Proteasome Inhibition Induces Nuclear Translocation and Transcriptional Activation of the Dioxin Receptor in Mouse Embryo Primary Fibroblasts in the Absence of Xenobiotics

Abstract

The aryl hydrocarbon receptor (AHR) is a transcription factor that is highly conserved during evolution and shares important structural features with the Drosophila developmental regulators Sim and Per. Although much is known about the mechanism of AHR activation by xenobiotics, little information is available regarding its activation by endogenous stimuli in the absence of exogenous ligand. In this study, using embryonic primary fibroblasts, we have analyzed the role of proteasome inhibition on AHR transcriptional activation in the absence of xenobiotics. Proteasome inhibition markedly reduced cytosolic AHR without affecting its total cellular content. Cytosolic AHR depletion was the result of receptor translocation into the nuclear compartment, as shown by transient transfection of a green fluorescent protein-tagged AHR and by immunoblot analysis of nuclear extracts. Gel retardation experiments showed that proteasome inhibition induced transcriptionally active AHR-ARNT heterodimers able to bind to a consensus xenobiotic-responsive element. Furthermore, nuclear AHR was transcriptionally active in vivo, as shown by the induction of the endogenous target gene CYP1A2. Synchronized to AHR activation, proteasome inhibition also induced a transient increase in AHR nuclear translocator (ARNT) at the protein and mRNA levels. Since nuclear levels of AHR and ARNT are relevant for AHR transcriptional activation, our data suggest that proteasome inhibition, through a transient increase in ARNT expression, could promote AHR stabilization and accumulation into the nuclear compartment. An elevated content of nuclear AHR could favor AHR-ARNT heterodimers able to bind to xenobiotic-responsive elements and to induce gene transcription in the absence of xenobiotics. Thus, depending on the cellular context, physiologically regulated proteasome activity could participate in the control of endogenous AHR functions.

The aryl hydrocarbon receptor (AHR) is a ligand-activated transcription factor known to mediate most of the toxic effects of a wide variety of environmental contaminants such as dioxin (2,3,7,8-tetrachlorodibenzo-[p]-dioxin [TCCD]). The AHR belongs to the basic helix-loop-helix (bHLH)/PAS (Period [Per]-Aryl hydrocarbon [Ah] receptor nuclear translocator [Arnt]-Single minded [Sim]) family of heterodimeric transcriptional regulators. bHLH/PAS proteins are involved in the control of diverse physiological processes such as circadian rhythms, organ development, neurogenesis, metabolism, and the stress response to hypoxia (reviewed in references 16, 29, 32, and 65). AHR activation is followed by changes in its compartmentalization within the cell. Thus, whereas a large fraction of the unliganded AHR resides in the cytosolic compartment, upon ligand binding the receptor translocates to the nucleus, heterodimerizes with the AHR nuclear translocator (ARNT), and binds to cognate regulatory elements (xenobiotic-responsive elements, XREs) located upstream on the promoter of target genes (CYP1A1, CYP1B1, CYP1A2, and UDP-glucuronosyl-transferase 1∗06) (47, 59, 64). Although the molecular events leading to AHR activation in the presence of xenobiotics are generally well understood, AHR signaling pathways in the absence of exogenous ligands remain essentially unknown. Nevertheless, increasing experimental evidence suggests physiological roles for the AHR during liver development and homeostasis (23, 42, 55), in immune system function (23, 58), in cell proliferation and differentiation (3, 21, 39, 66), and in retinoic acid metabolism (4). Molecular mechanisms for AHR activation in absence of xenobiotics remain elusive, mainly because no endogenous ligands have yet been identified. Among cellular processes that could be involved in physiological activation of the AHR, protein kinase C-mediated phosphorylation appears to be a potential candidate (8, 10). Recent data suggest that changes in the cellular levels of AHR and ARNT could participate in AHR activation since transient overexpression of both proteins in CV-1 and C57 cells leads to nuclear translocation and transcriptional activation of CYP1A1 in the absence of exogenous ligand (9).

Protein degradation through the proteasome constitutes an important cellular mechanism recently identified to regulate gene expression (reviewed in references 14, 15, and 18). This process has been implicated in controlling the half-life of the cell cycle regulators p27^Kip1 and cyclin D1 (7, 56), in the inactivation of transcription factors such as NF-κB, c-Jun, and c-Myc (33, 61), and in the degradation of estrogen receptor α (43). Proteasome inhibition is also relevant for certain cellular responses to stress, such as those taking place through activation of Jun N-terminal kinase (JNK-1), which leads to apoptotic cell death in numerous cell lines (41, 62). Recent reports from different cell systems have addressed the role of the proteasome in degradation of xenobiotic-bound AHR. Thus, it has been shown that xenobiotic-activated AHR is degraded in the cytoplasm after being exported from the nucleus (17); that translocation from the cytosol is enough to trigger AHR degradation, since the role of ligand binding and ARNT dimerization is marginal to the degradation process (53); and that proteolysis requires ubiquitin binding to the AHR (40, 53).

In this study, using mouse embryonic primary fibroblasts (MEF), we have analyzed the role of proteasome inhibition on AHR activation in absence of exogenous ligand. Our results indicate that proteasome inhibition depleted cytosolic AHR and induced receptor translocation to the cell nucleus independently of xenobiotic-mediated activation. Xenobiotic-free, nuclear AHR was able to heterodimerize with ARNT and to bind to a consensus XRE in vitro. Furthermore, nuclear AHR-ARNT heterodimers were transcriptionally active in vivo, as shown by the induction of the endogenous target gene CYP1A2. Whereas proteasome inhibition did not affect the total cellular AHR content, it induced a transient increase in ARNT protein and mRNA levels. These changes in ARNT expression were synchronized with AHR transcriptional activation. We suggest that regulation of ARNT expression by the proteasome could promote AHR translocation and accumulation into the nuclear compartment, resulting in transcriptionally active AHR-ARNT heterodimers able to induce the expression of target genes. Thus, regulated by cell status, physiological control of proteasome activity could be involved in AHR activation, allowing this receptor to participate in endogenous functions independently of xenobiotics.

MATERIALS AND METHODS

Chemicals and reagents.

TCDD and benzo[a]pyrene (BP) were purchased from AccuStandard. N-Carbobenzoxy-Leu-Leu-leucinal (MG132) and N-acetyl-Leu-Leu-norleucinal (LLnL, Calpain inhibitor I, or MG101) were obtained from Sigma Chemical Co. TCDD, BP, MG132 and LLnL were dissolved in sterile dimethyl sulfoxide (DMSO) and filtered before use. Poly(dI-dC)-poly(dI-dC) was from Pharmacia Biotechnology. BioTaq DNA polymerase was from Bioline, and Moloney murine leukemia virus reverse transcriptase was from Stratagene. Cell medium supplements were purchased from Life Technologies, except for fetal bovine serum, which was obtained from Sigma. Rabbit antibody against mouse AHR was generously provided by Richard Pollenz (University of South Carolina). Goat anti-ARNT and anti-mouse immunoglobulin G IgG-coupled-horseradish peroxidase (HRP) were from Santa Cruz Biotechnology. Rabbit antibody against mouse β-actin was obtained from Sigma. Anti-rabbit and anti-goat IgG-HRP were from Pierce and from DAKO, respectively. Rabbit anti-rat CYP1A1, which cross-reacts with mouse CYP1A2, was from Gentest Corp. Monoclonal antibody against CYP1A1 and vaccinia virus-expressed mouse CYP1A1 and CYP1A2 were generously provided by Harry Gelboin and Kristopher Krausz (National Cancer Institute, National Institutes of Health, Bethesda, Md.). The specificity of the anti-AHR, anti-CYP1A2, and anti-CYP1A1 antibodies has been described previously (26, 47).

Mice.

AHR-null and wild-type control mice of the same genetic background (C57BL6/N × 129/sV) were produced as previously described (23). The animals were housed in a germ-free facility in accordance with the Animal Care and Use Guidelines established by the University of Extremadura. Mice were fed with water and rodent chow ad libitum. The genotype of each mice was determined by restriction fragment length polymorphism of tail DNA as described previously (23).

Cell culture.

MEF were isolated from 14.5-day-postcoitum embryos (3). Briefly, embryos were separated from maternal tissues and yolk sac and the internal organs were removed. Carcasses were finely minced and incubated with gentle agitation at 37°C for 1 h in phosphate-buffered saline (PBS) containing 0.25% (wt/vol) trypsin-EDTA. Trysin was inactivated, and the cell suspension was briefly centrifuged to remove undigested tissue. Supernatant containing MEF was plated in Dulbecco's modified Eagle's medium supplemented with 10% FBS, 2 mM l-glutamine, 100 U of penicillin per ml, and 100 μg of streptomycin per ml. MEF were grown to confluence in tissue culture flasks at 37°C in a 5% CO₂ atmosphere (considered here to be passage 0). MEF from the first or second passage at around 80% confluence were used in all the experiments. Swiss 3T3 fibroblasts were propagated from frozen stocks in the same Dulbecco's modified Eagle's medium as mentioned above.

Construction of AHR-EGFP fusion protein and transient transfections.

Full-length AHR cDNA was isolated from mouse liver total RNA by reverse transcription-PCR (RT-PCR) using the primers described by Ma et al. (38), which hybridize to the 5′end (nucleotides 915 to 938) and 3′end (nucleotides 3306 to 3332) of the cDNA characterized from Hepa-1 cells (22). To make the AHR-enhanced green fluorescent protein (EGFP) fusion construct, mouse AHR cDNA with primer-added XbaI-HindIII flanking cloning sites was made blunt ended and ligated in frame into the blunt-ended HindIII site of pEGFP-C1 (Clontech Laboratories). The construct was subjected to endonuclease restriction analysis to confirm the location and orientation of the AHR cDNA. Transient transfections were performed with the fibroblast cell line Swiss 3T3 or with AHR-null MEF cultures using the LipofectAMINE Plus Reagent (Life Technologies) and 1 μg of either pEGFP or AHR-pEGFP DNA construct. Protein expression was allowed to continue for 24 h after transfection, and cultures were treated for up to 12 h with 8 μM MG132, 50 μM LLnL, or vehicle alone (DMSO). For xenobiotic-induced nuclear translocation, transfected cells were treated with 10 μM BP for 90 min. After treatment, the plates were washed with PBS and photographed under a Zeiss fluorescence microscope.

Preparation of cell lysates and immunoblot analysis.

Following treatment, cell monolayers were washed with PBS and scraped from the plates at 4°C into lysis buffer (50 mM Tris-HC1 [pH 7.5], 2 mM EDTA, 2 mM EGTA, 10 mM β-glycerophosphate, 5 mM sodium pyrophosphate, 50 mM sodium fluoride, 0.1 mM sodium orthovanadate, 1% Triton X-100, and Complete protease inhibitor cocktail [Boehringer-Mannheim]). Cell suspensions were subjected to two freezing-thawing cycles to obtain whole-cell extracts. Aliquots of whole-cell extracts were centrifuged at 14,000 × g for 10 min at 4°C, and the resulting supernatants were saved as cytosolic fractions. Nuclear extracts were obtained as indicated below. For the analysis of protein expression by immunoblotting, 10-μg portions of the corresponding cellular extracts were denatured, separated in sodium dodecyl sulfate (SDS)–8% polyacrylamide gels, and transferred to nitrocellulose membranes. The membranes were blocked at room temperature for 2 h in TBS-T (50 mM Tris-HC1 [pH 7.5], 10 mM NaCl, 0.5% Tween 20) containing 5% (wt/vol) nonfat milk and incubated with the corresponding primary antibodies for 2 h at room temperature or overnight at 4°C. After extensive washing in TBS-T, blots were incubated with the corresponding HRP-coupled secondary antibody. Following additional washing in TBS-T, the membranes were incubated with the SuperSignal chemiluminescence substrate (Pierce) and exposed to a chemiluminescence imaging screen. The screen was scanned using a Molecular Imager FX system (Bio-Rad). When required, signals were quantitated by volumetric integration of the raw data using Quantity One software (Bio-Rad) as specified by the manufacturer.

RT-PCR and Northern analysis.

Subconfluent MEF cultures were treated with 8 μM MG132 for 0, 3, 6, or 12 h, and total cellular RNA was extracted using the guanidinium thiocyanate method (12). A 7-μg portion of total RNA was reverse transcribed at 42°C for 60 min using oligo (dT) priming and Moloney murine leukemia virus reverse transcriptase. PCR amplification for the AHR cDNA was performed using the primers AhrF (forward) (5′ ATGAGCAGCGGCGCCAACATCACC 3′) and AhrR (reverse) (5′ AACATCAAAGAAGCTCTTGGCCCTCAG 3′); ARNT cDNA was amplified with the primers ArntF (forward) (5′ CCAGATGTGTAATGACAAGGAGCGG 3′) and ArntR (reverse) (5′ ATCGGAACATGACGGACAGCACCTG 3′). To check for RNA integrity and quantitation, β-actin cDNA was also amplified using the primers ActinF (forward) (5′ GGTCAGAAGGACTCCTATGTGG 3′) and ActinR (reverse) (5′ TGTCGTCCCAGTTGGTAACA 3′). Amplification was carried out for 40 cycles in a 50-μl reaction mixture containing 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.5 mM MgCl₂, 0.2 mM each deoxynucleoside triphosphate, 0.5 μM each primer, 2.5 U of Taq polymerase, and an aliquot of each reverse transcription reaction mixture as template. Cycling conditions were as follows: denaturation at 94°C for 1 min, annealing at 58°C for 1 min (60°C for Ahr), and extension at 72°C for 2 min (1 min for Ahr). PCR products were visualized in agarose gels stained with ethidium bromide.

Northern analysis was performed essentially as described previously (23). Briefly, total RNA was isolated by the guanidinium thiocyanate method (12) and 10 μg from each treatment was separated in 6% formaldehyde–1% agarose gels. The gels were transferred to Gene-Screen Plus nylon membranes, RNA was fixed by UV cross-linking, and blots were prehybridized at 65°C for 3 h in Rapid-Hyb buffer (Amersham). Mouse Cyp1a2 and β-actin cDNA probes were labeled by random priming using [³²P]dCTP and Klenow DNA polymerase. cDNA probes were added at 8 × 10⁵ cpm/ml in Rapid-Hyb buffer, and incubation was continued for 2 h at 65°C. Background was reduced by sequential washing in 2× SSC (3 M sodium chloride, 0.3 M sodium citrate [pH 7.5])–0.1% (wt/vol) SDS at room temperature for 30 min and in 0.1× SSC–0.1% (wt/vol) SDS at 65°C for two changes of 25 min each. Membranes were exposed to Kodak screens, which were scanned using a Molecular Imager FX imaging system.

Nuclear extracts and EMSA.

Electrophoretic mobility shift assay (EMSA) was performed on nuclear extracts isolated from MEF cultures. Following treatment, plates were washed with PBS and monolayers were scraped into ice-cold 10 mM EDTA (pH 8.0). The resulting cell suspensions were centrifuged at 400 × g for 5 min at 4°C and the cell pellets were resuspended in hypotonic MDH buffer (25 mM HEPES [pH 7.9], 3 mM MgCl₂, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride). After swelling on ice for 10 min, the cells were homogenized at 4°C with 25 strokes of a loose-fitting pestle Dounce homogenizer. Nonidet P-40 was added at 0.5% (vol/vol), and homogenization was continued for five more strokes as indicated above. Lysed MEF were then centrifuged at 15,000 × g for 30 s at 4°C. Pellets containing crude nuclei were carefully washed in MDH buffer, centrifuged, and resuspended in low-ion-strength HDEK buffer (25 mM HEPES [pH 7.9], 2 mM EDTA, 1 mM dithiothreitol, 0.1 M KCl). To extract nuclear AHR, the KCl concentration was increased to 0.4 M, glycerol was added to 10% (wt/vol), and the suspension was incubated for 30 min at room temperature with gentle rotation. Extracted nuclei were centrifuged at 15,000 × g for 60 min at 4°C, and the supernatant (nuclear fraction) was aliquoted and frozen at −80°C. The protein concentration was determined as indicated below. For EMSA, complementary oligonucleotides containing the double-stranded sequence for the XRE3 binding element (underlined (20) 5′ GATCGCGCTCTTCTCACGCAACTCCGAGCTCA 3′ and 5′ GATCTGAGCTCGGAGTTGCGTGAGAAGAGCCG 3′ were synthesized, annealed, and labeled at their 5′ ends using T4 polynucleotide kinase and [γ-³²P]ATP. Binding reactions were performed in 20 μl of binding buffer (20 mM HEPES [pH 7.9], 1 mM EDTA, 1 mM dithiothreitol, 10% [wt/vol] glycerol) containing 7 μg (for TCDD-treated cultures) or 10 μg (for MG132-treated cultures) of nuclear protein and 2.5 μg of nonspecific competitor poly(dI-dC)-poly(dI-dC). After a 20-min incubation at room temperature, 25,000 cpm of ³²P-labeled XRE3 was added and the incubation was continued at room temperature for an additional 20 min. DNA-protein complexes were resolved at room temperature in nondenaturing 6% polyacrylamide gels, using 1× TBE (89 mM Tris-borate, 89 mM boric acid, 2 mM EDTA) as running buffer. The gels were dried, exposed to Kodak screens, and scanned using a Molecular Imager FX imaging system. For competition experiments, extracts were preincubated at room temperature for 20 min with a 100-fold molar excess of unlabeled XRE3 or 2 μg of either anti-AHR or anti-ARNT antibodies before the addition of the labeled XRE3. Preimmune IgG was also preincubated with nuclear extracts to confirm the specificity of the inhibitory reaction by the anti-AHR and anti-ARNT antibodies.

Protein concentration and statistical analysis.

The protein concentration was determined by using the Coomassie Plus protein assay reagent (Pierce) and bovine serum albumin as standard. Statistical analyses were done using analysis of variance on the Instat software program (GraphPAD software, version 1.11). Data shown are mean ± standard deviation.

RESULTS

Proteasome inhibition induces AHR nuclear translocation in the absence of xenobiotics in MEF.

The addition of high-affinity exogenous ligands (10 nM TCDD or 10 μM BP) to MEF cultures induced AHR depletion from the cytosolic compartment to almost undetectable levels by 12 h (Fig. 1, upper panel). This effect was probably due to proteolysis of the receptor since AHR depletion was also evident in total-cell lysates from the same cultures (Fig. 1, lower panel). Thus, in agreement with results obtained with different cell lines (17, 40, 53), xenobiotics were able to induce ligand-dependent depletion of cellular AHR levels in MEF cultures. MEF treated with proteasome inhibitors (8 μM MG132 or 50 μM LLnL), in the absence of xenobiotics, also showed a significant decrease in cytosolic AHR content (Fig. 1, upper panel). However, in contrast to the results obtained with xenobiotic-treated cultures, proteasome inhibition did not significantly affect total cellular AHR levels (Fig 1, lower panel). Therefore, cytosolic AHR depletion by proteasome inhibitors could be due to receptor translocation from the cytosolic to the nuclear compartment of the cell. In an attempt to further define the cellular distribution of AHR during proteasome inhibition, we carried out transient-transfection experiments using the mouse fibroblast cell line Swiss 3T3 and the AHR-EGFP fusion protein. To test whether this construction was responsive to nuclear translocation, we analyzed the subcellular distribution of AHR-EGFP after addition of xenobiotics. As expected, addition of 10 μM BP to AHR-EGFP-transfected fibroblasts induced a marked nuclear localization of the AHR (Fig. 2F), indicating that the fusion protein was behaving as the wild-type receptor. Control experiments performed with Swiss 3T3 cells transfected with EGFP control construct showed a homogeneous cellular distribution of fluorescence that was not affected by treatment with xenobiotics or proteasome inhibitors (Fig. 2A to D). Swiss 3T3 cells transfected with the AHR-EGFP fusion protein and treated with 8 μM MG132 (Fig. 2G) or 50 μM LLnL (Fig. 2H) showed AHR-EGFP nuclear translocation that was maintained for at least 12 h after treatment. Considering that proteasome inhibition induced cytosolic AHR depletion in MEF cultures (Fig. 1) and nuclear localization in transfected Swiss 3T3 fibroblasts (Fig. 2), our data indicated that proteasome inhibition probably depleted cytosolic AHR by promoting receptor translocation to the nucleus. To analyze if the AHR-EGFP fusion protein was not only responsive to nuclear translocation but also able to activate gene expression, we performed transient transfections with MEF cultures isolated from AHR-null embryos and measured induction of the endogenous target gene CYP1A2 by immunoblotting (Fig. 3). Whereas basal CYP1A2 expression was undetectable in nontransfected AHR-null MEF (lane 1), it was readily induced in AHR-EGFP-transfected cultures treated for 6 h with 10 nM TCDD (lane 2). Transfected MEF, in the absence of TCDD treatment, did not express any CYP1A2 protein (lane 3). Since TCDD treatment increased CYP1A2 mRNA levels in wild-type MEF (see Fig. 7), these data suggest that AHR-EGFP-dependent CYP1A2 protein induction could be the result of transcriptional activation of the transfected receptor. However, we were unable to detect any constitutive or MG132-induced CYP1A2 expression in Swiss 3T3 fibroblasts (data not shown), suggesting that responsiveness to proteasome inhibition could be mainly associated with embryo-derived primary fibroblasts. Additional data supporting the ability of proteasome inhibition to induce AHR translocation were obtained by analyzing receptor levels in nuclear extracts from MEF cultures treated with 8 μM MG132 in the absence of xenobiotics. As shown in Fig. 4A, although total-cell AHR levels remained essentially constant with time in the presence of MG132, at 12 h after proteasome inhibition, a consistent slight decrease in cellular AHR was observed that could be due to protein degradation once transcriptional activation of target genes had occurred. This degradation process could be similar to that suggested for ligand-induced AHR proteolysis (17, 40, 53). In the nuclear compartment, the AHR content transiently increased with time, reaching a maximum at about 6 h after proteasome inhibition (Fig. 4B). As expected, treatment with a high-affinity exogenous ligand (10 nM TCDD) for 90 min induced high levels of nuclear AHR in wild-type (expressing a functional AHR) but not in AHR-null MEF cultures. Taken together, these results suggest that proteasome inhibition could contribute to AHR translocation and accumulation in the nucleus without the requirement for xenobiotic binding.

FIG. 1.

AHR is depleted from the cytosol of proteasome inhibitor-treated MEF. Cells were plated and treated with the indicated chemicals for 12 h. AHR expression was analyzed by immunoblotting with anti-mouse AHR antibody using 10 μg of cytosolic or total-cell extracts. (Upper panel) Immunoblot analysis for AHR levels in cytosolic fractions from MEF treated with solvent (DMSO), 10 μM BP, 10 nM TCDD, 8 μM MG132, or 50 μM LLnL. (Lower panel) Immunoblot analysis for AHR levels in the corresponding total-cell extracts. Experiments were performed in duplicate in at least three different MEF preparations.

FIG. 2.

Proteasome inhibitors induce AHR-EGFP nuclear translocation in the absence of xenobiotics. Swiss 3T3 fibroblasts were grown and transiently transfected with the EGFP control plasmid (A to D) or the AHR-EGFP fusion protein (panels E to H). Following a 24-h incubation to allow for protein expression, cells were treated for an additional 6 h with DMSO (A and E), 8 μM MG132 (C and G), or 50 μM LLnL (D and H) or for 90 min with 10 μM BP (B and F). Micrographs are representative of transfections performed in at least six different Swiss 3T3 cultures.

FIG. 3.

The AHR-EGFP fusion protein is transcriptionally active. MEF cultures isolated from AHR-null mouse embryos were transfected with the AHR-EGFP fusion construct as indicated in Materials and Methods. Reporter gene expression was analyzed following CYP1A2 induction by immunoblotting using 10 μg of protein. AHR-null MEF cultures were transfected and treated for 6 h with DMSO (lane 3) or 10 nM TCDD (lane 2). Nontransfected AHR-null MEF cultures were used as a negative control (lane 1). Vacciniavirus-expressed mouse CYP1A2 was included as a positive control (lane 4). Expression of mouse β-actin was included as a control for protein loading. The experiment was done in duplicate with two different AHR-null MEF preparations.

FIG. 7.

CYP1A2 induction by proteasome inhibition is due to AHR-dependent transcriptional activation. Wild-type and AHR-null MEF cultures were treated with 8 μM MG132 for the indicated times or with 10 nM TCDD for 6 h, and CYP1A2 mRNA expression was measured by Northern blot analysis as indicated in Materials and Methods. mCYP1A2 mRNA expression steadily increased with time by proteasome inhibition or TCDD treatment in control (Ahr+/+) but not in AHR-null (Ahr−/−) MEF cultures. Note that although AHR-null cells were not responsive for CYP1A2 induction by any of the treatments, their basal CYP1A2 mRNA levels (0 h) appeared to be higher than for wild-type (Ahr+/+) MEF. Expression of mouse β-actin was used as a control for RNA integrity and loading. The experiment was repeated twice with two different MEF preparations.

FIG. 4.

Proteasome inhibition induces AHR accumulation in the nucleus of MEF in absence of exogenous ligand. MEF were plated and treated with 8 μM MG132 for the indicated times. Nuclear and total-cell extracts were analyzed by immunoblotting using 10 μg of protein and an anti-mouse AHR antibody. (A) Total-cell extracts from cultures treated with MG132 for 0, 3, 6, or 12 h. (B) Nuclear extracts from MEF cultures treated as for panel A. AHR levels in nuclear extracts from wild-type and AHR-null MEF, both treated with 10 nM TCDD for 90 min, are included as positive and negative controls, respectively. Note the presence of basal levels of nuclear AHR in the absence of any treatment in wild-type MEF cultures (lane 0 h) but not in AHR-null MEF (lane Ahr−/−). The experiment was performed in duplicate with two different MEF preparations.

Proteasome inhibition-mediated transcriptional activation of nuclear AHR.

To analyze if proteasome inhibition-mediated nuclear translocation was followed by the formation of active AHR-ARNT heterodimers able to activate transcription in MEF cultures, we performed in vitro binding experiments by EMSA and in vivo immunoblot analysis for the expression of target genes (e.g., CYP1A2). The results obtained with EMSA using nuclear extracts are shown in Fig. 5. MEF cultures treated with 8 μM MG132 showed a time-dependent increase in the formation of specific AHR-ARNT nuclear complexes (arrow) able to bind to the consensus XRE3 (Fig. 5, lanes 2 to 5). The specificity of MG132-induced AHR-ARNT heterodimers hybridizing to XRE3 was confirmed by competition experiments in the presence of 100-fold molar excess of unlabeled XRE3 (lane 8) or by the addition of either anti-AHR (lane 9) or anti-ARNT (lane 10) antibodies. Additional positive and negative controls were performed by analyzing nuclear extracts from wild-type (lanes 6, 11, and 12) or AHR-null (lane 7) MEF cultures, both treated with 10 nM TCDD for 90 min. Preincubation of nuclear extracts with preimmune IgG did not affect the specific binding of AHR-ARNT complexes to the XRE3 consensus element (results not shown).

FIG. 5.

Proteasome inhibition induces AHR-ARNT binding to a consensus XRE3 element in vitro. Nuclear extracts from MEF cultures treated with 8 μM MG132 for 0 h (lane 2), 3 h (lane 3), 6 h (lane 4), or 12 h (lane 5) were prepared and analyzed by EMSA using the ³²P-labeled XRE3. Nuclear extracts from wild-type (lane 6) and AHR-null (lane 7) MEF cultures, both treated with 10 nM TCDD for 90 min, were used as positive and negative controls, respectively. The specificity of the shifted band (arrow) was confirmed by preincubating nuclear extracts from MEF cultures treated for 6 h with 8 μM MG132 or 90 min with 10 nM TCDD with a 100-fold molar excess of unlabeled XRE3 (lanes 8 and 11, respectively) or with 2 μg of anti-AHR antibody (lanes 9 and 12, respectively). Nuclear extracts from MEF cultures treated for 6 h with 8 μM MG132 were also incubated with 2 μg of anti-ARNT antibody (lane 10). Lane 1 contains a binding-reaction mixture in the absence of nuclear extract; 10 and 7 μg of nuclear protein were used for MG132- and TCDD-treated MEF cultures, respectively. The position of the specific AHR-ARNT-XRE3 complex is indicated by the arrow at the top of the gel. The position of the free probe is shown at the bottom of the gel. Preincubation of nuclear extracts with preimmune IgG did not affect specific AHR-ARNT binding to XRE3. EMSA was performed in duplicate with nuclear extracts from three different MEF preparations.

Although nuclear extracts from proteasome inhibitor-treated MEF cultures contained transcriptionally active AHR able to bind to the consensus XRE3 sequence in vitro, we also investigated if induction of target genes could take place in vivo under these experimental conditions. As shown in Fig. 6A, the constitutive level of CYP1A2 protein was undetectable by immunoblotting in MEF cultures (DMSO treated, lane 1). On xenobiotic treatment (10 nM TCDD or 10 μM BP), a marked induction of CYP1A2 protein production was observed (lanes 2 and 3). As an AHR-regulated gene, xenobiotic-dependent CYP1A2 induction was completely lost in AHR-null MEF cultures (lanes 5 and 6). In agreement with previous results obtained in fibroblast-derived cell lines (2, 13, 30), we were unable to detect any constitutive or xenobiotic-induced expression of the closely related CYP1A1 gene in MEF cultures. Proteasome inhibition by 8 μM MG132 induced CYP1A2 protein in wild-type MEF in a time-dependent manner, reaching a maximum level of steady-state protein expression by 12 h (Fig. 6B). The level of CYP1A2 induction after 12 h of MG132 treatment was comparable to that obtained by treatment with 10 nM TCDD for 6 h. To eliminate the possibility that MG132 could be stabilizing basal CYP1A2 protein rather than activating AHR-mediated transcription, treatments were also performed in AHR-null MEF cultures (Fig. 6C). It can be observed that MG132 was unable to induce CYP1A2 in MEF lacking a functional AHR, suggesting that proteasome inhibitor-mediated gene induction was regulated through the AHR. To further characterize the involvement of AHR in proteasome inhibitor-mediated gene induction, we analyzed CYP1A2 mRNA levels in wild-type and AHR-null MEF cultures after MG132 treatment (Fig. 7). The proteasome inhibitor increased CYP1A2 mRNA production in wild-type MEF from constitutive (0-h) to induced (12-h) levels. Xenobiotic treatment (10 nM TCDD for 6 h) also increased CYP1A2 mRNA expression to levels comparable to those obtained after 12 h of proteasome inhibition. AHR-null MEF, in contrast, did not support CYP1A2 mRNA induction and CYP1A2 mRNA remained at its constitutive level regardless of proteasome inhibitor or xenobiotic treatment. Interestingly, basal levels of CYP1A2 mRNA were higher in AHR-null than in wild type MEF, suggesting that the AHR could be involved in maintaining its endogenous transcription rate in embryo-derived fibroblasts. Taken together, these data indicate that proteasome inhibition in MEF cultures, in the absence of xenobiotics, localizes the AHR to the nucleus and activates the formation of active AHR-ARNT heterodimers able to bind to regulatory XRE sequences in vitro. Further, MG132-mediated AHR activation resulted in CYP1A2 induction at the mRNA level in wild-type but not in AHR-null MEF cells.

FIG. 6.

CYP1A2 protein induction by xenobiotics and proteasome inhibitors in MEF cultures. (A) MEF cultures from wild-type (lanes 1 to 3) or AHR-null (lanes 4 to 6) mice were treated with the indicated AHR ligands, and total-cell extracts were analyzed by immunoblotting using 10 μg of protein and an anti-CYP1A2 antibody. Total-cell extracts from cultures treated for 12 h with DMSO (lanes 1 and 4), 10 μM BP (lanes 2 and 5), or 10 nM TCDD (lanes 3 and 6) were used. (B) Wild-type MEF cultures were treated with 8 μM MG132 for the indicated times or with 10 nM TCDD for 6 h, and CYP1A2 expression was analyzed by immunoblotting. (C) To determine if CYP1A2 induction was an AHR-dependent process, AHR-null MEF cultures were treated with 8 μM MG132 for the indicated times or with 10 nM TCDD for 6 h and CYP1A2 expression was analyzed by immunoblotting. Expression of mouse β-actin was used as a control for protein loading. Vaccinia-expressed mouse CYP1A2 (r1A2) was used as positive control (lane 7). Experiments were done in duplicate with three different MEF preparations.

Proteasome inhibition-mediated AHR transcriptional activation could be a consequence of increased ARNT expression.

Since transcriptionally active AHR was found in MEF cultures in the absence of xenobiotics, an attempt was made to identify potential mechanisms involved in this process. ARNT is a bHLH/PAS protein and is a required partner for AHR-dependent transcriptional activation (reviewed in references 32, 44, 59, 64, and 65). The results of immunoblot analysis for AHR and ARNT during proteasome inhibition are shown in Fig. 8. Protein levels for AHR and ARNT were quantified by volumetric integration of the raw data from Western blots and normalized to mouse β-actin expression. Whereas the cellular AHR protein level was not significantly altered by proteasome inhibition (Fig. 8A), ARNT protein levels changed over time, increasing to close to twofold after 6 h of MG132 treatment (Fig. 8). Because ARNT is maintained in the nuclear compartment of the cell at relatively constant levels in the presence or absence of xenobiotics (40, 53), proteasome inhibitors could be altering ARNT protein levels by either increasing its endogenous rate of transcription or inhibiting its rate of degradation. To discriminate between these two possibilities, we analyzed ARNT expression at the mRNA level by RT-PCR in MEF cultures treated with 8 μM MG132 (Fig. 9). It could be observed that the ARNT mRNA level transiently increased with time, reaching about twofold induction by 3 h and fourfold induction by 6 h of proteasome inhibition (top panel). Low but detectable levels of ARNT mRNA were found constitutively (0 h) and 12 h after proteasome inhibition. The AHR mRNA level, however, did not significantly change during proteasome inhibition (middle panel). Taken together, these data suggest that ARNT expression in MEF cultures could be regulated at the transcriptional level by proteasome inhibition. Elevated rates of ARNT transcription could induce a transient increase in the levels of nuclear protein able to induce translocation and accumulation of AHR in the nuclear compartment. Stabilized nuclear AHR could potentiate the formation of AHR-ARNT heterodimers responsible for transcriptional activation of target genes.

FIG. 8.

The proteasome inhibitor induces ARNT protein expression. Wild-type MEF cultures were treated with 8 μM MG132 for the indicated times. ARNT and AHR expression were analyzed by immunoblotting using the corresponding primary antibodies. (A) Representative immunoblots obtained for each of the proteins. (B) Quantitative analysis by volumetric integration of the raw data from the immunoblots shown in panel A. Data were normalized by the levels of mouse β-actin expression. The experiment was performed in duplicate with three different MEF preparations.

FIG. 9.

The proteasome inhibitor induces ARNT transcription. Wild-type MEF cultures were treated with 8 μM MG132 for the indicated times. ARNT and AHR mRNA levels were analyzed by RT-PCR using 7 μg of total RNA and oligo (dT) priming. The specific oligonucleotides for PCR amplification of each gene are indicated in Materials and Methods. The control lane corresponds to an RT-PCR reaction performed in the absence of RNA template. Mouse β-actin mRNA expression was analyzed to verify total RNA integrity and concentration. Note that low but detectable levels of ARNT mRNA could be detected constitutively (0 h) and 12 h after proteasome inhibition. ARNT levels increased by about twofold at 3 h of MG132 treatment and by close to fourfold after 6 h of proteasome inhibition. The experiment was done in duplicate with three different MEF preparations.

DISCUSSION

Many cellular processes are regulated by a transcriptional machinery that recruits the appropriate transcription factors to the nucleus, where they will regulate the expression of specific target genes. Increasing evidence suggests that intracellular compartmentalization and localization could play a role in controlling transcriptional activity by allowing proper protein-protein and protein-DNA interactions and by maintaining adequate levels of transcription factors (5, 48, 49). Proteolysis mediated by the proteasome has emerged as an important mechanism regulating the function of proteins involved in many physiological processes (14, 15, 18). Furthermore, altered proteasome activity has been linked to cancer and to immune, inflammatory, and neurodegenerative diseases in humans (reviewed in reference 14). Most of the molecular information concerning AHR-mediated gene expression has been accumulated by the use of high-affinity exogenous ligands such as dioxin. Although increasing experimental data suggest that the AHR plays physiological and homeostatic functions independently of xenobiotics (3, 4, 21, 23, 39, 42, 55, 58, 66), little information is available regarding potential mechanisms of activation in the absence of exogenous ligand. In this work, we present experimental data suggesting that a proteolytic mechanism could be involved in the regulation of AHR activity in the absence of xenobiotics. Proteasome inhibition in embryonic primary fibroblasts was able to trigger AHR translocation into the nucleus and transcriptional activation of target genes without xenobiotic interaction. Specifically, we have made the following observations: (i) the proteasome inhibitors MG132 and LLnL induced AHR depletion from the cytosol of MEF cells without altering its total cellular level (Fig. 1). (ii) AHR-EGFP fusion protein was translocated to the nucleus in transiently transfected Swiss 3T3 cells after addition of proteasome inhibitors (Fig. 2)—this fusion protein was able to induce target gene expression in transfected AHR-null MEF lacking endogenous AHR (Fig. 3), and, further supporting a translocation process, proteasome inhibition induced a time-dependent accumulation of the AHR in the cell nucleus (Fig. 4); (iii) nuclear AHR was able to form functional AHR-ARNT heterodimers in vitro (Fig. 5) and to activate CYP1A2 expression in vivo (Fig. 6); (iv) the induction of endogenous target gene expression by proteasome inhibitor was observed at the protein and mRNA levels in wild-type but not AHR-null MEF (Fig. 6 and 7); and (v) proteasome inhibitor-dependent AHR activation, in the absence of xenobiotics, could be regulated through a transient increase in ARNT expression (Fig. 8 and 9). From these results, we suggest that physiological control of proteasome activity could play an important role in AHR-mediated transcriptional activation in the absence of exogenous ligand.

Recent studies of liver-derived (mouse Hepa 1c1c7 and human HepG2), Chinese hamster ovary (E36), and rat smooth muscle (A7) cell lines showed no significant activation of AHR by proteasome inhibitors in absence of xenobiotics (17, 40, 53). This apparent absence of activation could be related to differences in AHR signaling between immortalized cell systems and embryonic primary fibroblasts. In this regard, it has been extensively shown that AHR activation and AHR-mediated signal transduction vary among species (46, 59), between strains of the same species (44, 59, 65), and between different tissues (36, 46). Moreover, a marked difference in AHR activation and CYP1A1 induction has been observed among cell lines from diverse tissue origins (6, 24, 59). In agreement to these reports, we were unable to detect induction of target gene expression by proteasome inhibitors in the fibroblast cell line Swiss 3T3. Thus, the available data suggest that AHR activation by proteasome inhibition could be cell type specific, reported to date only in mouse embryo fibroblasts. The relevance of this regulatory mechanism in AHR-mediated processes during mouse development deserves further investigation. Nevertheless, although our data indicate that embryonic MEF were responsible for AHR translocation by proteasome inhibitors, we cannot exclude the possibility that additional factors such as potential endogenous ligands or phosphorylation-related processes (8, 10) could also participate in AHR activation. In support of such potential mechanisms, we have found that untreated MEF presented constitutively low although detectable amounts of nuclear AHR (Fig. 4, compare lanes 1 and 6) in a transcriptionally active conformation (Fig. 5, lane 2). Thus, the AHR could be involved in endogenous functions unrelated to P450 induction and xenobiotic binding. In agreement with this hypothesis, recent studies have reported the existence of nuclear AHR in the absence of exogenous ligand (9, 57) and physical interactions between AHR and the retinoblastoma protein independently of xenobiotics (25, 50).

Since nuclear localization does not imply a transcriptionally active AHR, we have used changes in CYP1A2 expression as an endogenous reporter for AHR-mediated transcription. CYP1A2 protein increased in wild-type but not AHR-null MEF following proteasome inhibition, indicating that target gene induction was AHR dependent (Fig. 6). Interestingly, proteasome inhibition or xenobiotic treatment induced comparable levels of nuclear AHR (Fig. 4) and CYP1A2 induction (Fig. 6) in MEF. However, whereas TCDD induced a maximum level of CYP1A2 expression within 3 to 6 h of treatment, proteasome inhibition required at least 12 h to reach similar levels of protein induction. Thus, xenobiotic-dependent activation may take place through a more direct process in which binding of saturating concentrations of high-affinity exogenous ligand to the AHR constitutes a stimulus potent enough to rapidly induce receptor translocation to the nuclear compartment. Proteasome inhibition, however, could represent a delayed process involving the induction of a transient increase in ARNT expression preceding nuclear accumulation of AHR.

Proteasome-dependent AHR activation was monitored following the induction of the endogenous target gene CYP1A2. Two experimental observations suggested that CYP1A2 induction by MG132 was due to an AHR-dependent transcriptional mechanism rather than to stabilization of basal protein content: (i) AHR-null MEF were unresponsive to proteasome inhibition, and (ii) the increase in target gene expression could be observed at the mRNA level. Interestingly, a higher level of constitutive CYP1A2 mRNA was found in AHR-null than in wild-type MEF, suggesting that the AHR could be involved in posttranscriptional regulation of CYP1A2 in embryonic fibroblasts. Additionally, the absence of inducible CYP1A2 protein in AHR-null MEF, even in the presence of mRNA expression, could involve the AHR, through an unknown mechanism, in the control of protein synthesis. Although several studies have reported the existence of posttranscriptional mechanisms regulating CYP1A2 expression (1, 19, 27, 28, 45, 52), further investigation are be needed to clarify the molecular mechanisms involving the AHR in such processes.

To investigate potential mechanisms participating in AHR activation by MG132, we analyzed the effect of proteasome inhibition on ARNT expression. ARNT is a common partner for transcription factors other than AHR (60). Specifically, hypoxia-inducible factor 1α (HIF-1α) is induced and heterodimerizes with ARNT under low-oxygen conditions and is rapidly degraded by the proteasome when conditions of normoxia return (35, 54). It was shown that HIF-1α–ARNT heterodimerization during hypoxia increased total cellular ARNT levels and stabilized HIF-1α in the nucleus (11). In contrast, other investigators had reported no significant change in ARNT expression during TCDD-induced AHR depletion (40, 53) or during normoxia-induced HIF-1α degradation (34, 51). In this study, we found a transient, time-dependent increase in ARNT protein expression (Fig. 8) that correlated with the pattern for ARNT mRNA expression (Fig. 9). These changes in ARNT levels were coincident in time with AHR nuclear translocation and accumulation (Fig. 4), transcriptional activation (Fig. 5), and CYP1A2 induction (Figs. 6 and 7). Thus, proteasome inhibitors could be acting on AHR signaling by transiently increasing nuclear ARNT levels, which, in turn, could constitute a stimulus able to promote translocation of the AHR from the cytosol to the nucleus. Once the concentration of AHR in the nucleus increased, there could be a shift in the equilibrium between monomeric and heterodimeric AHR-ARNT complexes (17) toward the stable, heterodimeric form of the receptor able to bind the promoter of target genes and to increase transcription. Therefore, a major effect of proteasome inhibitors on AHR activation, in the absence of xenobiotics, could be to increase the time of residence of AHR in the cell nucleus, favoring AHR-ARNT interaction and DNA binding. In agreement, it has been suggested that ARNT binding to AHR increases the stability of the receptor in the nuclear compartment, contributing to the maintenance of adequate levels of transcription (11, 16, 29, 32, 59, 65). Additional data supporting the role of AHR-ARNT levels in transcriptional activity have been obtained after overexpression of both proteins in CV-1 and C57 cells, which resulted in AHR-ARNT nuclear translocation and induction of CYP1A1 in the absence of xenobiotics (9). Moreover, proteasome inhibition and a thiol-reducing agent were enough to translocate HIF-1α from the cytosol during normoxia, indicating that escape from proteolysis was sufficient for HIF-1α accumulation in the cell nucleus (11). Although nuclear translocation is an essential step in AHR signaling, ligand binding seems to play important roles in receptor activation, translocation, and heterodimerization (37). In this context, our results do not exclude the possibility that once increased ARNT levels had established the signal to internalize the receptor to the nucleus, additional factors (e.g., unidentified endogenous ligands or phosphorylation-dependent mechanisms) could contribute to AHR translocation in the absence of xenobiotics. The molecular level at which proteasome inhibition could increase ARNT transcription is unknown. As a hypothesis, it could involve inhibition of proteolysis of constitutive factors responsible for ARNT transcription such as Sp1, Sp3, and CAAT box binding factor A (63). In particular, Sp1 degradation could be relevant to this process since its cellular level has been shown to be regulated through the proteasome (31).

In summary, AHR nuclear translocation, heterodimerization with ARNT, and transcriptional activation were observed in MEF cultures as a result of proteasome inhibition in the absence of xenobiotics. This process was correlated in time with a transient increase in ARNT expression at transcriptional level. We suggest that proteasome inhibition-mediated increase in nuclear ARNT could constitute a stimulus capable of inducing AHR translocation into the nuclear compartment. Increased nuclear AHR could favor the formation of active AHR-ARNT heterodimers, binding to regulatory sequences, and transcriptional activation of target genes. Although additional factors may participate in this process, physiological regulation of proteasome activity could be a mechanism used by the cell to control AHR-dependent transcriptional activity. This could allow AHR to participate in endogenous cellular functions independently of xenobiotics.