Nuclear Localization and Export Signals of the Human Aryl Hydrocarbon Receptor*

The aryl hydrocarbon receptor (Ahr) is a ligand-acti-vated transcription factor that binds DNA in the form of a heterodimer with the Ahr nuclear translocator (hy-poxia-inducible factor 1 b ). We found in this study that Ahr contains both nuclear localization and export signals in the NH 2 -terminal region. A fusion protein com- posed of b -galactosidase and full-length Ahr translocates from the cytoplasm to the nucleus in a ligand-dependent manner. However, a fusion protein lacking the PAS (Per-Ahr nuclear translocator-Sim homology) domain of the Ahr showed strong nuclear localization activity irrespective of the presence or absence of ligand. A minimum bipartite Ahr nuclear localization signal (NLS) consisting of amino acid residues 13–39 was identified by microinjection of fused proteins with glutathione S -transferase-green fluorescent protein. A NLS having mutations in bipartite basic amino acids lost nuclear translocation activity completely, which may explain the reduced binding activity to the NLS receptor, PTAC58. A 21-amino acid peptide (residues 55–75) containing the Ahr nuclear export signal is sufficient to direct nuclear export of a microinjected complex of glutathione S -transferase-Ahr-green fluorescent protein. These findings strongly suggest that Ahr act as a ligand-and signal-dependent nucleocytoplasmic cytosol. The nuclear rim targeting assay was performed by the addition of purified recombinant PTAC58 and PTAC97 proteins in the testing solution followed by incubation at 4 °C for 30 min. After incubation, the cells were fixed with 4% formaldehyde, and localization of GST-Ahr-GFP fusion proteins was examined under a fluorescent microscope. Binding Assay— GST-Ahr-GFP proteins, which were adjusted with the same amounts by monitoring the absorbance at 280 nm, were incubated with glutathione-Sepharose 4B (Pharmacia) for 1 h at 4 °C and then washed three times with transport buffer. The Sepharose was incubated with purified PTAC58 and PTAC97 in transport buffer containing 1 mg/ml BSA, 1 m M of phenylmethylsulfonyl fluoride, and 0.05% CHAPS. After incubation for 1 h at 4 °C, the Sepharose was washed three times with transport buffer, added to lysis buffer (50 m M Tris- HCl, pH 8.3, containing 500 m M NaCl, 1 m M EDTA, 2 m M dithiothreitol, and 0.2 m M phenylmethylsulfonyl fluoride) and boiled for 10 min. The eluted proteins were separated by 8% SDS-polyacrylamide gel electro-phoresis, and blotted onto a nitrocellulose membrane. After incubation with blocking buffer containing 3% gelatin in TBS (20 m M Tris-HCl, pH 7.5, 150 m M NaCl) for 30 min, the filter was rinsed by TBS and incubated with anti-PTAC58 rabbit antibody (25) for 1 h. The filter was washed three times with TBS containing 0.05% Tween 20 and then incubated with alkaline phosphatase-labeled anti-rabbit IgG (Bio-Rad) for 1 h. After washing with the same buffer followed by TBS, the bound antibodies were detected using an alkaline phosphatase conjugate substrate kit (Bio-Rad). Cells— The amplified fragment of Ahr(55–75) wt using F57/R67 as primers encodes the 21-amino acid sequence of PFPQDVINKLDKLSVLRLSVS. GST-Ahr(55–75) wt -GFP or GST-Ahr(55–75) mut -GFP, which have the substitutions of Leu 70 and Leu 72 for alanines, were injected into MDBK cell nuclei along with Texas Red-labeled BSA, which was coinjected to ensure a clean nuclear injection without leakage. Cells were either injected then and incubated at 37 °C for 30 min or were equilibrated to 4 °C for 15 min, injected, and then incubated at 4 °C for 30 min. After incubation, the cells were fixed with 4% formaldehyde in phosphate-buffered saline, pH 7.4, for 15 min, and localization was observed by fluorescent microscopy. The NES peptide of PKI corresponding to the sequence of residues 35–49 of NELALKLAGLDINKT was synthesized. The synthetic peptide was conjugated to BSA with the bifunctional cross-linking reagent sulfo- succinimidyl-4-( N -maleimidomethyl)-cyclohexane-1-carboxylate.

The aryl hydrocarbon receptor (Ahr) 1 binds a variety of environmentally important carcinogens, including polycyclic aromatic hydrocarbons and certain halogenated aromatic hydrocarbons such as 2,3,7,8-tetrachlorodibenzo-p-dioxin. Before binding ligands, Ahr is located in the cytoplasm as one component of a complex that has a molecular mass of about 280 kDa (1). This complex is composed of Ahr, two molecules of the 90-kDa heat shock protein, and possibly a 43-kDa protein (2). After ligand binding, Ahr dissociates from the complex and translocates to the nucleus (3). The heterodimer of Ahr and Ahr nuclear translocator (ARNT) constitutes a transcription factor and binds specific DNA sequences called XREs (xenobioticresponsive elements) in the enhancer regions of the CYP1A1 and several other proteins involved in xenobiotic metabolism (4). Because these enzymes are involved in the metabolism of polycyclic aromatic hydrocarbons to active genotoxic metabolites, Ahr plays an important role in carcinogenesis caused by these compounds (5)(6)(7).
Because ARNT was first cloned as a factor required for ligand-dependent nuclear translocation of Ahr from the cytoplasm to the nucleus (8), the subcellular localization of ARNT was believed to be cytoplasmic. In fact, most ARNT was recovered in the cytosolic fraction by cell fractionation. However, immunohistochemical analysis has shown that ARNT is localized predominantly in the nucleus, regardless of the presence or absence of ligands (9,10). This controversial subject was clarified by our recent study in which a nuclear localization signal (NLS) of the amino acid residues between 39 and 61 of human ARNT was found to be a novel bipartite type recognized by the two components of nuclear pore-targeting complex (11). Because the heterodimeric partner Ahr is present in the cytoplasm in the absence of ligands and translocates to the nucleus upon binding of ligands even in the ARNT-deficient cell line, Hepa-1 c4, these two subunits may translocate independently to the nucleus, where they may form a heterodimer to bind to the cognate DNA sequence, XRE. This finding prompted us to investigate the molecular translocation mechanism of Ahr from the cytoplasm to the nucleus in a ligand-dependent manner.
Active transport of proteins between the nucleus and cytoplasm is a major process in eukaryotic cells (12,13). Transport of proteins across the nuclear pore is generally selective and signal-dependent. Active import of proteins into the nucleus requires the presence of NLS. NLSs of various proteins identified so far can be classified mainly into two classes: 1) a single cluster of basic amino acids represented by the SV40 large T antigen NLS, and 2) a bipartite type, in which two sets of adjacent basic amino acids are separated by a stretch of approximately 10 amino acids (12,13). On the other hand, nuclear export signals (NESs) have been found recently in the human immunodeficiency virus (HIV) Rev protein (14), a cAMPdependent protein kinase (protein kinase A) inhibitor (PKI; 15), the fragile X mental retardation protein (FMRP) (16), and mitogen-activated protein kinase kinase (MAPKK) (17). The characteristic of the NESs defined was certain leucine-enriched amino acid stretches (18). Two novel signals found to direct both import and export were the M9 domain of human nuclear ribonucleoprotein A1 (19) and KNS of human nuclear ribonucleoprotein K (20).
The NLS-dependent nuclear import process requires at least four different proteins that act in a sequential manner with NLS-containing proteins. There appear to be several discrete steps in the import process which involve: 1) binding of the NLS receptors (importin ␣, karyopherin ␣, PTAC58) to an NLS; 2) complex formation in conjunction with importin ␤ (karyopherin ␤, PTAC97); 3) targeting nuclear pore proteins; and 4) ATP/ GTP-dependent translocation through the nuclear pore mediated by Ran (12,13,21). A number of NLS receptors have been identified recently, suggesting that there is a family of these NLS-binding proteins (22)(23)(24)(25)(26)(27)(28). Differential expression and sequence-specific interaction of NLS receptors with various types of NLSs have been reported recently (29), indicating that NLS receptors may play a role in regulating nuclear protein transport. A novel receptor-mediated nuclear protein import pathway recognized by transportin has also been reported in the case of M9-containing human nuclear ribonucleoprotein A1 protein (30). However, the molecular mechanism for the export of proteins from the nucleus is much less well understood than the import process at present.
In the present study we investigated the nuclear translocation of Ahr using transient expression of chimeric constructs of Ahr and ␤-galactosidase (␤-gal) in the presence or absence of ligand, clarifying the ligand-dependent nuclear localization of the Ahr protein. Subsequent analysis of various portions of Ahr using ␤-gal fusions as well as fusion protein with GST-GFP (green fluorescent protein) gave the minimum NLS consisting of amino acids 13-39, which completely overlaps the DNA binding domain of Ahr. We investigated the molecular mechanisms of the nuclear translocation of Ahr further using an in vitro system and also found signal-dependent nuclear export activity.

MATERIALS AND METHODS
Cell Cultures-Cell lines used for this study were the mouse hepatoma Hepa-1 clone Hepa 1c1c7, Hepa-1 c4 mutant, which lacks ARNT expression, generously provided by Dr. O. Hankinson, as well as HeLa and Madin-Darby bovine kidney (MDBK) cells. Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum at 37°C with 5% CO 2 atmosphere.
Plasmid Construction-Human Ahr cDNA was prepared by polymerase chain reaction amplification of reverse transcribed products of total RNA from HepG2 cells using specific primers and Pfu DNA polymerase (CLONTECH); this DNA was then inserted in the pGEM-7Zf(ϩ) vector (Promega). The sequence of the construct was confirmed by sequencing using fluorescein-labeled SP6 and T7 primers, AutoRead Sequencing kits, and an A.L.F.II DNA sequencer (Pharmacia Biotech Inc.). Subsequent cloning into the pSV␤-gal (Promega) was performed as described previously (11). The BglII-BglII fragment of the Ahr cDNA was ligated to the BglII site of the modified ␤-gal control vector to generate the ␤-gal/Ahr(1-848) vector.
Various portions of Ahr cDNA were amplified by polymerase chain reaction using the ␤-gal/Ahr(1-848) vector as a template and Pfu DNA polymerase with specific sets of primers to generate artificial BglII sites at both ends. The sequences of the primers used for the preparation of fragments of Ahr were as follows: F1, GTC TGG TGT CAA AAA CAG ATC TGC ATG and R24, TTCAGA TCT TAA GGG ATC CAT TAT GGC AGG for Ahr(1-427); F1 and R6, TTC AGA TCT TAA TAA GAA TTC TCC TTC TTG for Ahr(1-119); F4, TAT AAG ATC TGC CAG GCT CTG AAT GGC TTT and R5, TTC AGA TCT TAG TGG TCT CTG AGT TAC AAT for Ahr(120 -386); F5, TAT AAG ATC TGC ACA GAT GAG GAA GGA ACA and R4, CCT GCC CGG TTA TTA TTA AGA TCT CAG for Ahr(387-848); F1 and R48, TTC AGA TCT TCG GTC TCT ATG CCG CTT GGA for Ahr . After cleavage with BglII, the fragments were ligated to the BglII site of modified pSV␤-gal control vector to give in-frame fusion genes.
To construct GST-Ahr-GFP fusion genes, the GST-GFP2 cassette vector was prepared as described previously (11). Sequences of the primers used for the preparation of the fragments of Ahr were as follows: F36, TAT AAG ATC TGC CGC AAG CGG CGG AAG CCG and R48 for Ahr(13-42); F36 and R51, TTC AGA TCT GTC TCT ATG CCG CTT GGA AGG for Ahr(13-41); F36 and R52, TTC AGA TCT TCT ATG CCG CTT GGA AGG ATT for Ahr ; F36 and R53, TTC AGA TCT ATG CCG CTT GGA AGG ATT TGA for Ahr ; F36 and R54, TTC AGA TCT CCG CTT GGA AGG ATT TGA CTT for Ahr ; F36 and R55, TTC AGA TCT CTT GGA AGG ATT TGA CTT GAT for Ahr ; F40, TAT AAG ATC TGC AAG CGG CGG AAG CCG GTG and R48 for Ahr (14 -42); F36 and R58, TTC AGA TCT TCG GTC TGC ATG CCG CTT GGA for Ahr  After cleavage with BglII, the fragments were ligated to the BamHI site of the GST-GFP2 vector as described previously (11). The direction of inserts was determined by sequencing. For the preparation of Ahr(13-42, ⌬19 -35), two oligonucleotides F38, GAT CTC TCG CAA GCG GCG GAA GCC GTC CAA GCG GCA TAG AGA CCG AA, and R57, GAT CTT CGG TCT CTA TGC CGC TTG GAC GGC TTC CGC CGC TTG CGA GA, were annealed and phosphorylated. Construction of GST-NLSc-GFP, which contained the core sequence of NLS of SV40 large T antigen (31), was prepared as described previously (11).
DNA Transfections into HeLa, Hepa-1, and Hepa-1 c4 Mutant Cells-Cells (3.5 ϫ 10 6 ) in 400 l of potassium phosphate-buffered saline were transfected with 15 g of each ␤-gal fusion protein expression vector by electroporation using the Gene Pulser (Bio-Rad) with the voltage and capacitance set at 450 V and 960 microfarads, respectively. After transfection, cells were cultured for 48 h with or without 1 M MC (3methylcholanthrene). In situ staining of ␤-gal was carried out as described previously (11). Immunofluorescence staining of Ahr was carried out as described previously (10) using polyclonal antibody against Ahr (BIOMOL) as a primary antibody and fluorescein isothiocyanate-labeled anti-rabbit IgG (Organon Teknika-Cappel) as a secondary antibody.
Preparation of GST-Ahr-GFP Fusion Proteins-The GST-Ahr-GFP vectors described above were introduced into the Escherichia coli strain BL21. Purification of expressed GST-Ahr-GFP fusion proteins was carried out as described previously (11). The purified proteins were dialyzed against 20 mM Hepes buffer, pH 7.3, containing 110 mM potassium acetate, 5 mM sodium acetate, 2 mM magnesium acetate, and 1 mM EGTA. The purified fusion proteins were analyzed using 10% acrylamide gel and showed a major protein band of 55 kDa (data not shown).
Microinjection into Cytosol of HeLa Cells-Microinjection experiments were performed essentially as described previously (32). Various GST-Ahr-GFP fusion proteins were injected into the cytoplasm of HeLa cells. After microinjection, the cells were incubated at 37°C for 30 min before fixation with 4% formaldehyde. Localization of the fusion proteins was examined by fluorescent microscopy.
Recombinant Expression and Purification of PTAC58 and PTAC97-Recombinant PTAC58 and PTAC97 (33) were expressed in BL21 as GST fusion protein as described previously (25). The fusion proteins were purified using glutathione-Sepharose affinity chromatography. Finally, recombinant proteins of PTAC58 and PTAC97 were obtained by cleavage with thrombin to release the GST portion.
Cell-free Import Assay-Preparation of total cytosol of Ehrlich ascites tumor cells was conducted as described previously (34). Digitonin-permeabilized MDBK cells were prepared as described previously (35) based on the method of Adam et al. (36). The testing solution (10 l) consisted of GST-Ahr-GFP and transport buffer (20 mM Hepes, pH 7.3, 110 mM potassium acetate, 2 mM magnesium acetate, 5 mM sodium acetate, 1 mM EGTA, and 2 mM dithiothreitol) containing 1 g/ml each aprotinin, leupeptin, and pepstatin. Transport assay was performed in the presence or absence of cytosol with 1 mM ATP, 5 mM creatine phosphate, and 20 units/ml creatine phosphokinase at 37°C for 30 min. For the nuclear transport assay using purified preparations, recombinant PTAC58, PTAC97, Ran and p10 proteins were added instead of cytosol. The nuclear rim targeting assay was performed by the addition of purified recombinant PTAC58 and PTAC97 proteins in the testing solution followed by incubation at 4°C for 30 min. After incubation, the cells were fixed with 4% formaldehyde, and localization of GST-Ahr-GFP fusion proteins was examined under a fluorescent microscope.
Binding Assay-GST-Ahr-GFP proteins, which were adjusted with the same amounts by monitoring the absorbance at 280 nm, were incubated with glutathione-Sepharose 4B (Pharmacia) for 1 h at 4°C and then washed three times with transport buffer. The Sepharose was incubated with purified PTAC58 and PTAC97 in transport buffer containing 1 mg/ml BSA, 1 mM of phenylmethylsulfonyl fluoride, and 0.05% CHAPS. After incubation for 1 h at 4°C, the Sepharose was washed three times with transport buffer, added to lysis buffer (50 mM Tris-HCl, pH 8.3, containing 500 mM NaCl, 1 mM EDTA, 2 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride) and boiled for 10 min. The eluted proteins were separated by 8% SDS-polyacrylamide gel electrophoresis, and blotted onto a nitrocellulose membrane. After incubation with blocking buffer containing 3% gelatin in TBS (20 mM Tris-HCl, pH 7.5, 150 mM NaCl) for 30 min, the filter was rinsed by TBS and incubated with anti-PTAC58 rabbit antibody (25) for 1 h. The filter was washed three times with TBS containing 0.05% Tween 20 and then incubated with alkaline phosphatase-labeled anti-rabbit IgG (Bio-Rad) for 1 h. After washing with the same buffer followed by TBS, the bound antibodies were detected using an alkaline phosphatase conjugate substrate kit (Bio-Rad).
Microinjection into Nuclei of MDBK Cells-The amplified fragment of Ahr(55-75) wt using F57/R67 as primers encodes the 21-amino acid sequence of PFPQDVINKLDKLSVLRLSVS. GST-Ahr(55-75) wt -GFP or GST-Ahr(55-75) mut -GFP, which have the substitutions of Leu 70 and Leu 72 for alanines, were injected into MDBK cell nuclei along with Texas Red-labeled BSA, which was coinjected to ensure a clean nuclear injection without leakage. Cells were either injected then and incubated at 37°C for 30 min or were equilibrated to 4°C for 15 min, injected, and then incubated at 4°C for 30 min. After incubation, the cells were fixed with 4% formaldehyde in phosphate-buffered saline, pH 7.4, for 15 min, and localization was observed by fluorescent microscopy. The NES peptide of PKI corresponding to the sequence of residues 35-49 of NELALKLAGLDINKT was synthesized. The synthetic peptide was conjugated to BSA with the bifunctional cross-linking reagent sulfosuccinimidyl-4-(N-maleimidomethyl)-cyclohexane-1-carboxylate.

Requirement of the NH 2 -terminal Portion of Ahr for Nuclear
Localization-To study the molecular mechanism of ligandand NLS-dependent nuclear translocation of Ahr, fusion protein composed of ␤-gal and full-length Ahr was transiently expressed in HeLa, Hepa-1, and ARNT-deficient Hepa-1 c4 mutant cells. Representative profiles of expressed fusion proteins visualized by in situ staining of ␤-gal with X-gal are shown in Fig. 1A. No staining of the nucleus was observed for the expressed ␤-gal alone (Fig. 1A, a), whereas the fusion protein of ␤-gal with the NLS of SV40 large T antigen (␤-gal/SV40 NLS) showed strong nuclear localization in transfected cells (Fig. 1A, b). When the chimeric gene of ␤-gal/Ahr(1-848) was expressed in the absence of MC, the fusion protein was localized in the cytoplasm. However, when the chimeric gene was expressed in the presence of MC, the fused product was clearly localized in the nucleus of all three cell lines tested even in the ARNT-deficient Hepa-1 c4 mutant (Fig. 1A, c).
To determine the region of Ahr required for nuclear localization, various portions of the human Ahr cDNA were amplified by polymerase chain reaction and ligated to a modified ␤-gal expression vector. The chimeric constructs were introduced into HeLa cells, and their localization was examined (Fig. 1B). Deletion of the transactivation domain located in the COOHterminal portion did not affect the MC-dependent nuclear localization (Fig. 1B, b). However, when both transactivation and PAS domains were deleted, resulting in a lack of the ligand and hsp90 binding domains (37), the fused product was clearly detected in the nucleus regardless of the absence or presence of MC (Fig. 1B, c). In contrast, fusion proteins of either ␤-gal/ Ahr(120 -386) (Fig. 1B, d) or ␤-gal/Ahr(387-848) (Fig. 1B, e) showed cytoplasmic localization irrespective of the ligand, con-firming the absence of NLS in these regions. Because the NLS identified so far contained a cluster(s) of basic amino acids, two candidates of NLS of Ahr can be estimated in the NH 2 -terminal portion (Ahr(1-119)), one of which is located between 13 and 17 (RKRRK), and the other is between 37 and 42 (KRHRDR) in the bHLH domain. Accordingly, we focused our attention on the relationship between the two clusters of basic amino acids in the NH 2 -terminal portion.
Region of Ahr Required for Nuclear Localization- Fig. 2A shows a schematic representation of Ahr , which includes the two clusters of basic amino acids. When a chimeric gene of ␤-gal/Ahr(1-42) was expressed transiently in the absence of MC, the fusion protein was localized in the nucleus visualized both by in situ staining with X-gal (Fig. 2B, a) or by immunohistochemical staining with antibody to Ahr (Fig. 2B, b). The weak cytoplasmic staining by the antibody may be explained either by the endogenous expression of Ahr in HeLa cells or nonspecific interaction between the antibody and cells (Fig.  2B, c). To confirm the direct nuclear translocation activity of Ahr(1-42), we next examined the fate of purified recombinant protein of GST-Ahr(1-42)-GFP microinjected into the cytoplasm of HeLa cells as described previously (11). As was seen for the transient expression of ␤-gal fusions (Fig. 2B, a and b), the GST-GFP fusion protein showed efficient nuclear localization within a 30-min incubation at 37°C (Fig. 2B, d), confirming that the fragment serves as an NLS.
To determine the minimum length of NLS of Ahr, we first constructed a plasmid by insertion of the Ahr(13-42) fragment into the fusion gene of GST-GFP, using a minimum fragment containing two complete clusters of basic amino acids. A representative profile of nuclear localization of the microinjected fusion protein is shown in Fig. 3A, a). Next, GST-GFP fused proteins, which contain a series of deleted fragments from the COOH or NH 2 terminus of Ahr , were purified and microinjected into the cytoplasm of the HeLa cells (Fig. 3A). When amino acid residues from Arg 42 to Arg 40 were deleted successively from the COOH terminus of Ahr(13-42), clear nuclear localization was observed (Fig. 3A, b-d). However, when a deletion from Arg 42 to His 39 was carried out, both nuclear and cytoplasmic localizations were observed (Fig.  3A, e). On the other hand, GST-Ahr(13-37)-GFP did not translocate to the nucleus (Fig. 3A, f). We next constructed a deletion mutant of NH 2 -terminal Arg 13 and found clear cytoplasmic localization of the gene product (Fig. 3A, g), showing that Arg 13 was unequivocally essential for the nuclear translocation of Ahr. Furthermore, when the two separated basic regions were linked directly (Ahr 13-42, ⌬19 -35), the fusion protein was localized in the cytoplasm (Fig. 3A, h), indicating that amino acids between 19 and 35 were also required as spacers to provide enough length between the two separated clusters of basic amino acids to interact with the receptor(s). Using an in vitro nuclear transport assay system (Fig. 3B), we also confirmed clear nuclear translocation activity in Ahr(13-39) incubated with cytosol (Fig. 3B, b) or mixture of purified proteins PTAC58, PTAC97, Ran, and p10 (Fig. 3B, d) in the presence of ATP at 37°C. To conclude, the minimum length of Ahr NLS was estimated to be composed of 27 amino acid residues between Arg 13 and His 39 .
Essential Amino Acids Involved in Nuclear Translocation of Ahr-To identify which amino acid residue(s) are essential for full NLS activity, mutational analysis in the region of Ahr  was performed, and the results obtained are shown in Fig.  4A. When substitutions of Arg 13 , Lys 14 , and Arg 15 to Ala were performed and the mutated gene products were microinjected into the cytoplasm of HeLa cells, they completely lost their nuclear translocation activity (Fig. 4A, b-d, respectively). The gene product with R16A was localized in both the nucleus and cytoplasm (Fig. 4A, e), and clear nuclear localization was observed in the case of fused protein containing K17A (Fig. 4A, f), thereby indicating that the three basic amino acids between residues 13 and 15 were essential in NLS activity as part of the basic amino acid cluster. An essential role of basic amino acids Lys 37 and Arg 38 in another part of bipartite NLS was similarly observed (Fig. 4A, i and j). The H39A mutant also had reduced the NLS activity (Fig. 4A, k), since both cytoplasmic and nuclear localizations were observed.
We developed great interest in Ser 36 , which lies within a recognition sequence for phosphorylation by protein kinase C, since nuclear localization is often regulated by phosphorylation of Ser or Thr close to the cluster of basic amino acids (38). However, nuclear localizations of the fused protein containing S36A and wild type protein were indistinguishable, indicating that Ser 36 does not participate in the NLS activity (Fig. 4A, h).
Mutations in the NLS of Ahr Resulted in Reduced Targeting to the Nuclear Rim-Because only one amino acid located in the two basic regions (amino acid residues between 13 and 15, and residues 37 and 38) replaced by alanine resulted in complete loss of nuclear targeting activity of Ahr  as judged by microinjection (Fig. 4A), the effect of mutations on the nuclear transport activity was investigated in vitro. Fig. 4B shows a comparison of the binding activity between the wild type of GST-Ahr(13-39) wt -GFP and GST-Ahr(13-39) mut -GFP, which has the double mutation of R15A/R38A in Ahr , to PTAC58 in the presence of PTAC97. Ahr with double mutations R15A/R38A (Fig. 4B, lane 4) showed drastically reduced binding to the NLS receptor PTAC58 compared with Ahr(13-39) wt (Fig. 4B, lane 3).
Next, we analyzed the nuclear rim targeting activity of GST-Ahr(13-39) wt -GFP or GST-Ahr(13-39) mut -GFP using an in vitro nuclear transport assay (Fig. 4C). We observed clear targeting to the nuclear rim of GST-Ahr(13-39) wt -GFP (Fig. 4C, g) as well as the control substrate of GST-NLSc-GFP (Fig. 4C, f) incubated with purified PTAC58/PTAC97 at 4°C. In contrast, however, there was no accumulation of GST-Ahr(13-39) mut -GFP to the nuclear rim (Fig. 4C, h). These results clearly indicate that the deficiency of nuclear localization of alanine-substituted Ahr in the two bipartite parts of NLS by microinjection (Fig. 4A) may be explained in part by reduced interaction between mutated NLS and NLS receptor PTAC58 resulting in the loss of nuclear rim targeting.
Ahr Contains Both NLS and NES, Suggesting That It Func-tions as a Shuttle Protein-During the course of experiments designed to elucidate the mechanism of nuclear translocation of Ahr, we found that Ahr has in its helix 2 region a short sequence of NES that regulates nuclear export of some proteins. The core sequence of NES, like the NES of PKI, Rev, or MAPKK, is rich in leucine residues, which were found to be crucial for NES activity (Fig. 5A). To test whether the leucinerich sequence in Ahr can act as an NES, the fused protein of GST-GFP containing Ahr(55-75) wt was expressed in BL21 and purified by glutathione-Sepharose. When injected into the nucleus of MDBK cells, the GST-Ahr(55-75) wt -GFP was found to be present in the cytoplasm and excluded from the nucleus almost completely within 30 min (Fig. 5B, a). In contrast, coinjected Texas Red-labeled BSA localized in the nucleus (Fig.  5B, b). The nuclear export of GST-Ahr(55-75) wt -GFP was sensitive to low temperature; it remained in the nucleus at 4°C (Fig. 5B, c). Because the two leucines in the core sequence of NES were shown to be prerequisite for NES activity (15), we introduced substitutions to the corresponding amino acids Leu 70 and Leu 72 for Ala. When injected into the nucleus, the GST-Ahr(55-75) mut -GFP was unable to cross the nuclear membrane and remained in the nucleus (Fig. 5B, e) as well as Texas Red-labeled BSA (Fig. 5B, f). Furthermore, when GST-Ahr(55-75) wt -GFP was injected into the nucleus together with an excess amount of NES peptide of PKI conjugated with BSA ( Fig.  5B, g), the export of GST-Ahr(55-75) wt -GFP from the nucleus was inhibited completely. In contrast, the same amount of BSA alone did not inhibit the nuclear export of GST-Ahr(55-75) wt -GFP to any extent (Fig. 5B, h). Therefore, it is likely that the nuclear export of GST-Ahr(55-75) wt -GFP may be a signal-dependent active transport mediated by NES-binding protein(s) (39,40).

DISCUSSION
Ahr and ARNT belong to the same bHLH/PAS family and have similar modular structures (8,41,42). The bHLH domains are located toward the amino termini of the proteins; the HLH regions mediate dimerization between Ahr and ARNT, whereas the basic regions are involved in DNA recognition by the Ahr⅐ARNT heterodimer. The PAS regions, two repeats homologous with Drosophila Per and Sim proteins, may contribute to Ahr-ARNT dimerization and ligand/hsp90 binding. The carboxyl regions of Ahr contain a transactivation domain that contributes to transcriptional control by the Ahr⅐ARNT com-plex. The heterodimer Ahr⅐ARNT complex recognizes the cisacting DNA sequence termed XREs, which acts upstream of several drug-metabolizing enzymes including the CYP1A1 gene, to induce transcription (5,7,43). The Ahr⅐ARNT system is also thought to mediate the various biological effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin-like environmental pollutants, including carcinogenesis, teratogenesis, tumor promotion, and immunosuppression (44,45). In fact, recent reports showed that disruption of the Ahr or ARNT gene caused impairment of the immune system and liver fibrosis (46) or abnormal angiogenesis (47) in mice, respectively, suggesting that the Ahr⅐ARNT system participates essentially in developmental processes.
Because the first step of these biological effects seems to be caused by signal transduction mediated by activated Ahr, investigation of nuclear translocation of Ahr and its heterodimer partner ARNT should provide useful information on the regulation of gene transcription. We recently showed a clear nuclear localization of ARNT in the absence of ligands to Ahr and identified an NLS of amino acid residues 39 -61 (11). These observations led to the notion that Ahr translocates ligand and NLS dependently from the cytoplasm to the nucleus irrespective of ARNT. We observed ligand-dependent nuclear translocation of full-length Ahr (Fig. 1A), but ligand-independent nuclear translocation activity was observed when the PAS domain was deleted (Fig. 1B). These observations support the notion that the NLS of Ahr is masked by hsp90 molecules in the absence of ligands, resulting in disturbance of recognition by its NLS receptor(s). On the contrary, when ligand is present in the cells, Ahr dissociates from the two molecules of hsp90 resulting in unmasking of the NLS, allowing it to translocate to the nucleus with formation of a nuclear targeting complex. A detailed analysis on the mechanisms of ligand-dependent nuclear translocation is now under investigation.
Ahr contains both NLS  and NES (55-75) in the NH 2 -terminal region as shown in Fig. 6, in which the previously reported DNA binding domain is also shown (48 -51). In the case of ARNT, two basic amino acid clusters are separated by 30 amino acid residues; these participated in different functions with one serving as a variant bipartite core of NLS (39 -61) (11) and the other (91-102) involved in binding to DNA (48,49,51). Because it has been reported that about 80% of the nucleic acid-binding proteins contain overlapping or flanking (less than 10-amino acid separation) NLSs and DNA or RNA binding regions (52), the distal location between NLS and the DNA binding domain in ARNT is a rare case. On the other hand, a typical bipartite NLS of Ahr(13-39) overlaps completely with its DNA binding domain of amino acid residues 10 -44 (48,49). Although ARNT and Ahr belong to the same bHLH/PAS family, the different assignment of the two basic amino acid clusters near the NH 2 terminus may be explained by a different evolutional aspect of the association between NLS and DNA binding motifs in one modular domain (52). Furthermore, it is noteworthy that some basic amino acids (Arg 15 , Lys 37 , His 39 ) and a spacer region (residues 19 -35) between the two basic amino acid clusters are required both for NLS function (Figs. 3A and 4A) and DNA binding activity (48) of the Ahr. The complete loss of nuclear targeting activity of the mutants in basic amino acids correlates with reduced nuclear rim targeting activity as shown in Fig. 4C, thereby confirming the requirement of these basic amino acids for the first step of nuclear localization as shown previously with ARNT (11). In addition, the intervening amino acids may be critical for establishing a precise protein conformation of Ahr not only for the association between NLS and NLS receptor(s), but also for the protein-DNA interaction.
The finding of a functional leucine-rich NES in Ahr (Fig. 5) might provide new insight into the mechanisms of Ahr-mediated gene regulation. It is of interest that the core sequence of the NES (residues 64 -72) was localized in the helix 2 region of Ahr, which is involved essentially in heterodimer formation with ARNT (53). When activated, Ahr translocated to the nucleus to induce transcription by Ahr⅐ARNT⅐XRE complex formation, hence it is likely that the NES may be subject to steric hindrance masking resulting in disturbance of interaction with NES receptor(s). When inducible gene expression of target genes, such as CYP1A1, has proceeded and Ahr dissociates from the Ahr⅐ARNT heterodimer complex, a presumable NES receptor(s) might recognize this common leucine-rich domain of the NES to export Ahr from the nucleus (39,40). That the NES peptide of PKI inhibited the nuclear export of GST-Ahr(55-75) wt -GFP suggests the existence of a common mechanism of nuclear export using a leucine-rich NES as shown in MAPKK (17). Although the biological significance of the NES in Ahr should be elucidated in more detail in the future, it is important to note that no other bHLH/PAS proteins except for Ahr have been found to contain the NES motif at present. To conclude, subcellular distribution of Ahr may be regulated by masking and unmasking of the two different signals of NLS and NES in response to ligands, resulting in nucleocytoplasmic shuttling of the protein.