcriptional Activity-AhR-deficient cells (2.0 ϫ 10 4 ) were seeded into each well of 24-well plates and incubated at 37°C. 60 -80 percent confluent cells were washed with serum-free medium and transfected with 250 l of serum-free medium containing 400 ng of the designated pHM6/HA-AhR-HIS construct, 100 ng of the DREdriven firefly luciferase reporter construct, p2DLuc, and 15 ng of Renilla luciferase vector pRL-TK (Promega) that was preincubated with 2.6 l of GenePORTER transfection reagents at room temperature for 45 min. Four hours later, another 250 l of growth media containing 20% fetal bovine serum was added into each well to achieve the final serum concentration of 10%. Fresh medium was added 24 h later. Forty-eight hours post-transfection, the triplicate wells were treated with either vehicle (0.15% Me 2 SO), 10 nM TCDD, 100 nM PMA, a combination of 10 nM TCDD and 100 nM PMA, or a combination of 10 nM TCDD, 100 nM PMA, 1 M GF 109203X, and 1 M RO 31-8220. PMA, GF 109203X, and RO 31-8220 were purchased from BioMol. In a separate experiment, similarly transfected cells were pretreated for 15 min with 4 M chelerythrine chloride (not shown) (Sigma) before the treatment performed above, as described previously (30). All treatment groups were incubated for 4 h before the cells were washed with phosphate-buffered saline and lysed in 100 l of lysis buffer as described by the manufacturer (Promega). The activities of firefly and Renilla luciferase were determined using the Dual-Luciferase reporter assay system. Twenty microliters of total cell lysate from each well were added into 80 l of luciferase reaction assay buffer, the luminescence was read using a TD20/20 Luminometer, and then 80 l of Stop and Glow reagent were added to the same tube to read luminescence representing Renilla luciferase activity. The ratio of firefly to Renilla activities was used for the final analysis and expressed relative to AhRWT treated with vehicle set at 100%.

AhRY9F Exhibits Normal Ligand Binding, Subcellular Localization, and ARNT Interaction-
We examined the possibility that the AhR tyrosine 9 may be required for ligand binding, nuclear localization, and/or ARNT interaction. The inhibition of DNA binding by the AhRY9F mutant was not due to decreased ligand binding, because specific binding of [ 3 H]TCDD to in vitro expressed WT and Y9F AhR-ARNT was equivalent as measured by the hydroxylapatite assay (21) (not shown). Tyrosine 9 is located adjacent to the basic cluster in the N terminus that contains the nuclear localization signal (see Fig.  1, NLS); thus, mutation of this residue may have the potential to disturb nuclear localization processes. The N-terminal 495 amino acid fragment of AhR is known to contain all domains for ligand binding, proper nuclear translocation, heterodimerization with ARNT, and association with the DRE sequence in the enhancer region of the CYP1A1 gene upon TCDD treatment (22). To monitor the intracellular location of the AhR, the N-terminal 495 residues of WT and the Y9F mutant AhRs was fused to the N terminus of green fluorescent protein cDNA to produce AhR495-GFP. Fluorescence microscopic analyses of transiently transfected Hepa1c1c7 cells with these AhR-GFP fusion proteins revealed that GFP, AhR495WT-GFP, and AhR495Y9F-GFP were present both in the cytosol and the nucleus in vehicle-treated cells. After TCDD treatment, however, both AhR495WT-GFP and AhR495Y9F-GFP were compartmentalized mainly in the nucleus, whereas the unconjugated GFP still remained distributed in both the cytosol and the nucleus (Fig. 2A). These data indicate that the nuclear localization signal in AhRY9F is still functional and recognized by the nuclear translocation machinery.
Studies were also conducted to determine whether DNA binding of the AhRY9F is interrupted by a defect in heterodimerization between the ligand-activated AhR and ARNT. In a co-precipitation assay with His 6 -tagged AhRs and [ 35 S]methionine-labeled full-length ARNT, the AhRY9F mutant was able to heterodimerize with ARNT as efficiently as wild-type AhR (Fig. 2B). Similarly, extracts were also prepared from TAO cells that had been transfected with AhR expression vectors. AhR was immunoprecipitated, and the level of AhR-associated ARNT was determined by Western blotting. The amount of ARNT co-immunoprecipitated with both WT and Y9F AhR was similar (not shown). Therefore, the interaction of the AhR with ARNT was not affected by the mutation of tyrosine 9 to phenylalanine, which is consistent with mutations to alanine, tryptophan, or serine (9, 10). Together, these data suggest that tyrosine 9 is not involved in ligand binding, nuclear localization, or ARNT interaction.
Decreased DNA Binding Activity of AhRY9F for DRE D, E, and A-Six distinct DREs (designated DRE A-F), containing a common core sequence but different flanking sequences, are in the upstream DNA region of the TCDD-responsive CYP1A1 gene (17). Southwestern analysis of nuclear extracts from wildtype, AhR-deficient, and ARNT-deficient mutant mouse hepatoma cell lines showed that DRE D is associated with two proteins that are neither AhR nor ARNT (23). One of them is C/EBP␣, which interacts with the CAACT sequence present in the DRE D but not in DRE A, C, or E. Compared with other DRE sequences, DRE D is the most active (17). Interaction of C/EBP␣ with the DRE D sequence was also verified in the glutathione S-transferase Ya subunit gene enhancer region, where C/EBP␣ further stimulates TCDD-induced AhR transcriptional activation most likely through further stabilization of the AhR-ARNT-DRE complex (24). The other molecule had an approximate mass of 95 kDa and was found to bind both DRE D and A (23). Dissociation of this protein from DNA may be necessary for the access of the AhR-ARNT complex to the same DNA binding sites. Thus, competition with this 95 kDa protein for these sites might determine the DNA binding potential of the AhR-ARNT complex (23). It is possible that the pattern of complex formation of the AhR-ARNT dimer with DRE D may be different from that with DRE E or A. Therefore, it was of interest to test whether the defect in DNA binding by AhRY9F is restricted specifically to association with DRE D. However, EMSAs demonstrated that the AhRY9F is not capable of binding to any DRE tested as compared with wild-type, suggesting that tyrosine 9 is required for intrinsic AhR DNA binding ability (Fig. 3A).
Disruption of AhR Transcriptional Activity by Mutation of Tyrosine 9 to Phenylalanine Is Not Cell-type Specific-To recapitulate the effect of tyrosine 9 mutation to phenylalanine in intact cells, both wild-type and AhRY9F were transiently expressed along with a DRE reporter gene in either AhR-deficient mouse hepatoma (TAO) or monkey kidney cells (COS-7). As expected, the AhRY9F exhibited significantly decreased transcriptional activity in both cell types (Fig. 3, B and C, respectively). These data demonstrate that the decrease in transcriptional activity because of a tyrosine 9 mutation is not restricted to a specific cell line.
AhRY9F Mutation Disrupts the Ligand-independent DNA Binding Activity of AhR Truncation Mutants-Next, we asked whether tyrosine 9 is required for constitutive DNA binding activity of a PAS-B domain-deleted AhR, because it might utilize a different tertiary structure for DNA binding that no longer requires the first nine amino acids (11). To address this question, a constitutively activated AhR, AhR346, was generated by eliminating only a minimal amount of the ligand binding domain. Truncation mutants with the wild-type sequence and the Y9F mutation were expressed in vitro and incubated with the similarly expressed full-length mouse ARNT in the presence or absence of TCDD. The AhR346 truncation mutant that has the wild-type sequence exhibited strong DNA binding activity in the vehicle-treated samples (Fig. 4A). The level of this constitutive DNA binding activity could not be further elevated by incubation with 10 nM TCDD. Mutation of tyrosine 9 to phenylalanine in the AhR346 PAS-B deleted receptor abolished the constitutive DNA binding activity and TCDD treatment failed to overcome the inhibition of DNA binding activity of AhR346Y9F mutant (Fig. 4A, lanes 7-10). Similar results were obtained with AhR331, even in the presence of the AhR antagonist, 3Ј-methoxy-4Ј-nitro-flavone (Fig. 4A, MNF; lanes 5 and 6 and Ref. 25). These data suggest that the DNA binding form of this truncated AhR shares a similar threedimensional conformation as that of the full-length AhR induced by ligands in which tyrosine 9 plays an essential role.
The N Terminus of the AhR Is Not Cleaved Prior to DNA Binding-It was previously suggested that the first 9 amino acids, including tyrosine 9, are not present in the active AhR due to proteolytic cleavage (15). This hypothesis was based on the N-terminal sequencing of the purified AhR from mouse liver that suggested the first residue to be what is now known as alanine 10. However, it is not clear whether the loss of these residues in the purified AhR was due to experimental artifacts or a reflection of a true physiological event (10). An AhR expression vector was constructed to express an N-terminally HA-tagged AhR to examine the hypothesis that the N-terminal of the AhR is proteolytically cleaved to allow normal DNA binding and that tyrosine 9 is required for this speculated cleavage. A supershift EMSA was performed to determine if the N terminus is absent from the AhR in the DNA-bound AhR-ARNT complex. In vitro translated AhRs were treated with 10 nM TCDD, and an EMSA was performed in the presence or absence of an HA antibody. Wild-type AhR tagged on either the N or C terminus (as a control) were supershifted by an HA antibody. Similarly, the AhRY9F, which only binds up to 20% of the level of wild-type AhR (13), was also completely supershifted with the HA antibody (Fig. 4B). These data demonstrate that the N terminus is not required to be cleaved from wild type or AhRY9F prior to/during DNA binding in vitro. Therefore, tyrosine 9 is not required for this speculated Nterminal cleavage to occur prior to DNA binding.
AhRY9F Reacts Similarly to Wild-type AhR with Anti-phosphotyrosine Antibodies, but Exhibits Altered Charged Forms-Previous reports indicated that the AhR requires tyrosine phosphorylation for its DNA binding activity (12,26). Furthermore, mutational analysis of mouse AhR tyrosine residues suggested that only a mutation at tyrosine 9, a computationally predicted site of phosphorylation, significantly decreases DNA binding activity and DRE-driven luciferase expression (13). These data raise the possibility that phosphorylation at tyrosine 9 may play a role in AhR DNA binding. Phosphorylation of the tyrosine 9 residue was investigated by immunoblot analysis using an anti-phosphotyrosine antibody, mutation to the phosphomimetic residue glutamate and two-dimensional gel electrophoresis. Both wild-type AhR and AhRY9F were conjugated with six consecutive histidines at their C-terminal ends to facilitate purification, and were transiently expressed in the TAO cell line. AhRs fractionated on Ni-NTA-agarose were sep-  5A). We tested several different anti-phosphotyrosine antibodies, some of which are known to have differential affinity and specificity for varying substrates. However, we were unable to detect a significant difference in the phosphotyrosine signals between wild-type AhR and AhRY9F (not shown). These data are consistent with the results from anti-phosphotyrosine immunoblotting of wild-type AhR and AhRY9F when they are synthesized in rabbit reticulocyte lysate (13). In addition, phosphotyrosine signals in both WT and AhRY9F remained similar before and after TCDD treatment (Fig. 5A). This indicates that the majority of tyrosine phosphorylation occurs constitutively, and/or tyrosine phosphorylation at several specific sites may change upon TCDD treatment allowing total tyrosine phosphorylation to remain quantitatively unchanged. These data suggest that, although the AhR is tyrosine-phosphorylated, tyrosine 9 may not be phosphorylated at all or, if it is, may not be a major phosphotyrosine residue in the AhR.
On the other hand, the fact that tyrosine 9 cannot be replaced with phenylalanine, which is unable to be phosphorylated but is the structurally most similar amino acid to tyrosine, suggests that phosphorylation of this position may affect the interaction of the AhR-ARNT complex with DNA. In some cases, phosphorylated amino acid residues can be mimicked by phosphomimetic amino acids, such as aspartate or glutamate, whose side chains possess negatively charged carboxyl groups. Therefore, site-directed mutagenesis was performed to replace tyrosine 9 with glutamate to determine whether DNA binding of the AhR involves the negatively charged phosphate group at tyrosine 9. EMSAs showed that in vitro transcribed and translated AhRY9E was unable to bind the DRE sequence when incubated with similarly expressed ARNT in the presence of TCDD (not shown). It also lacked transcriptional activity when it was transiently expressed in CV-1 cells and TAO cells with a DRE-driven reporter gene (not shown). These data are consistent with the suggestion that it is not just the negative charge of a putatively phosphorylated tyrosine 9 that is critical for DNA binding.
The AhR isolated from cells or synthesized in vitro has been shown to exhibit charge heterogeneity that is consistent with phosphorylation as determined by two-dimensional gel electrophoresis (13,27,28). Cytosolic 125 I-dioxin-labeled AhR exhibited an apparent isoelectric point (pI) range of 5.2-5.7, while the apparent pI nuclear AhR ranged from 5.5 to 6.2 (28). This differential charge heterogeneity is consistent with a means of dynamic post-translational modification occurring at some point between the cytosol and the nucleus, though no evidence has been generated to directly implicate phosphorylation at specific residues. The 35 S-labeled wild-type AhR and AhRY9F were separately synthesized in vitro and analyzed by twodimensional gel electrophoresis. Both AhRs showed charge heterogeneity similar to previous reports, but AhRY9F isoforms were more basic than the wild-type AhR (Fig. 5B). These results suggest that either tyrosine 9 is a phosphorylated site and/or that mutating tyrosine 9 results in a change in charge (for example, by post-translational modification) at other unidentified residues.
The Unoccupied, TCDD-bound, and DNA-binding Forms of the AhR Are Not Phosphorylated on Tyrosine 9 -To determine whether or not phosphorylation of AhR tyrosine 9 is required for DNA binding, we produced an antibody (anti-AhRpY9) for immunodetection of the putatively phosphorylated AhR tyrosine 9. The polyclonal antibody recognized specifically a phosphorylated tyrosine residue on synthetic peptides corresponding to AhR residues 4 -16, containing AhR tyrosine 9, and did not recognize a non-phosphorylated negative control peptide using both ELISA and immunoblot methodologies (not shown). To verify the specificity of the anti-AhRpY9 antibody for whole protein containing a phosphorylated tyrosine 9, BSA was conjugated to the phosphopeptide antigen (BSA-pY9). The anti-AhRpY9 antibody did react specifically with the BSA-pY9 (Fig.  6A, top panel, lane 1) but did not recognize the BSA-pY9 conjugate when it was dephosphorylated (Fig. 6A, top panel, lane  2). Interestingly, another polyclonal AhR antibody (immunogen, recombinant AhR N-terminal fragment (residues 1-402) obtained from BioMol) only recognized the dephosphorylated BSA conjugate (Fig. 6A, bottom panel, lane 2). These results suggest that the anti-AhRpY9 antibody is both specific and sensitive enough to detect an AhR-phosphotyrosine 9 sequence. However, the anti-AhRpY9 antibody did not recognize vehicle (not shown) or TCDD-treated AhRs synthesized in vitro, even when incubated with TCDD and ARNT (Fig. 6A, lanes 3 and 4). Also, the anti-AhRpY9 antibody did not recognize AhRs from FIG. 5. Analysis of phosphorylation differences between wildtype AhR and AhRY9F. A, anti-phosphotyrosine blot analysis of AhRWT and AhRY9F. TAO cells were transiently transfected with either pcDNA3/Neo (lanes 1 and 2), pcDNA3/␤AHR-HIS (lanes 3 and 4), or pcDNA3/␤AhRY9F-HIS (lanes 5 and 6). After 48 h of transfection, the cells were treated with either 0.1% Me 2 SO (lanes 1, 3, and 5) or 4 nM TCDD (lanes 2, 4, and 6) for 1 h prior to the lysis. The histidine-tagged AhR protein was purified with Ni-NTA-agarose. The purified AhR proteins were separated by SDS-PAGE, transferred to PVDF membrane, and probed with a non-sequence-specific anti-phosphotyrosine antibody (anti-PY) (top panel). The same membrane was stripped and reprobed with an anti-AhR antibody (bottom panel). B, effect of the AhRY9F mutation on charge heterogeneity of the AhR. AhRs were expressed in rabbit reticulocyte lysate supplemented with [ 35 S]methionine and solubilized for two-dimensional gel electrophoresis. Two-dimensional patterns were aligned within the pH range used of 3-9, and cropped within an approximate pH range of 5-9 to show clearly the individual charged forms of the receptors. The basic end (Ϫ) and the acidic end (ϩ) are marked below the figure. The streak at the basic extreme represents the end of the IPG strips and was used in part for alignment of the gels. Arrows are used to indicate the orientation for the 1st (isoelectric focusing, IEF) and 2nd dimensions (SDS-PAGE), as well as to designate individual charged forms of the WT AhR and the mutant AhRY9F (Y9F). 35 S-labeled AhR was visualized via a PhosphorImager. These data are representative of four individual experiments. transiently transfected and vehicle or TCDD-treated TAO cells (Fig. 6B).
It is possible that phosphorylation at AhRY9 might occur in only a small population of the AhR pool, specifically in the DNA binding form. We addressed this issue in two different ways. First, we labeled AhR with [ 35 S]methionine and attempted to concentrate any AhR phosphorylated on tyrosine 9 by immunoprecipitating with an excess amount of anti-AhRpY9. The HA-tagged AhRs that had been incubated with ARNT and TCDD to produce the DNA binding form of the AhR could be immunoprecipitated by an anti-HA antibody (Fig. 7A, lanes  1-4) but not by the anti-AhRpY9 antibody (Fig. 7A, lanes 5-8). Also, we hypothesized that the anti-AhRpY9 antibody would result in either a supershift or an attenuation of the TCDDmediated AhR-ARNT-DRE complex if the DNA binding form of the AhR is phosphorylated on tyrosine 9. However, the anti-AhRpY9 antibody failed to alter the normal mobility of the wild-type AhR-ARNT-DRE ternary complex when AhR and ARNT were synthesized in rabbit reticulocyte lysate (Fig. 7B,  pY9, lane 7) or prepared from Hepa1c1c7 cells (not shown). The anti-HA positive control antibody supershifted this complex (Fig. 7B, HA, lane 8) whereas the anti-AhR antibody, which recognizes residues within the DNA binding region of AhR, attenuated DNA binding as expected (Fig. 7B, Rpt9, lane 6). Together, these data are consistent with the interpretation that the AhRY9 is not a phosphorylated residue required for DNA binding.
Immunopurified AhR Is Phosphorylated by PKC, and AhRY9F Inhibits PKC-elicited Phosphorylation-The inability of the anti-AhRpY9 antibody to recognize the AhR together with the two-dimensional gel electrophoresis data suggesting that a mutation of tyrosine 9 still results in a less negatively charged receptor support a hypothesis that tyrosine 9 is required for normal post-translational modification of other AhR residues. There is much published data implicating the importance of phosphorylation for AhR activity and, in particular, PKC has been observed to be important in the regulation of AhR activity in whole cells (for example, Refs. 26 and 29 -32).
In an attempt to determine if PKC was capable of phosphorylating the full-length AhR, wild-type and mutant AhRs were immunoprecipitated from rabbit reticulocyte lysate and incubated with purified PKC and [ 32 P]ATP. Notably, wild-type AhR was phosphorylated by the purified PKC, demonstrating for the first time that PKC can directly phosphorylate the AhR (Fig. 8A). AhR from Hepa1c1c7 cells similarly immunopurified (using the anti-AhR antibody, Rpt-9) also demonstrated PKCelicited phosphate incorporation (not shown). Furthermore, AhRY9F incorporated less phosphate relative to wild-type AhR (Fig. 8A) even when immunopurified AhRs were treated with alkaline phosphatase prior to PKC incubation (not shown) indicating a difference in PKC substrate recognition and/or kinetics. The AhRY9F did incorporate a similar amount of labeled phosphate compared with AhRWT after a 16-h incubation with PKC. This demonstrates that the mutant is capable of incorporating phosphate but at a slower rate than the wild-type AhR (not shown). The control lane (Fig. 8A, Cntrl, lane 7) represents an immunopurification of lysate containing no AhR and subsequent incubation with PKC. The lower band that incorporates phosphate represents PKC autophosphorylation. These data suggest that tyrosine 9 is required for proper kinase substrate recognition of serine(s)/threonine(s) within the full-length AhR.
Synthetic AhRY9-mutated Peptides Are Phosphorylated Less Efficiently by PKC-To test the hypothesis that tyrosine 9 may be required for optimal phosphorylation of serine(s) and/or threonine(s) in the N terminus of AhR, we utilized peptides corresponding to the wild-type AhR residues 1-23 (AhR-(1-23)WT), the mutant (AhR-(1-23)Y9F) sequence, and a phosphorylated tyrosine 9 (AhR-(1-23)pY9) sequence. Interestingly, if we consider the phosphate incorporated into the AhR-(1-23)WT peptide as 100%, then significantly less phosphate was incorporated into the peptides that had the tyrosine mutated to a phenylalanine (69%) or to a phosphorylated tyrosine (36%) (Fig. 8B). These data, together, suggest that tyrosine 9 itself is not phosphorylated, but that a tyrosine at position 9 is required for normal substrate recognition by serine/threonine kinase(s), such as PKC.
PMA Augments AhRY9F Transactivation Activity-To test the hypothesis that the decreased activity of the AhRY9 mutant is, at least in part, due to less efficient PKC-elicited phosphorylation, we treated AhR-deficient cells that were transfected with the AhRs and a DRE-driven luciferase construct FIG. 6. The anti-AhR phosphotyrosine 9 antibody (anti-AhRpY9) does not recognize denatured AhRs. A, immunoblot controls and detection of denatured AhR with the anti-AhRpY9 antibody. BSA was conjugated to a peptide corresponding to AhR residues 4 -16, containing a phosphorylated tyrosine at the position representing tyrosine 9 to produce BSA-pY9. An aliquot of BSA-pY9 (lane 1) was dephosphorylated with a mixture of alkaline phosphatase (AP) and potato acid phosphatase (PAP) as a control (lane 2) for the specificity of the anti-AhRpY9 antibody. AhRWT (lane 4) and AhRY9F (lane 3) were translated separately in vitro in rabbit reticulocyte lysate and incubated with similarly translated ARNT in the presence of 10 nM TCDD for 90 min. The BSA conjugates and the AhR samples were loaded on a 7.5% denaturing gel for SDS-PAGE, and the protein was transferred to a PVDF membrane. The anti-AhRpY9 antibody was used for the immunoblot in the top panel, while polyclonal anti-AhR antibody (BioMol) was used for immunoblot detection in the bottom panel after stripping the membrane. B, immunodetection of AhR isolated from transiently transfected AhR-deficient cells with the anti-AhRpY9 antibody. TAO cells were transfected with AhR expression vectors (pHM6/AHRWT, lanes 4 and 5 and pHM6/AhRY9F, lanes 6 and 7). After 48 h of transfection, the cells were treated with either vehicle (0.1% Me 2 SO) (lanes 4 and 6) or 10 nM TCDD (lanes 5 and 7) for 1 h prior to cell lysis. The HA-tagged AhR protein was purified by immunoprecipitation using the anti-HA antibody. The purified AhR proteins were separated by SDS-PAGE, transferred to PVDF membrane, and probed with anti-AhRpY9 antibody (top panel). The same membrane was stripped and reprobed with the anti-AhR antibody (bottom panel). These data are representative of three independent experiments. with the PKC activator, PMA. The co-treatment of cells with TCDD and PMA has been shown to result in augmentation of TCDD-induced AhR-mediated gene transcription (30,31). This effect has been dubbed the "PMA effect," which has been suggested to be the result of PKC-elicited phosphorylation of the AhR and/or ARNT protein (30) (31,33). Compared with TCDD treatment alone, cells expressing the wild-type receptor demonstrated enhanced luciferase activity when co-treated with TCDD and PMA (Fig. 9). Consistent with the stated hypothesis, even cells expressing tyrosine 9 mutants (Y9A not shown) and co-treated with TCDD and PMA demonstrate a 30% increase in transcriptional activity. However, this activity was still below that observed for the cells containing the wild-type AhR co-treated with TCDD and PMA. This effect could be decreased by co-incubation with PKC inhibitors, GF 109203X and RO 31-8220, or pretreatment with chelerythrine chloride (not shown), as has been demonstrated previously (30,31). Cells expressing the canonical AhR DNA binding mutant, AhRR39A, a mutation of arginine 39, which is thought to directly contact DNA (8,9,11), do not demonstrate a PMA effect (Fig. 9). These results suggest that tyrosine 9 is important for phosphorylation of serine(s)/threonine(s) on the AhR that influences the induction of gene transcription.

DISCUSSION
AhR tyrosine 9 is conserved across species, suggesting a crucial role for this residue in AhR function. Previously, we have indicated that the DNA binding activity of a purified AhR-ARNT complex is controlled by tyrosine phosphorylation apparently directly on the AhR (12). We have demonstrated that the AhR is tyrosine-phosphorylated when synthesized in both rabbit reticulocyte lysate (13) and cells (Fig. 5A). In an attempt to identify the putative phosphotyrosine(s) in the AhR, we carried out investigations using site-directed mutagenesis. Of all the tyrosines located in the N-terminal region of the AhR (amino acids 1-399), only mutation of tyrosine 9 leads to significant loss in AhR DNA binding and transcriptional activity, suggesting that tyrosine 9 may be the phosphotyrosine required for AhR DNA binding (13). Additionally, two-dimensional separation based on size and charge of the AhR showed that there is a basic shift in charge distribution due to a single mutation at tyrosine 9, which suggests that this residue is phosphorylated itself and/or that the AhRY9F mutation alters normal covalent modification(s) at other sites (Fig. 5B). To determine conclusively if tyrosine 9 is itself phosphorylated or if it is otherwise involved in maintaining charge distribution on the AhR, an antibody was raised using a peptide antigen (AhR residues 4 -16 with a phosphotyrosine corresponding to tyrosine 9). This anti-AhRpY9 antibody did not immunoreact with any AhR examined, while all the controls indicate it is capable of specific immunodetection of a phosphorylated tyrosine 9 (Figs. 6 and 7). We conclude that tyrosine 9 is not a phosphorylated residue required for DNA binding but that it is required to maintain normal post-translational modification of other AhR residues.
Following our determination that a mutation of tyrosine 9 to phenylalanine did not interfere with ligand-activated nuclear localization or ARNT heterodimerization, we sought to determine a mechanism by which tyrosine 9 may be responsible for maintaining post-translational modification of the AhR. Treatment of AhRs synthesized in rabbit reticulocyte lysate with tyrosine or serine/threonine phosphatases resulted in a basic shift of the receptor as determined by two-dimensional gel electrophoresis (13). While continuing to analyze the hypothesis that tyrosine 9 is a phosphoresidue, peptides corresponding to AhR residues 1-23 (MSSGANITYASRKRRKPVQKTVK) were synthesized in order to evaluate whether or not tyrosine  1, 3, 5, and 7) and AhRY9F (lanes 2, 4, 6, and 8) were immunoprecipitated with either an anti-HA antibody (lanes 1-4) or the anti-AhRpY9 antibody (lanes 5-8).
The purified AhR proteins were separated by SDS-PAGE, transferred to PVDF membrane, and detected using a PhosphorImager. B, supershift EMSA of the AhR-ARNT DRE-bound complex using the anti-AhRpY9 antibody. HA-tagged wild-type AhR was translated in vitro in rabbit reticulocyte lysate and incubated with similarly translated ARNT in the presence of vehicle (0.1% Me 2 SO) or 10 nM TCDD for 90 min. The DNA binding activities of the AhR-ARNT complexes were measured with 32 P-labeled DRE D probe in a non-denaturing gel in the absence/presence of anti-AhR antibody (Rpt-9, lanes 3 and 6), the anti-HA tag antibody (HA, lanes 5 and 8), or the anti-AhRpY9 antibody (pY9, lanes 4 and 7). Expression level of AhR was monitored by immunoblot analysis of in vitro translated proteins (not shown). Lanes 1-2, vehicle and TCDD-treated AhR, respectively, with no antibody used. Lanes 3-5, vehicle-treated AhR with Rpt-9, anti-AhRpY9, or anti-HA antibody, respectively. Lanes 6 -8, TCDD-treated AhR with Rpt-9, anti-AhRpY9, or anti-HA antibody, respectively. These results are representative data of four independent experiments. 9, the only tyrosine in this sequence, could serve as a substrate for the unknown tyrosine kinase(s) present in the lysate. Under in vitro kinase assay conditions suitable for supporting tyrosine kinase activity, we detected no significant levels of ␥-32 P incorporation into these peptides (not shown). However, in a similar experiment using assay conditions suitable for maximal serine/ threonine kinase activity, we observed high ␥-32 P incorporation (not shown). In an attempt to identify the serine/threonine kinase(s) activity in the reticulocyte lysate responsible for this peptide and full-length AhR phosphorylation, kinase assays were performed utilizing synthetic peptide substrates for various serine/threonine kinases. We observed moderate PKC-like activity in reticulocyte lysate, and subsequently demonstrated the presence of PKC in rabbit reticulocyte lysate by immunoblotting. 2 To determine if PKC was capable of phosphorylating the AhR peptides, we incubated purified PKC with the peptides and [␥-32 P]ATP. Peptides with a mutated (Y9F) or modified (phosphorylated) tyrosine 9 incorporated significantly less phosphate than the wild-type sequence (Fig. 8B). These data suggest that tyrosine at position 9 may be critical for serine/ threonine kinase(s) recognition and that it may be this ability that accounts for some of the altered charge forms of AhRY9F consistent with a less phosphorylated AhR.
It was of particular interest to determine if PKC could directly phosphorylate the full-length AhR and whether or not a mutation at tyrosine 9 in the sequence interfered with this phosphorylation. We demonstrate for the first time that the AhR can be phosphorylated by PKC (Fig. 8). Additionally, as hypothesized from the peptide data, AhRY9F was phosphoryl-  9. PMA augments transactivation activity of the AhRY9 mutant. TAO cells were transiently transfected with AhR expression vectors, DRE-driven luciferase reporter gene, and Renilla luciferase normalization vector using GenePORTER. After recovery, the triplicate wells were treated with either vehicle (0.15% Me 2 SO), 10 nM TCDD, 100 nM PMA, or a combination of 10 nM TCDD and 100 nM PMA. All treatment groups were incubated for 4 h prior to cell lysis and the luciferase assay. Relative firefly luciferase activity, normalized to Renilla luciferase activity, is shown as the percent of vehicle (Me 2 SO)treated wild-type AhR (AhR-WT) (ϮS.D. of triplicate analyses). ated less efficiently as compared with wild type (Fig. 8A). AhRs immunopurified with the anti-AhR antibody, Rpt-9, from Hepa1c1c7 cell extracts were also phosphorylated by PKC (not shown). Overall, it is apparent that PKC phosphorylation of AhR serine(s)/threonine(s) can be decreased by mutation of tyrosine 9.
PKC activity has been linked to TCDD-mediated AhR-dependent processes (26, 29 -32). The AhR and PKC signaling pathways apparently converge as TCDD increases PKC activity that then enhances TCDD-elicited AhR transcriptional activity. Furthermore, compounds structurally related to TCDD also increase PKC-elicited phosphorylation activity in vitro and in vivo, in various cell lines (34). Several studies have suggested that PKC-elicited phosphorylation causes a synergistic increase in TCDD-induced, AhR-mediated CYP1A1 gene induction (29,30,31,35). Co-treatment of cells with TCDD and PMA induces AhR-mediated gene transduction above levels induced by saturating concentrations of TCDD (30,31). This PMA effect was not caused by message/gene product stabilization or stimulation of basal transcription machinery. Also, this PKC stimulation did not alter the protein levels of AhR or ARNT or the TCDD-induced down-regulation of the AhR, nor did PKC stimulation result in changed nuclear accumulation levels of the AhR-ARNT heterodimer (30,31). Furthermore, domain-swapping analyses revealed that neither the transactivation domain (TAD) of AhR (see Fig. 1) or ARNT was required for the PMA effect (31). These results suggest that the PKC activity is required either for activation of co-activators that interact with the N terminus of the AhR and/or ARNT and/or that PKC can directly phosphorylate the N terminus of the AhR and/or ARNT. Together, the results presented in this work demonstrate that PKC can directly phosphorylate the AhR (Fig. 8A). Furthermore, the AhRY9F "DNA binding mutant" can be made to induce transcription when PKC is stimulated in cells while mutation of a residue that is thought to directly contact DNA, AhRR39A, cannot be made to induce gene transcription (Fig.  9). These results suggest that enhanced PKC activity may partially overcome the inhibitory effect of a mutation of tyrosine 9, and PKC may be directly phosphorylating the AhR to mediate this effect.
By what mechanism may PKC-elicited phosphorylation or similar serine/threonine kinase-elicited phosphorylation be responsible for AhR activity? It has been shown that a basic leucine zipper transcription factor, Nrf2, could be phosphorylated on serine 40 by PKC. Lack of phosphorylation on a Nrf2-S40A mutant resulted in enhanced interaction with a cytosolic dimerization partner, Keap1, resulting in decreased antioxidant response element-mediated transcription (36). It may be that altered phosphorylation status of the AhR results in changed affinity for one or multiple associating proteins, such as XAP2 and/or the importins, in the cytoplasm and/or the nucleus resulting in altered DNA binding and transcriptional activity. Also, although this distinct composition of the AhR DNA binding domain has been proposed to be essential for binding to a unique sequence (5Ј-GCNA-3Ј) of the DRE rather than a conventional E-box sequence, it is not known how two basic clusters separated by about 20 amino acids can make contact with a short DNA sequence of about four nucleotides. In order for these two separate domains to recognize the same four-nucleotide-long sequence (5Ј-GCNA-3Ј), there must be a mechanism that brings the two clusters into the small region of the DNA double helix. Current findings suggest that the tyrosine 9 in the vicinity of the first basic region may fulfill this role. It is conceivable that tyrosine 9 participates in a conformational change, possibly induced by phosphorylation at adjacent serine(s)/threonine(s) that are necessary to pull the first basic cluster, containing arginine 14, into the proximity of the second basic region, containing arginine 39, perhaps requiring another associated protein, and thus maintains an optimal three-dimensional structure of the unique AhR DNA binding domain.
We conclude that AhR tyrosine 9, a conserved residue, is not a phosphorylated residue but that it is required for appropriate AhR phosphorylation by serine/threonine kinase(s). These data allow us to reexamine the hypothesis that tyrosine phosphorylation of the AhR is required for DNA binding, because the two most likely candidates, tyrosine 372 in the major region of phosphorylation (12) and now tyrosine 9, can be ruled out. The major evidence in support of this hypothesis was the data demonstrating that AhR not ARNT phosphorylation is required for DNA binding (26) and the ability of tyrosine-but not serine/threonine-specific phosphatase treatment of purified rat AhR-ARNT to disrupt DRE binding (12). A possible explanation for these results is that multiple tyrosines, not just a single tyrosine, were dephosphorylated leading to a significant conformational change and inability to bind DRE. The data presented here also focus on the potential functional role of AhR serine/threonine(s) phosphorylation sites. There are five sites of the AhR that correspond to a PKC motif: serine 11, threonine 21, serine 35, threonine 376, and threonine 429. Data already exist suggesting that DNA binding can be altered by mutation of serine 11, threonine 21, or serine 35 to alanines (8 -10). It is important to note that although these residues are predicted PKC target sites, each has a context sequence suitable for several serine/threonine kinases, including multiple PKC isotypes. For example, serine 11 also has a sequence context that may serve as a substrate for cyclin-dependent protein kinase 5 (Cdk5) and casein kinase 1 (CK1), while the sequence context of serine 35 is suitable for phosphorylation by Cdk1 and CK1. 2 It is becoming clear that many transcription factors are phosphorylated and dephosphorylated within and/or surrounding their nuclear import signals and nuclear export signals to regulate protein-protein interactions and compartmentalization (37). While this manuscript was in preparation, for example, it was reported that phosphorylation of human serine 68 within the AhR nuclear export signals by p38 mitogen-activated protein kinase plays a role in subcellular localization of the AhR (38). Studies are under way to address the hypothesis that altered phosphorylation at possibly more than one residue by kinases in addition to PKC, because of a mutation of tyrosine 9 near the AhR nuclear import signals, can significantly impair the ability of the AhR to interact with certain proteins, bind to DNA, and induce transcription.