Treatment of Mice with 2,3,7,8-Tetrachlorodibenzo-p-Dioxin Leads to Aryl Hydrocarbon Receptor-Dependent Nuclear Translocation of NF-κB and Expression of Fas Ligand in Thymic Stromal Cells and Consequent Apoptosis in T Cells¹

Toxic manifestations of many halogenated aromatic hydrocarbons, such as 2,3,7,8-tetrachlorodibenzo-p-dioxin ((TCDD)³ or dioxin), primarily depend on the presence of the aryl hydrocarbon receptor (AhR), which is a ligand-dependent transcription factor belonging to the superfamily of basic-helix-loop-helix DNA-binding proteins (1). Unligated AhR resides in the cytosol and its inactivity is maintained by interactions with chaperone proteins and heat shock protein 90 (2). Binding of TCDD to AhR displaces chaperone proteins and heat shock protein 90 from the complex, generating a ligated receptor capable of translocating into the nucleus and associating with the AhR nuclear translocator (Arnt) protein. TCDD-induced modulation of gene expression is mediated through interactions between the Arnt/ligated AhR complex and a 5′-GCGTG-3′ DNA sequence, which is the core binding motif of dioxin responsive elements (DRE) located in the promoter regions of AhR-responsive genes (1). DREs have been identified in several TCDD-induced genes and are believed to contribute to TCDD’s potent modulation of gene expression. Furthermore, AhR knockout (KO) mice are resistant to TCDD-mediated liver toxicity, thymic atrophy, and suppression of cell-mediated and humoral immune responses, which prove that AhR is necessary for TCDD’s action (3, 4).

T cells in the thymus and periphery have been shown to be highly sensitive to TCDD-induced apoptosis. Studies from our laboratory and elsewhere, using Fas- or FasL-mutant mice, demonstrated that TCDD induces thymic atrophy and deletion of activated T cells by apoptosis regulated by Fas-FasL interactions (5, 6, 7, 8, 9, 10). In addition, we demonstrated that Fas gene promoter contained a functional DRE inasmuch as the binding of the AhR complex to this DRE led to the induction of the Fas gene in thymocytes (11). Also, while TCDD has been shown to increase the expression of the FasL gene in the thymus (5), it does not have a DRE in its promoter. In addition to Fas and FasL, TCDD induces the transcriptional expression of several proapoptotic genes in the thymus and spleen (12). In vitro studies using human leukemic T cell lines and EL-4 mouse thymoma cell lines demonstrated enhanced apoptosis following TCDD treatment via AhR-independent but caspase-dependent mechanisms (13, 14). Interestingly, TCDD-exposed leukemic T cell lines exhibited down-regulation of bcl-2 and induction of the JNK pathway (14), implicating a possible role for the mitochondrial pathway. Despite these studies, the precise molecular mechanisms underlying TCDD-induced apoptosis in vivo remain unresolved.

The aim of the present study was to elucidate the molecular pathways of TCDD-induced apoptosis in the thymus in vivo. We investigated whether TCDD-induced apoptosis is an AhR-dependent process and involves activation of the death-receptor and/or the mitochondrial pathway. We also studied the precise contributions of thymic stromal cells and T cells, in the context of Fas-FasL interactions, in the induction of apoptosis in the thymus following exposure to TCDD.

The following reagents were obtained from Invitrogen Life Technologies: RPMI 1640, DMEM, PBS, FBS, PCR reagents, and TRIzol reagent. The following mouse mAbs were purchased from BD Pharmingen: anti-CD4-FITC (L3T4), purified anti-CD3 (145-2C11), anti-CD3-FITC (ε-chain), anti-CD8-FITC (Ly-2), anti-CD44-PE (IM7), anti-αβTCR-FITC (H57-597), anti IL-2R-FITC (7D4), anti-J11d, anti-rat IgM-FITC, anti-CD90.1 Thy 1.1-PE (OX-7), anti-FasL (Kay-10), anti-mouse IgG, FcBlock, anti-FasL-PE (Kay-10), anti-TSA-1-FITC (MTS35, clone104), anti-Fas-PE (Jo2), anti-CD69-FITC, biotinylated anti-CD45.2 (clone 104), and PE-conjugated streptavidin. For Western blotting, the following primary Abs were used: caspase-9 (1:2,000; Cell Signaling), Bax (1:1,000; Cell Signaling), Bcl-x_L (1:1,000; Cell Signaling), Bid (1:1,000; R& D Systems), caspase-8 (3:1,000; LabVision), caspase-2 (1:2,000; Alexis Biochemicals), Smac (1:1,000; Alexis Biochemicals), β-actin (1:50,000; Sigma-Aldrich), and cytochrome c (2:1,000; BD Pharmingen). HRP-conjugated secondary Abs were obtained from Jackson ImmunoResearch Laboratories. Confocal microscopy studies were performed with the following Abs and reagents: tosyl-phe-chloromethylketone (TPCK) (Sigma-Aldrich), goat anti-NFκB p65 (c-20), rabbit anti-NFκB p50 (c-19) (Santa Cruz Biotechnology), anti-goat Cy3 (Jackson ImmunoResearch Laboratories), anti-rabbit Cy5 (Jackson ImmunoResearch Laboratories), Hoechst 33258 (Molecular Probes), and ProLong Antifade mounting medium (Molecular Probes). Transfection and luciferase assays were done with the following reagents: pGL3-basic vector (Promega), pCMV-β-galactosidase (BD Clontech) vector, the QuickChange II site-directed mutagenesis kit (Stratagene), the Effectene transfection kit (Qiagen), and the Dual-Light system (Applied Biosystems). The following reagents were used in RT-PCR: RNeasy Mini kit (Qiagen), RT Kit (Qiagen), Epicentre’s PCR premix F, and Platinum Taq Polymerase (Invitrogen Life Technologies).

AhR-deficient (AhR KO) mice created by deletion of exon 2 of the AhR gene (15) were generously provided as breeders by Dr. C. A. Bradfield (McArdle Laboratory for Cancer Research, University of Wisconsin Medical School, Madison, WI). These mice have been backcrossed >10 generations onto the C57BL/6 background and were generated by breeding heterozygous (AhR^+/−) females with homozygous (AhR^−/−) males. For genotyping of AHR KO mice, the QIAamp DNA mini kit (Qiagen) was used to prepare genomic tail DNA. PCR genotypic analysis of AHR KO mice was performed as previously described (3). Bid-deficient (Bid KO) mice were obtained as breeders from the laboratory of Dr. S. J. Korsmeyer (Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA) (16). These mice were backcrossed >9 generations onto the C57BL/6 background and bred as homozygotes in our animal facility. C57BL/6 mice, referred as C57BL/6 wild-type, were purchased from the National Institutes of Health (Bethesda, MD) and used as control mice. B6.PL-Thy1^a/Cy (Thy 1.1⁺) female mice were obtained from The Jackson Laboratory and used as a source of thymocytes expressing the Thy 1.1 allele. Female C57BL/6 lpr/lpr (lpr) and C57BL/6 gld/gld (gld) mice were purchased from The Jackson Laboratory and bred in our animal facilities. All strains of mice were used at ∼6 wk of age. At this age, the lpr and gld mice on the C57BL/6 background do not develop lymphoproliferative disease and have normal thymus (5). All animals were housed in polyethylene cages equipped with filter tops and wood shavings. Each animal cage had rodent chow and tap water ad libitum. Mice were housed in an environment of constant temperature (23 ± 2°C) and on a 12 h-light:12 h-dark lighting schedule.

TCDD was generously donated by Dr. K. Chae (National Institute on Environmental Health Sciences, Research Triangle Park, NC). TCDD was dissolved in acetone and diluted in corn oil. The solution was gently heated to evaporate the acetone. Animals were injected i.p. with either a single dose of TCDD (1–50 μg/kg) or the vehicle control (corn oil).

TCDD- or vehicle-exposed animals were euthanized, weighed, and dissected to remove the thymi. The relative thymic weights were calculated as follows: (organ weight, mg)/(body weight, g). Single cell suspensions were prepared as previously described (5). Cell viability was determined on a hemacytometer by using trypan blue dye and observing the cells in an inverted phase contrast microscope. Cell viability data from four to six experiments were pooled and depicted as mean cellularity or mean relative organ weights ± SEM.

Thymocytes (1 × 10⁶) were washed with PBS and incubated for 30 min on ice with 0.5 μg of the following primary mAbs: FITC-anti CD4 (L3T4), FITC-anti CD3 (ε-chain), FITC-anti CD8 (Ly-2), PE-anti CD44 (IM7), FITC-anti αβTCR (H57-597), FITC-anti IL-2R (7D4), anti-Fas-PE (Jo2), or anti-CD69-FITC. Detection of J11d marker was done by staining cells with J11d mAb followed by FITC- anti-rat IgM mAbs. After incubation with the mAbs, cells were then washed once with PBS. Negative controls consisted of cells that were stained with appropriate isotype-specific Abs. Stained cells were analyzed by a BD Biosciences FACScan flow cytometer. Ten thousand cells were analyzed per sample. Dead cells, clumps, and debris were excluded electronically by gating on forward vs side scatter. Data were depicted as percentage positive cells expressing the surface marker, or as mean intensity of fluorescence (MIF), which represents the density of expression of the surface marker. The data from four to six independent experiments were pooled and depicted as mean MIF ± SEM.

Thymi from TCDD- or vehicle-treated mice were dissected and immediately placed in liquid nitrogen. Total RNA was isolated from thymi using the TRIzol Reagent according to the manufacturer’s instructions (Invitrogen Life Technologies). To assess RNA integrity, RNA (1–2 μg) was mixed with an equal volume of denaturing loading buffer (2× Tris-borate-EDTA buffer, pH 8.3, 13% Ficoll, 0.01% bromphenol blue, and 7 M urea), incubated at 70°C for 10 min followed by visualization of the 28 S and 18 S ribosomal bands on a 1% agarose gel (1× Tris-borate-EDTA buffer) at 80 V for 1 h. The concentration of RNA was determined spectrophotometrically. The AmpoLabeling-LPR kit (SuperArray) was used to convert RNA (1.5 μg) to biotinylated-cDNA probe. Labeled probes were hybridized to the GEArray Q Series Mouse Apoptosis Gene arrays (SuperArray) according to the manufacturer’s instructions.

After washing, the arrays were visualized by chemiluminescence and gene expression was quantified by scanning densitometry. A complete list of the 96 apoptosis-related genes contained in each array can be found in a recent publication from our laboratory (11) or at 〈www.superarray.com〉. Data analysis was done using the GEAarray Analyzer software developed by SuperArray. All raw signal intensities were corrected for background by subtracting the signal intensity of the negative control (pUC18 DNA). Loading was also normalized based on intensity of hybridization signals to the housekeeping gene β-actin. The relative levels of cDNA expression were assessed by comparing the signals from TCDD-treated wild-type or AHR KO thymocytes to those obtained by the respective control sample. Microarray data were expressed as fold increase of the signal intensity of the TCDD sample relative to the control sample.

The thymic stromal cells, which constitute ∼1–2% of the cells found in the thymus, were isolated from AhR^+/+ or AhR KO mice as described by Gray et al. (17). In some experiments, nonlymphoid stromal cells were identified by the deficiency of CD45 using flow cytometry, which enabled us to distinguish these cells from the T cells which express CD45, and constitute > 98% of thymocytes. Also, the CD45-deficient stromal cells expressed the thymic shared Ag (TSA) marker. As a source of T cells, we used whole thymocytes by isolating the thymi from mice and subjecting them to gentle homogenization in 10% FBS/RPMI 1640 medium with subsequent determination of cell viability. For coculture studies, 4 × 10⁶/ml stromal cells were cultured overnight with 2 × 10⁶ thymic T cells/ml obtained from Thy 1.1 or lpr mice in 24-well plates. For blocking studies, 5 μg/ml of either anti-mouse FasL mAb (Kay-10) or a control anti-mouse IgG mAb was added to the stroma/T cell cultures. In some experiments, stromal cells from untreated wild-type mice were cultured for 24 h with various concentrations of TCDD (0.1, 1, or 10 nM) or the vehicle (0.01% DMSO). To stain stromal cells (1–2 × 10⁶), they were preincubated with FcBlock for 20 min to block nonspecific binding to FcRs. Subsequently, cells were incubated for 30 min on ice with either PE-conjugated anti-mouse FasL (Kay-10) or FITC-conjugated anti-mouse TSA-1 (MTS35) (clone104). For CD45 staining, stromal cells were first incubated with biotinylated anti-mouse CD45.2 mAb (clone 104) followed by PE-conjugated streptavidin. After final incubation, cells were washed and resuspended in 200 μl of EDTA/FACS buffer for flow cytometric analysis. Fifty thousand cells were analyzed per sample. Immunofluorescense stainings for FasL, TSA-1, and CD45 were replicated at least three times and representative data were depicted as the percentage of cells expressing the marker.

Total RNA was isolated from thymocytes using the TRIzol reagent according to the manufacturer’s instructions (Invitrogen Life Technologies). Five hundred nanograms of RNA were subjected to cDNA synthesis using the Omniscript RT kit (Qiagen). The reverse transcriptase reaction was run at 37°C for 1 h. First-strand cDNA was amplified in 50-μl final volume containing 5 μl of each primer (final concentration 0.2 μM), 0.5 μl of TaqDNA polymerase, 2 μl of 50 mM MgCl₂, 4 μl of 2.5 mM dNTP, 5 μl of 10× PCR buffer, 1 μl of cDNA, and water. The amplification conditions were 2 min at 94°C, 29 cycles at 94°C for 30 s, annealing temperature at 40 s, 72°C for 1 min, and one final cycle at 72°C for 2 min. The following primer sequences were used: Cyp1a1 U (5′-CCC ACA GCA CCA CAA GAG ATA-3′); Cyp1a1 L (5′-AAG TAG GAG GCA GGC ACA ATG TC-3′); β-actin U (5′-AAG GCC AAC CGT GAA AAG ATG ACC-3′); and β-actin L (5′-ACC GCT CGT TGC CAA TAG TGA TGA-3′). The annealing temperatures and product sizes for these primers are as follows: Cyp1a1 (62°C, 499 bp); and β-actin (55 or 62°C, 427 bp). Ten microliters of β-actin product and 20 μl of the other mouse gene products were resolved on a 1.5% agarose gel. The density of the bands for vehicle vs TCDD treatment was calculated using Scion Image software available at 〈http://rsb.info.nih.gov/nih-image〉. To normalize RNA loading and PCR variations, the signals of target genes were corrected with the signals of β-actin mRNA. Differences between vehicle and TCDD treatments were expressed as a ratio and percent change caused by TCDD treatment. Gene expression was checked at least in three independent experiments and two representative data were depicted.

FITC-dUTP nick-end labeling or TUNEL assay (Roche Applied Science) was used to detect apoptosis in thymic T cells as previously described (5, 7). In addition, double-staining studies were conducted to determine the amount of apoptosis in Thy 1.1⁺ thymic T cells cultured with TCDD- or vehicle-exposed stromal cells. After overnight incubation of stromal cells with Thy 1.1⁺ thymic T cells, cells were harvested and washed with PBS. Cells were incubated with 0.5 μg of PE-conjugated anti-mouse CD90.1 Thy 1.1 (OX-7) mAb for 30 min on ice followed by one wash with PBS. Similarly, stromal cultures cultured with lpr thymocytes were harvested, washed with PBS, and stained with PE-conjugated anti-mouse Fas mAb (0.5 μg; BD Pharmingen) for 30 min on ice followed by one wash with PBS. Next, cells were subjected to the TUNEL protocol as described above. Fluorescence of 10,000 cells was collected by flow cytometry. Electronic compensation for fluorochrome spectral overlap was performed during flow cytometric analysis of double-stained cells. Gated Thy 1.1⁺ thymocytes were analyzed for TUNEL positivity. When stromal cells were cultured with lpr thymocytes, gated Fas⁻ cells were analyzed for TUNEL positivity. Data from four independent experiments were pooled and depicted as mean percentage of apoptosis ± SEM.

Freshly isolated thymic T cells (1 × 10⁶) from TCDD- or vehicle-treated mice were resuspended in PBS and incubated with 40 nM of 3,3′-dihexyloxacarbocynine (DIOC₆; Sigma-Aldrich) and 50 μg/ml propidium iodide (final concentrations) for 15 min at 37°C. Ten thousand cells were analyzed by flow cytometry. Δψ_m was evaluated in propidium iodide-negative (live) cells and at least three independent experiments were done.

For preparation of tissue lysates, thymi were excised and homogenized in cold PBS. Thymocytes (50–100 × 10⁶ cells) were lysed with 100–200 μl of CHAPS cell extract buffer (Cell Signaling). Lysates were clarified by centrifugation at 14,000 rpm for 5 min, and their protein content was determined by the Bradford protein assay (Bio-Rad). Ten or 50 μg of total protein was fractionated in a SDS-acrylamide gel, transferred to nitrocellulose, and blocked for 1 h with 5% nonfat dry milk in 1× TBS/0.1% Tween 20. Membranes were exposed to the primary Abs overnight at 4°C on a shaker, then incubated with HRP-conjugated secondary Abs and processed with the ECL system according to the manufacturer’s protocol (Amersham Biosciences). Protein visualization of β-actin expression was used as a loading control. The density of the bands was quantified by densitometry. β-actin was used to normalize the samples. Western data were expressed as percentage change of the band density of the TCDD sample relative to the vehicle control sample.

Caspase-3/7 activity was measured in thymocytes from TCDD- or vehicle-exposed mice using the Apo-ONE Homogeneous Caspase-3/7 Assay according to the manufacturer’s instructions (Promega). A Wallac 1420 multilabel counter (PerkinElmer) was used to measure the relative fluorescence units of each sample at an excitation wavelength of 485 nm and at an emission wavelength of 535 nm. Data from four to six mice were pooled and depicted as mean fluorescence units ± SEM.

Stromal cells were prepared from AhR^+/+ or AhR KO mice treated with TCDD or the vehicle for 6 h. In some experiments, stromal cells were isolated from untreated AhR^+/+ mice and then cultured overnight with the following treatments: DMSO, 1 nM TCDD, 20 μM TPCK (NF-κB inhibitor; Sigma-Aldrich) or 1 nM TCDD + 20 μM TPCK. The cells were treated with TPCK 1 h before addition of TCDD. After treatment, stromal cells were resuspended at 5 × 10⁶ cells/ml and 100 μl of this cell suspension was layered onto polylysine slides. Cells were allowed to adhere for 30 min at 37°C. Next, cells were fixed in 3.7%-buffered paraformaldehyde at room temperature for 15 min and permeabilized in 0.2% Triton X-100 in PBS at room temperature for 5 min. Confocal microscopy studies were performed with the following Abs: goat anti-NFκB p65 (c-20; Santa Cruz Biotechnology), rabbit anti-NFκB p50 (c-19; Santa Cruz Biotechnology), or purified hamster anti-mouse CD3 (145-2C11; BD Pharmingen). After blocking the FcRs with rat anti-mouse CD16/CD32 (Fc Block; BD Pharmingen), Abs were added at appropriate concentrations for 1 h at room temperature. Subsequently, cells were washed twice in PBS and stained with anti-goat Cy3, anti-rabbit Cy5, or anti-hamster Cy2 for 1 h at room temperature. Cells were washed twice with PBS and nuclei were stained with Hoechst 33258 for 1 min before slides were mounted in ProLong Antifade mounting medium and examined using a Zeiss LSM 510 Meta laser scanning confocal microscope. Images were acquired using thin optical sections (∼0.3 μm) through the x-y axis. Images were collected both simultaneously and sequentially to rule out bleed through from one fluorescence channel to another. Three independent samples were evaluated and representative data were depicted. In NF-κB nuclear translocation studies, 50 or more cells exposed to TCDD or the vehicle were individually analyzed and the percentage of thymocytes showing activation of NF-κB was calculated. Also, a representative picture of the stromal cells was depicted.

The FasL promoter-reporter construct (pGL-3-FasL720) made in pGL3-basic vector, a luciferase reporter plasmid (Promega), was kindly provided by Dr. T. Ratliff (University of Iowa, Iowa City, Iowa). In this construct, FasL promoter region (720 bp) spans −689 to +65 (in reference to the translational start site) and was generated using the upstream primer 5′-GTACCTCAGTTTTCATCTGGTGACCAGAAG-3′ and the downstream primer 5′-GCACCCAGCCCCAGGAAAGG-3′ (18). Site-directed mutations to NF-κB sites (NF-κB1 and NF-κB2) present in FasL promoter were generated using the QuickChange II Site-Directed Mutagenesis kit and according to manufacturer’s instructions (Stratagene). pGL-3-FasL720 was used as the template to generate mutations. The primers used to generate mutations as indicated by underlining the nucleotides in NF-κB1 were 5′-GAGAAAGGTGTTTAAATTGACTGC-3′ and its antiparallel sequence and for NF-κB2 were 5′-CCTTGGTCTTTTAAACATGCCTCAGC-3′ and its antiparallel sequence. Positive clones were sequenced to confirm mutations in NF-κB1 and NF-κB2. In this study, we used pGL-3 basic, pGL-3 control (Promega) vectors as controls, and pcDNA3.1 with pCMV-β-galactosidase (BD Clontech) vector to normalize the transfection efficiency.

EL-4 cells were cultured in complete medium (complete medium contains DMEM supplemented with 10% FBS, 10 mM HEPES, 10 mM l-glutamine, and 10 μg/ml gentamicin) at 37°C, 5% CO₂. A series of FasL promoter-luciferase constructs (pGL-3-FasL270), pGL-3 basic, pGL-3 control, and pcDNA3.1 with pCMV-β-galactosidase were used in transient transfection assays. Transfections into EL-4 cells were performed using the Effectene Transfection kit and according to the manufacturer’s instructions (Qiagen). Briefly, EL-4 cells were collected, washed twice with sterilized PBS, and cultured in six-well plates 24 h before transfection. The following day, EL-4 cells at 50–70% confluence were transfected with 5–10 μg/well pGL-3-FasL270, pGL-3 basic, and pGL-3 control plasmid vectors and 1 μg/well pCMV β-galactosidase control vector. Two days after transfection, the cells were split in 24-well plates and were exposed to 100 nM/ml TCDD (dissolved in DMSO) for 18 h. The cells without treatment or those treated with DMSO were used as controls. Luciferase and β-galactosidase assays were performed using Dual-Light System and according to manufacturer’s instructions (Applied Biosystems). In other sets of experiments, we used 100 ng/ml α-naphthoflavone (an antagonist for TCDD) with or without TCDD in culture and Luciferase and β-galactosidase assays were performed. The light units were assayed by luminometry (Monolight 2010; Analytical Luminescence Laboratory). Luciferase activity was normalized by dividing the mean control luciferase relative light unit (RLU) by the mean β-galactosidase RLU. The normalized luciferase RLU from the samples were divided by the normalized RLU of the untreated sample and values were expressed as “normalized fold increase.” Data obtained represent the average of three transfection experiments, each conducted in triplicate and depicted as mean ± SD.

To detect the expression of AhR in EL4 cells, total RNA was isolated using the RNeasy Mini kit according to the company’s protocol (Qiagen). cDNA was synthesized using the Sensiscript RT kit (Qiagen). RT-PCR were prepared using Epicentre’s PCR premix F, Platinum Taq Polymerase (Invitrogen Life Technologies), and mouse AhR primers (mAhR forward primer: 5′-GCGGCCGCAGGAAGTGAGG-3′ and mAhR reverse primer: 5′-GTGCCGTTGATT TGCGTGTGCT-3′). PCR was performed for 30 cycles using the following parameters; 30 s at 95°C, 40 s at the 60°C annealing temperature, and 60 s at 72°C, with a final incubation at 72°C for 10 min. PCR product was resolved using 1.5% agarose gel. The expected size of the AhR PCR product was 482 bp.

Lymphoid cells (5 × 10⁵ cells/well) were cultured in 96-well plates containing 0.2 ml of medium alone or with anti-CD3 (5 μg/ml; BD Pharmingen). Cells were incubated for 48 h at 37°C in 5% CO₂ and pulsed with 2 μCi [³H]thymidine during the last 8 h of incubation. Cells were harvested using an automated cell harvester (Skatron). The amount of radioactivity was determined using a scintillation counter (TM Analytic 6895) and the mean cpm ± SEM of triplicate cultures was calculated. Data obtained represent the average of four independent experiments, each conducted in triplicate and depicted as mean ± SD.

Results are presented as the mean ± SEM of four to six independent experiments per treatment group. Statistical analyses were performed using JMP IN statistical software (SAS Institute). ANOVA was performed with a significance level of α = 0.05. Comparison among means were made using the Tukey-Kramer honestly significant difference test, with values of p < 0.05 considered to be statistically significant.

To investigate whether TCDD-induced apoptosis in thymocytes is an AhR-dependent process, we used AhR KO mice. All mice were genotyped as well as examined for Cyp1a1 mRNA expression in the thymus, to confirm the genotype. Data from a representative experiment shown in Fig. 1,A demonstrated that TCDD treatment induced the expression of Cyp1a1 in AhR^+/+ (wild-type) but not AhR KO mice. Next, AhR^+/+ and AhR KO mice were injected with TCDD (50 μg/kg) or the vehicle and 5 days later, thymus was analyzed for weight, cellularity, and apoptosis. Significant reductions in relative thymic weights (Fig. 1,B) and thymic cellularity (Fig. 1 C) were observed in TCDD-treated AhR^+/+ but not TCDD-treated AhR KO mice, when compared with the vehicle-treated control groups. We also observed that vehicle-treated AhR KO mice expressed significantly lower cell numbers in the thymus when compared with vehicle-treated AhR^+/+ mice. This resulted from the fact that AhR KO mice had smaller size thymi when compared with the AhR^+/+ mice.

We next examined whether the resistance to TCDD-induced thymic atrophy exhibited by AhR KO mice correlated with inability of thymocytes to undergo apoptosis. To this end, freshly isolated thymocytes from TCDD- or vehicle-treated AhR^+/+ and AhR KO were cultured for an additional 24 h to evaluate apoptosis using the TUNEL assay. This strategy offers the advantage of studying apoptotic cells in the absence of being rapidly cleared by phagocytic cells, as known to occur in vivo (7). As seen from a representative experiment shown in Fig. 1,D (upper panel) and pooled data from four to six experiments in Fig. 1,D (lower panel), TCDD-treated AhR^+/+ but not AhR KO mice showed a significant increase in percentage of apoptotic thymocytes when compared with the vehicle-treated mice. These data were corroborated by the observation that caspase-3/7 enzymatic activity was significantly increased in thymocytes from AhR^+/+ but not AhR KO mice treated with TCDD (Fig. 1 E).

Because the AhR KO mice had smaller sized thymi, we performed additional studies to test whether they had normal phenotypic and functional distribution of T cells. Double staining for CD4 and CD8 markers revealed that thymocytes from AhR KO had normal proportions of double-positive, double-negative, and single-positive T cells when compared with the AhR^+/+ mice (Fig. 2,A). Moreover, the thymocytes from AhR KO mice showed similar levels and percentages of Fas⁺ and CD69⁺ T cells when compared with the AhR^+/+ mice (Fig. 2,B). Lastly, when thymocytes were activated through the TCR, the proliferative response seen using thymocytes from AhR KO mice was comparable to that seen using AhR^+/+ mice (Fig. 2 C). Together, these data suggested that the thymocytes from AhR KO mice are phenotypically and functionally similar to the wild-type mice.

Apoptotic thymocytes including those exposed to TCDD are known to up-regulate the expression of CD3, αβTCR, IL-2R, and CD44 markers while down-regulating the expression of CD4, CD8, and J11d markers (6, 19). It was possible that the thymocytes from the AhR KO mice had altered distribution of such apoptotic markers leading to differential susceptibility to TCDD-induced apoptosis. To address this, AhR^+/+ and AhR KO mice were injected with TCDD or the vehicle and analyzed for the above markers. As seen from Fig. 3, vehicle-treated AhR^+/+ and AhR KO mice exhibited similar percentages and MIF of various T cell markers. Furthermore, TCDD-treated AhR^+/+ mice showed phenotypic alterations characteristic of apoptotic cells such as up-regulation of CD3, αβTCR, IL-2R, and CD44 markers and down-regulation of CD4, CD8, and J11d as indicated by alterations in MIF values. Interestingly, thymocytes from TCDD-treated AhR KO mice failed to exhibit phenotypic changes when compared with their respective vehicle controls. Together, these data suggested that deficiency of AhR does not alter the expression of several of the surface markers that we studied in the thymus and, furthermore, it confers resistance to the phenotypic changes in the thymus induced by TCDD.

cDNA microarray analysis was used to profile apoptotic genes regulated through the AhR-signaling pathway. To this end, thymi were collected from AhR^+/+ and AhR KO mice, 24 h after treatment with 50 μg/kg TCDD, and analyzed for 96 genes regulating apoptosis. A list of the genes is available at the manufacturer’s website (〈www.superarray.com〉). Of these genes, the expression of 14 genes appeared to be dependent on AhR expression as they were increased >2-fold in TCDD-treated AhR^+/+ but not AhR KO mice when compared with the respective vehicle controls. Significant decrease in gene expression (>2-fold) was not detected after TCDD treatment in either AhR^+/+ or AhR KO mice. We categorized the AhR-regulated genes into five groups as shown in Table I: 1) Bcl-2 family (Bad, Bcl-w, bid, and bok/mtd); 2) TNF superfamily (FasL, OX40L, OPG, and TNFSF11); 3) apoptosis-related enzymes (caspase-1 and DAP kinase); 4) inhibitors of apoptosis (NAIP1 and IAP2); and 5) adaptor molecules (CRADD and TRAF5). In addition, we used MatInspector software (20) to analyze for the presence of potential DREs in those genes that were up-regulated by TCDD and had published promoter sequences. At least one potential DRE sequence was found in the following genes: Bad, Bid, and OPG (data not shown). No DREs were predicted in the promoter regions of FasL. The rest of the genes in Table I were not analyzed as promoter sequences are not currently available.

Interactions between thymic stromal cells and immature T cells are known to play a critical role during T cell development. FasL, one of the triggers of the death receptor pathway, is expressed on stromal cells but not T cells from the thymus (21). In the current study, we therefore tested whether TCDD treatment would up-regulate the expression of FasL on thymic stromal cells and, if so, whether this would induce increased apoptosis in thymic T cells. To this end, we examined CD45 expression on cells isolated from the thymus, which enabled us to discriminate between lymphoid CD45⁺ (thymocytes) and nonlymphoid CD45⁻ cells (stromal cells). To further confirm that the CD45⁻ cells were stromal cells, we determined the expression of TSA, which is a specific marker for thymic stromal cells. Thus, when the CD45⁻ cells were gated and analyzed for TSA, over 90% of the cells were TSA⁺ (Fig. 4,A). Next, we enriched the thymocytes for stromal cells 24 h after injecting wild-type mice with vehicle or TCDD and examined the levels of FasL by flow cytometry. When the CD45⁻ cells were gated and analyzed for FasL by flow cytometry, we noted a significant increase in the percentage of FasL⁺ cells following exposure to TCDD along with an increase in the levels of FasL, when compared with vehicle-treated controls (Fig. 4,B). When CD45⁺ cells from the thymus were similarly analyzed, the cells from vehicle-treated mice failed to express FasL (Fig. 4,C), consistent with the observation that FasL is expressed on thymic stromal but not T cells (21). Moreover, TCDD treatment failed to induce FasL in these cells (Fig. 4 C). Together, these data suggested that TCDD up-regulates the expression of FasL on thymic stromal cells but not on T cells.

Thymic T cells express high levels of Fas. Thus, we wanted to determine whether the TCDD-induced increase in expression of FasL on stromal cells was sufficient to mediate higher levels of apoptosis in T cells that come in contact with stromal cells. To this end, we tested whether coculture of FasL⁺ stromal cells from TCDD-treated wild-type mice with normal untreated Fas⁺ thymic T cells would lead to increased apoptosis in the latter population. To ensure that we were not analyzing any T cells found in the stromal cell preparations from TCDD-treated mice, which could influence the results on apoptosis, we conducted mixing experiments in which Thy 1.2⁺ stromal cells isolated from TCDD- or vehicle-treated mice exposed for 24 h were mixed with freshly isolated Thy 1.1⁺ thymic T cells from untreated Thy 1.1 mice. Next, we analyzed the Thy 1.1⁺ T cells for apoptosis as described below. To confirm that the interactions between stromal cells and T cells involved FasL, the cultures were also incubated with isotype control Abs or anti-FasL mAbs. After overnight coculture, the cells were double-stained with PE-anti-Thy 1.1 mAb and FITC-dUTP (TUNEL) to selectively detect apoptosis in Thy 1.1⁺ thymic T cells. Analysis of gated Thy 1.1⁺ cells showed that overnight culture of untreated Thy 1.1⁺ thymic T cells with vehicle-exposed Thy 1.2⁺ stromal cells along with control Abs resulted in significant apoptosis in Thy 1.1⁺ cells (64.3%) (Fig. 5,A). This can be attributed to spontaneous apoptosis and/or FasL-induced apoptosis because normal stromal cells express FasL as shown earlier (Fig. 4) (21). Interestingly, culture of untreated Thy 1.1⁺ thymic T cells with TCDD-exposed Thy 1.2⁺ stromal cells along with control Abs resulted in higher levels of apoptosis in Thy 1.1⁺ cells (83.2%) (Fig. 5,A). When similar mixing experiments were performed in the presence of anti-FasL (Kay-10) Abs, there was ∼30% decrease in Thy 1.1⁺ apoptotic cells cultured with vehicle-treated stromal cells, whereas, ∼50% blocking was displayed in apoptotic Thy 1.1⁺ cells cultured with TCDD-treated stromal cells. These data suggested that the apoptosis induced in Thy 1.1⁺ thymic T cells following culture with TCDD-exposed stromal cells resulted at least in part from Fas-FasL interactions. Furthermore, when untreated Thy 1.1⁺ thymic T cells were mixed with TCDD-treated stromal cells from FasL-defective (gld/gld) mice, no significant increase in apoptosis was detected in Thy 1.1⁺ thymic T cells when compared with the controls (Fig. 5,B). Similarly, when TCDD-exposed wild-type stromal cells were mixed with Fas-deficient thymic T cells, there was no significant increase in apoptosis in the latter cells when compared with the controls (Fig. 5,B). Moreover, when untreated Thy 1.1⁺ thymic T cells from wild-type mice were mixed with TCDD-treated stromal cells from AHR KO mice, no significant apoptosis was detected in Thy 1.1⁺ thymocytes when compared with the controls (Fig. 5 B). These studies conclusively demonstrated that interactions between TCDD-treated FasL⁺ stromal cells and untreated Fas⁺ T cells can trigger increased levels of apoptosis in the latter cells. Also, AhR expression in stromal cells played a critical role in the apoptosis of thymocytes following TCDD treatment.

Previous studies have demonstrated that NF-κB transcription factors are involved in the expression of FasL, consequently increasing apoptosis (22, 23). Thus, we next investigated whether TCDD induces NF-κB activation in stromal cells. To this end, we used confocal microscopy to study nuclear translocation of NF-κB in individual stromal cells. As seen from Fig. 6,A, morphologically, the stromal cells were easily distinguishable from T cells inasmuch as stromal cells were CD3 negative, larger in size and were more granular with abundant cytoplasm, while T cells were CD3⁺, smaller in size with large nucleus. Analysis of NF-κB subunits, p50 and p65, showed that they were expressed in the cytoplasm of stromal cells and T cells of vehicle-treated mice as previously described (24, 25). Interestingly, following in vivo TCDD treatment, both p50 and p65 subunits translocated to and colocalized in the nucleus of stromal cells but not thymic T cells from AhR^+/+ and AhR KO mice (Fig. 6,B). Upon enumeration, 38% of the stromal cells from vehicle-treated AhR^+/+ mice were found to exhibit nuclear translocation of NF-κB as against 78% in TCDD-treated AhR^+/+ mice (data not depicted). It should be noted that unlike the stromal cells, only 9% of T cells from vehicle-treated and 11% from TCDD-treated AhR^+/+ mice showed nuclear translocation of NF-κB. In vitro treatment of stromal cells with TCDD also showed strong nuclear translocation and colocalization of NF-κB subunits (p50 and p65), which was blocked by the known NF-κB inhibitor TPCK (Fig. 7). Together, the confocal studies suggested that TCDD activates NF-κB in thymic stromal cells through an AhR-dependent pathway.

As mentioned earlier, we were unable to detect DREs in the promoter sequence of the mouse FasL gene as analyzed using MatInspector software (20). However, we were able to identify two NF-κB sites on the promoter designated NF-κB1 and NF-κB2 (Fig. 8,A). Therefore, we conducted a series of experiments to determine the role of NF-κB, activated by TCDD, on FasL induction using the luciferase reporter system. First, to further corroborate the ability of TCDD to induce FasL, we used a mouse FasL promoter reporter construct (pGL-3-FasL720) to transfect EL-4 cells. This transformed T cell line expresses FasL and has previously been used in studying the regulation of gene expression (26, 27, 28). Moreover, EL-4 cells also expressed the AhR as detected by RT-PCR analysis (Fig. 8,B). We observed a >15-fold increase in luciferase expression when EL-4 cells transfected with FasL promoter were subjected to TCDD treatment (Fig. 8,C). In the same experimental settings, we observed very minimal (1- to 2-fold) expression of luciferase when the cells were not treated or treated with DMSO (Fig. 8,C). When α-naphthoflavone, an antagonist for TCDD that interacts with AhR, was used in combination with TCDD, luciferase induction was significantly inhibited. Furthermore, we observed no or minimal effects on luciferase induction (1- to 2-fold) when α-naphthoflavone was used alone in the culture (Fig. 8 C).

Mouse FasL promoter contains at least two NF-κB sites (Fig. 8,A) which we designated as NF-κB1 (GGTGTTTCCC) and NF-κB2 (GGTCTTTTCCC). To determine the role of NF-κB in up-regulation of the FasL promoter, we generated pGL-3-FasL720 constructs that carried mutations as indicated by underlining the nucleotides in NF-κB1 (GGTGTTTAAATT), NF-κB2 (GGTCTTTTAAACAT), and in both NF-κB1 and NF-κB2. These mutant constructs were then used in transfection of EL-4 cells. The data shown in Fig. 8 D indicated that mutations in either NF-κB1 or NF-κB2 caused a significant decrease in TCDD-induced luciferase expression (from ∼15- to ∼1–2-fold). Furthermore, mutations in both NF-κB1 and NF-κB2 almost completely blocked TCDD-induced luciferase activity. These data showed that TCDD may induce the expression of FasL by directly regulating FasL promoter activity through NF-κB.

To test whether FasL-mediated apoptotic signaling in thymocytes occurred via activation of caspase-8, Western blot analysis was conducted with thymocytes isolated from vehicle or TCDD-treated wild-type mice (Fig. 9,A). At 18 h post-TCDD exposure, a significant increase in cleavage of caspase-8 (47 kDa) was detected in thymocytes, when compared with levels in the vehicle controls (Fig. 9).

To address whether the mitochondrial pathway was also involved, Western blot analysis was conducted to evaluate the expression of critical proteins involved in this pathway of apoptosis. Data shown in Fig. 9, B and C, revealed that thymocytes from TCDD-treated wild-type mice exhibited release of cytochrome c at 12 h. Moreover, TCDD treatment caused increased expression of the Bax protein at 12 h and caspase-9 at 18 h. Cleavage of Bid, a substrate of caspase-8, was enhanced at 18 h in TCDD-exposed thymocytes, suggesting cross-talk between the two pathways of apoptosis. No significant changes in protein expression were detected in Smac, Bcl-x_L, and the cleaved form of caspase-2 in TCDD-exposed thymocytes when compared with control groups. It should be noted that the increased levels of caspase cleavage and induction of other apoptotic molecules following TCDD treatment was not robust when compared with the vehicle controls. This may have resulted from the fact that the vehicle controls also showed significant levels of caspase cleavage, which was expected due to spontaneous ongoing apoptosis in normal thymus. Moreover, TCDD-induced apoptotic thymocytes are rapidly cleared in vivo by phagocytic cells, thereby leaving very few apoptotic cells to be detected upon isolation from the mice (7).

Next, we analyzed changes in the Δψ_m using the DIOC₆ dye. When thymocytes from wild-type mice harvested 6–18 h following injection with 50 μg/kg TCDD were screened, a striking reduction in Δψ_m was observed at 12 h (Fig. 10,A). At longer exposure times (>18 h), no changes in Δψ_m were detected in TCDD-exposed thymocytes when compared with control cells (data not shown), possibly due to clearing of apoptotic cells in vivo. Moreover, a dose-dependent loss of Δψ_m was observed in TCDD-treated thymocytes starting at a dose of 10 μg/kg or higher (Fig. 10 B).

To further address the role of Bid in TCDD-induced apoptosis, wild-type or Bid KO mice were exposed to 50 μg/kg TCDD or the vehicle for 3 days. Both strains exhibited significant reduction in thymic cell numbers (Fig. 11,A) and showed enhanced apoptosis (Fig. 11 B) following TCDD treatment. It should be noted that the thymus from Bid KO mice was also sensitive to atrophy induced by TCDD when administered at 10 μg/kg (data not shown). These data suggested that Bid is dispensable for TCDD-induced apoptosis in the thymus.

Previous studies from our laboratory have shown that C57BL/6 mice with mutations in Fas and FasL are resistant to TCDD-induced thymic atrophy at concentrations <50 μg/kg body weight (5, 6, 7, 8, 9, 10). However, at higher than this dose, these mice were sensitive to thymic atrophy, thereby suggesting that additional alternative pathways may play a role (5). This observation was also supported by our recent demonstration that TCDD induces expression of several proapoptotic genes besides Fas and FasL (11, 12). The precise mechanisms by which TCDD-induced Fas/FasL interactions or other apoptotic genes trigger apoptosis in thymic T cells are unclear. In the current study, we conducted in-depth analysis of the in vivo effects of TCDD on thymic apoptotic signaling pathways. First, our results indicated that susceptibility to TCDD-induced apoptosis in thymocytes is dependent on AhR expression. Second, we showed that this apoptotic process might involve members of the death receptor pathway and the recruitment of caspase 8. Third, we demonstrated that TCDD treatment might increase FasL expression on the surface of thymic stromal cells, and that Fas⁺ T cells might undergo enhanced apoptosis when interacting with such stromal cells. Fourth, we found that TCDD treatment leads to nuclear translocation of NF-κB in thymic stromal cells but not T cells through an AhR-dependent process that regulates FasL promoter activity. Last, TCDD, at higher doses, was also able to induce mitochondrial pathway of cell death in the thymus by activating proapoptotic members of the Bcl-2 family, promoting loss of Δψ_m, inducing cleavage of caspase-9, and releasing cytochrome c.

Our laboratory demonstrated for the first time that thymic atrophy induced by TCDD can be accounted for, at least in part, by induction of apoptosis (5, 6, 7, 8). Subsequently, several studies have extended this observation to a wide range of target tissues. TCDD is known to induce apoptosis in human T lymphoblastic and leukemic cell lines (14, 29), luteinized granulosa cells (30), central nervous tissue of zebra fish embryos (31), several tissues of developing Fundulus heteroclitus embryos (32), vascular tissue of medaka embryos (33), and dental epithelium (34). Despite this evidence, the mechanisms of apoptosis induced by TCDD are not clear and may vary based on the cell type. For example, in vitro studies using T cell lines such as Jurkat, have shown that TCDD activates MAPK-signaling pathways including JNK and ERK, suggesting that MAPK pathways may be involved in TCDD-induced cell death (35). In EL-4 T cells, it was shown that insulin-like growth factor-binding protein-6 was found to mediate the immunotoxic effects of TCDD in an AhR-independent pathway (13).

The current study investigated the in vivo effects of TCDD and demonstrated that up-regulation of FasL on thymic stromal cells plays a crucial role in T cell apoptosis. Previous studies from our laboratory demonstrated that FasL was up-regulated in the thymus following TCDD treatment (5). In these studies, we had used the whole thymus and analyzed for FasL expression using RT-PCR. In the current study, we used purified stromal cells and T cells from the thymus and found that FasL is expressed only in the stromal but not T cells. These data are consistent with a recent study showing that FasL is constitutively expressed in thymic stromal cells but not in T cells. Moreover, we noted that upon TCDD treatment, FasL was up-regulated only in thymic stromal cells but not in T cells. These data suggested that TCDD may facilitate the expression of FasL on cells that already express this molecule rather than induce its expression on cells that fail to express FasL.

Signaling through AhR may be necessary to induce the expression of FasL inasmuch as FasL was not up-regulated in TCDD-treated AhR-deficient mice. We confirmed these data using the luciferase reporter assay. It is interesting to note that the FasL promoter does not have a DRE site as analyzed by MatInspector software, thereby suggesting that its activation may be mediated through other transcription factors. The FasL promoter contained two NF-κB sites and mutations in either or both of these sites led to virtual abrogation of FasL promoter activity. These data suggested that TCDD-induced activation of NF-κB plays a crucial role in the up-regulation of FasL on thymic stromal cells. NF-κB transcription factors have been previously reported to mediate FasL up-regulation and induction of apoptosis in T cells (22, 23). Exposure of hepatoma cells to TCDD activates NF-κB and AP-1 in an AhR-dependent manner (36), consistent with the current observation. Induction of NF-κB activity has also been shown in the immature rat thymus following TCDD treatment (37). It should be noted that TCDD has also been shown to inhibit the activation of NF-κB in hepatoma cells, thereby suggesting that TCDD may exert differential effects in different cell types (38). NF-κB is constitutively expressed in some subsets of immature T cells in the thymus but not in majority of the mature T cells (25). This is consistent with our results that NF-κB was not expressed at significant levels in thymic T cells, whereas in stromal cells, increased levels of NF-κB were detected. Furthermore, TCDD treatment led to nuclear translocation of NF-κB in the stromal cells but not T cells by an AhR-dependent manner. NF-κB has been primarily known to function as an antiapoptotic transcription factor, although some studies have suggested that it can also act as a proapoptotic protein depending on cell type and the nature of apoptotic inducers (39, 40). The current study is novel inasmuch as it demonstrates that activation of NF-κB by TCDD in thymic stromal cells may regulate apoptosis induction in the thymus.

In the current study, we noted that the thymi from AhR KO mice were smaller in size and cellularity. However, comprehensive analysis revealed that AhR deficiency did not alter the proportions of various T cell subsets and the expression of Fas and other molecules were normal. Also, the T cells from the thymus responded normally to activation through the TCR. These data suggested that AhR deficiency might not affect the T cell maturation or functions in the thymus. Thus, it is possible that the decreased size of the thymus in AhR KO mice may result from decreased homing of bone marrow stem cell precursors to the thymus, an aspect that needs further investigation. Our studies are also consistent with the previous findings that peripheral immune functions in AhR KO mice are normal (3). It should be noted that in an earlier study, we reported that TCDD increased the expression of Fas in the thymus in an AhR-dependent fashion (11). The Fas gene expressed DRE in its promoter and chromatin immunoprecipitation studies showed that the AhR complex binds to potential DREs in the Fas gene promoter (11). Thus, while up-regulation of Fas on T cells may further facilitate induction of apoptosis, our studies suggest that increased expression of FasL on thymic stromal cells is sufficient to induce enhanced apoptosis inasmuch as when Fas⁺ untreated thymic T cells were added to TCDD-treated stromal cells, the T cells underwent increased apoptosis. Fig. 12 shows a schematic representation of the mechanisms of up-regulation of Fas and FasL on thymic T cells and stromal cells leading to induction of apoptosis.

The current study made the novel observation that TCDD activates both the death receptor and mitochondrial pathways of apoptosis in thymocytes. Hepatocytes have been shown to resemble type II cell lines in which Fas-induced death is dependent on mitochondria through cross-talk (16). Thus, while wild-type mice injected with anti-Fas Abs die from hepatocellular apoptosis, Bid-deficient mice survive from such a challenge (16). In contrast, thymocytes may resemble type I cells inasmuch as they do not absolutely require Bid for Fas-induced death, although Bid deficiency delays apoptosis (16). Thus, in the current study, we investigated whether TCDD-induced apoptosis was Bid dependent. Although we noted that Bid was cleaved in TCDD-exposed thymocytes, it was dispensable inasmuch as Bid KO mice were as sensitive to TCDD-induced thymic atrophy and apoptosis as the wild-type mice. Thus, in TCDD-induced apoptosis of thymocytes, the death receptor and the mitochondrial pathways may be triggered independently. The coactivation of these two pathways offers an explanation as to why earlier studies found that Bcl-2 transgenic mice were sensitive to TCDD-induced thymic atrophy (41, 42).

Previous studies from a laboratory failed to detect apoptosis in thymocytes in vivo when tested for apoptosis immediately following their isolation (41, 42, 43). These studies are similar to our previous studies that freshly isolated thymocytes from TCDD-treated mice do not exhibit apoptosis due to rapid clearance of apoptotic cells in the thymus by phagocytic cells, in vivo (7). Detecting apoptosis can be improved by incubating freshly isolated thymocytes for 24 h in vitro because committed apoptotic cells can die without being removed by phagocytic cells (5, 7, 9, 10, 44). Also, it is necessary to use multiple approaches (5, 6, 7, 8, 11, 12, 44) and not just a single assay to detect apoptosis (41, 42, 43).

A recent study also demonstrated that TCDD treatment led to an increase in the percentage of thymocytes in the G₁ phase of the cell cycle and a significant decrease in the percentage of S + G₂/M thymocytes, especially in the CD4⁻CD8⁻CD3⁻ triple-negative intrathymic progenitor cell population (43). Although, clearly, multiple mechanisms of thymic atrophy induced by TCDD could be operating in vivo, it is also likely that the TCDD-induced cell cycle arrest can subsequently trigger activation of death pathways.

As seen from the current study, in addition to Fas and FasL, we believe that DRE-linked transcriptional targets of AhR may also include other molecules that regulate apoptosis in thymocytes through death receptors and/or the mitochondria. For instance, two potential DRE sequences were reported in the Bax promoter (45), although we discovered a third DRE site using MatInspector software (20). Our analysis also revealed that promoter sequences of Bad, Bid, and OPG genes contained potential DNA binding sequences for the Arnt/AhR heterodimeric complex (data not shown). Furthermore, our studies raised the possibility that Bax and Bid genes are directly regulated by TCDD-activated AhR as supported by increased levels of Bax and Bid proteins in TCDD-exposed thymocytes.

In the current study, we also noted using cDNA microarray analysis that several genes regulating apoptosis were induced in AhR^+/+ but not AhR KO mice, thereby suggesting that the induction of these genes may be regulated by activation through the AhR. However, it should be noted that the microarray analysis was performed at 24 h post-TCDD treatment and may have missed additional genes activated in the early or intermediate hours. In an earlier study, we found that, in the thymus, 23 of 37 apoptotic genes screened were up-regulated by TCDD by a factor of two or more when compared with the vehicle-treated controls (12). In the current study, although we used a different array for analysis consisting of 96 apoptotic genes, it was worth noting that several of the genes that were previously shown to be induced (12) were also found to be regulated by the AhR. Some of these molecules included Bad, Bcl-w, FasL, Caspase-1, and caspase and RIP adaptor with death domain.

The nature of cells sensitive to TCDD-induced toxicity and their role in thymic atrophy has been extensively debated. Some studies have suggested that TCDD-induced thymic atrophy mainly results from AhR/Arnt activation on hemopoietic cells and/or intrathymic progenitor cells (43, 46, 47), while others have proposed that TCDD may target thymic stromal cells (48, 49, 50, 51, 52), which in turn may contribute to thymic atrophy. However, the precise mechanisms through which alterations in stromal cells lead to an effect on T cell differentiation or thymic atrophy is unclear. It should be noted that thymic stromal cells consist of a phenotypically diverse group of cells, including epithelial cells, fibroblasts, macrophages, and dendritic cells (17). Thus, while some studies have suggested that TCDD may not affect thymic dendritic/macrophages (46), others have noted that TCDD may alter the morphology and functions of thymic epithelial cells (48, 49, 50, 51, 52). In a recent study, deficiency of ARNT protein in T cells but not in thymic epithelial cells was shown to play a critical role in TCDD-induced thymic atrophy (47). However, the latter findings were based on in vitro organ culture studies and ARNT deficiency in normal epithelial cells was found to have profound influence on the functions of epithelial cells (47), which could have affected the TCDD-induced toxicity. Based on the current and past studies (43, 46, 49), we suggest that TCDD exerts an AhR-dependent modulatory effect on both thymic T cells and stromal cells. In T cells, TCDD induces expression of Fas following activation through DREs found on the Fas gene promoter and makes them more sensitive to apoptosis (11). Moreover, TCDD enhances the expression of FasL on thymic stromal cells in an AhR-dependent manner possibly through NF-κB activation. These events are likely to induce enhanced apoptosis in T cells during T cell differentiation. For example, T cells differentiating in the thymus go through a positive selection process involving cell-cell interactions with cortical epithelial cells (53). Also, the T cells undergo negative selection mediated by contact through dendritic cells and medullary epithelial cells (53). We suggest that T cells, during such contacts with stromal cells, may undergo enhanced apoptosis due to TCDD-induced up-regulation of Fas and FasL on T cells and stromal cells, respectively.

The authors have no financial conflict of interest.