Abstract
The B-cell, a major cellular component of humoral immunity, has been identified as a sensitive target of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD). The actual molecular mechanism responsible for the immunotoxic effects produced by TCDD is unclear; however, many of the biological effects produced by TCDD are thought to be mediated by the aryl hydrocarbon receptor (AhR). Using the CH12.LX B-cell line, the present studies show that inhibition of μ gene expression and IgM protein secretion by polychlorinated dibenzo-p-dioxin congeners follow a structure-activity relationship for AhR binding. Furthermore, these effects may be mediated by the two dioxin-responsive enhancer (DRE)-like sites that were identified within the Ig heavy chain 3′α-enhancer. Electrophoretic mobility shift assay-Western analysis demonstrated TCDD-induced binding of the AhR nuclear complex to both DRE-like sites as well as TCDD-induced binding of several nuclear factor-κB/Rel proteins to a κB site, which overlaps one of the DRE-like sites. Interestingly, κB binding in the AhR-deficient BCL-1 B-cells was also induced by TCDD, demonstrating an AhR-independent effect of TCDD on κB binding. Taken together, these results support an AhR/DRE-mediated mechanism for TCDD-induced inhibition of IgM expression.
The immune system is a sensitive target organ of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), a persistent environmental contaminant. Immune suppression has been observed in virtually every species studied and occurs at doses that do not produce obvious signs of toxicity (for review, see Holsapple et al., 1991). Effects on various components of the immune system have been observed; however, the B-cell, a major cellular component of humoral immunity, has been identified by cell-type fractionation studies as a highly sensitive cellular target for the direct immunotoxic effects (i.e., inhibition of Ig secretion) of TCDD (Dooley and Holsapple, 1988).
The most extensively characterized mechanism thought to be responsible for the biological responses induced by TCDD involves transcriptional regulation through the aryl hydrocarbon receptor (AhR) (for review, seeRowlands and Gustafsson, 1997). The cytosolic AhR is complexed with heat shock protein-90 and other partially characterized proteins. Binding of ligand, such as TCDD, to the AhR results in translocation of the liganded AhR into the nucleus where it forms a heterodimer with the aryl hydrocarbon receptor nuclear translocator (ARNT). The TCDD-AhR/ARNT nuclear complex can act as a transcription factor by binding specific DNA motifs termed dioxin-responsive enhancers (DREs) in the promoter/enhancer regions of sensitive genes. This mechanism has been primarily characterized in hepatic tissue and hepatic cell lines through studies aimed at elucidating the mechanism for the induction of drug-metabolizing enzymes, such as cytochrome P450 1A1 (CYP1A1). A relationship between an up-regulation of metabolic enzymes and the immunotoxic effects produced by TCDD has yet to be identified. Indeed, several other genes such as those encoding plasminogen activator inhibitor-2, interleukin-1β, transforming growth factor-α and -β, epidermal growth factor receptor, estrogen receptor,c-fos, c-jun, and recombination-activating gene have been shown to be up-regulated or down-regulated after TCDD treatment (Astroff et al., 1990; Sutter et al., 1991; Gaido et al., 1992; Puga et al., 1992; Lu et al., 1994; Silverstone et al., 1994). The modulation of nonmetabolic genes such as those named above may account, in varying degrees, for the various toxicities observed in animals treated with TCDD. In addition, DRE-like sites have been found in the promoter regions of several of these genes, supporting the possibility for transcriptional regulation through the AhR, which would be analogous to CYP1A1 induction (Lai et al., 1996).
There is evidence for an AhR dependence of TCDD-mediated immune suppression based on earlier studies that used the following: 1) Ah high-responsive (Ahrbb) and Ah low-responsive (Ahrdd) mouse strains (Vecchi et al., 1983); 2) congenic mice at the Ah locus (Kerkvliet et al., 1990); and 3) structure-activity relationships (SARs) between various AhR ligands and inhibition of the plaque-forming antibody response (Davis and Safe, 1988). In addition, this laboratory has recently demonstrated an AhR-dependent inhibition of IgM secretion by TCDD using a cell line model consisting of AhR-expressing CH12.LX B cells and AhR-deficient BCL-1 B-cells (Sulentic et al., 1998). The AhR may directly regulate transcription of immunological genes important to B-cell activation and differentiation. Alternatively, the AhR may mediate inhibition of IgM secretion through an interaction with signaling pathways of other proteins such as c-Src kinase, nuclear factor-κB (NF-κB), Sp1, transcription factor IIB, and retinoblastoma protein, all of which have been shown to directly associate with the AhR (Enan and Matsumura, 1996; Kobayashi et al., 1996; Ge and Elferink, 1998; Swanson and Yang, 1998; Tian et al., 1999).
In the present studies, we describe the identification of DRE-like sites within the Ig heavy chain 3′α-enhancer to which TCDD induced AhR binding. The 3′α-enhancer is composed of four functional enhancer domains [Cα3′E, 3′αE(hs1,2), hs3, and 3′α-hs4]. It has been suggested that these enhancers form a locus control region because they can act synergistically to regulate the expression of the Ig heavy chain genes, μ, δ, γ, ε, and α, which encode heavy chain proteins for IgM, IgD, IgG, IgE, and IgA, respectively (Pettersson et al., 1997). Inhibition by TCDD of IgM secretion may be, at least in part, due to an AhR/DRE-mediated effect on μ expression, perhaps through an effect at the 3′α-enhancer. In this report, we further characterize and support our previous observation in the CH12.LX/BCL-1 cell line model of AhR dependence for TCDD-mediated immune suppression. We demonstrate that μ gene expression is inhibited by TCDD and that this inhibition as well as inhibition of IgM protein secretion follows an SAR for AhR binding. Consistent with a common AhR-mediated mechanism of action, there was a general concordance between the IC50 values for inhibition of μ expression and IgM secretion and the EC50 for induction of CYP1A1 expression. In addition, two DRE-like binding sites within the 3′α-enhancer demonstrate TCDD-inducible binding of the AhR nuclear complex, implicating a direct transcriptional effect on μ expression. Interestingly, TCDD also induced NF-κB/Rel protein binding to a full κB site within the 3′α-enhancer. Although NF-κB/Rel proteins have been shown to be transcriptional regulators of this enhancer, the induction of κB binding also occurred with the AhR-deficient BCL-1 cells.
Materials and Methods
Chemicals.
TCDD, 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin (HxCDD), 2,3,7-trichlorodibenzo-p-dioxin (TriCDD), and 1-monochlorodibenzo-p-dioxin (MCDD), in 100% dimethyl sulfoxide (DMSO), were purchased from AccuStandard (New Haven, CT). The certificate of product analysis stated the purity of TCDD, HxCDD, TriCDD, and MCDD to be 99.1, 100, 99.6, and 100%, respectively, as determined by AccuStandard using gas chromatography/mass spectrometry. DMSO and lipopolysaccharide (LPS) were purchased from Sigma (St. Louis, MO).
Cell Lines
The CH12.LX B-cell line derived from the murine CH12 B-cell lymphoma, which arose in B10.H-2aH-4bp/Wts mice (B10.A × B10.129), has been previously characterized (Bishop and Haughton, 1986) and was a generous gift from Dr. Geoffrey Haughton (University of North Carolina, Research Triangle Park, NC). The BCL-1 B-cell line was derived from a murine B-cell lymphoma that spontaneously arose in a BALB/c mouse (Slavin and Strober, 1978). This cell line has been previously characterized (Gronowicz et al., 1980) and was generously provided by Dr. Kathryn H. Brooks (Michigan State University, East Lansing). The CH12.LX and BCL-1 cell lines were maintained as previously described (Sulentic et al., 1998).
Quantitative Reverse Transcription-Polymerase Chain Reaction (RT-PCR).
Total RNA from each sample was isolated using the High Pure RNA isolation system (Boehringer Mannheim, Indianapolis, IN). RNA samples were quantified by spectrophotometry and then analyzed for DNA contamination by PCR analysis without reverse transcriptase. RNA samples containing DNA were incubated with RNase-free DNase as previously described (Williams et al., 1996). Quantitative RT-PCR was performed as outlined in Gilliland et al. (1990a,b), except that the recombinant RNA was used as an internal standard (IS) instead of genomic DNA. ISs were generated as previously described (Vanden Heuvel et al., 1993) and contain specific PCR primer sequences for CYP1A1 or μ and a spacer gene of rat β-actin or rat β-globin, respectively. Total DNA-free RNA (100 ng) and IS (recombinant RNA) were reverse transcribed simultaneously, in the same reaction tube, into cDNA using oligo(dT)15 as primers. The forward and reverse primers for μ were TGAGCAACTGAACCTGAGG and TGCATACACAGAGCAACTG, respectively. Final reaction concentrations for the PCR reaction were 3 mM MgCl2 and 2.5 U Taq DNA polymerase (Promega, Madison, WI). Samples were cycled 30 times with each cycle consisting of 94°C for 15 s, 60°C for 30 s, and 72°C for 45 s. Primers for the CYP1A1 gene were a generous gift from Dr. Dale Morris (G. D. Searle, St. Louis, MO). The CYP1A1 PCR reaction was performed as described above for the μ PCR reaction, except that the MgCl2 concentration was 4 mM, the annealing temperature was 56°C, and the samples were cycled 32 times. PCR products were visualized by ethidium bromide staining and quantitation was performed by assessing the optical density for both the target and IS DNA using a Gel Doc 1000 video imaging system (Bio-Rad, Hercules, CA). The number of transcripts was calculated from a standard curve generated from the density ratio between the gene of interest and a specific IS concentration.
Enzyme-Linked Immunosorbent Assay (ELISA).
Supernatants were harvested from naive or LPS (3 or 30 μg/ml)-stimulated CH12.LX or BCL-1 cells after a 72-h incubation at 37°C in 5% CO2 and were analyzed for IgM by sandwich ELISA as described previously (Sulentic et al., 1998). Colorimetric detection was performed over a 1-h period using an EL808 automated microplate reader with a 405-nm filter (Bio-Tek, Winooski, VT). The DeltaSoft 3 computer analysis program (BioMetallics, Princeton, NJ) calculated the concentration of IgM in each sample from a standard curve generated from the absorbance readings of known IgM concentrations.
Nuclear Protein Preparation.
CH12.LX or BCL-1 cells were incubated with DMSO (0.01%) or 30 nM TCDD in DMSO for 1 h at 37°C. Cells were harvested by centrifugation at 1200 rpm for 10 min, washed once with 1× PBS, and then incubated in 10 mM HEPES (pH 7.5) for 5 min on ice, and centrifuged at 1200 rpm for 5 min. One milliliter of MDH/LAP [3 mM MgCl2, 1 mM dithiothreitol (DTT), 25 mM HEPES, 100 μM leupeptin, 40 U/ml aprotinin, and 200 μM phenylmethylsulfonyl fluoride (PMSF)] was added to the cell pellet and homogenized with a tight-fitting pestle. Nuclei were pelleted by centrifuging at 1,000g for 5 min, washed twice with MDHK/LAP (3 mM MgCl2, 1 mM DTT, 25 mM HEPES, 100 mM KCl, 100 μM leupeptin, 40 U/ml aprotinin, and 200 μM PMSF), and then resuspended in 100 μl of HEDGK/LAP (25 mM HEPES, 1 mM EDTA, 1 mM DTT, 10% glycerol, 400 mM KCl, 100 μM leupeptin, 40 U/ml aprotinin, and 200 μM PMSF), incubated on ice with agitation for 40 min, and centrifuged at 14,000g for 15 min. The supernatant was aliquoted and stored at −80°C before use in the electrophoretic mobility shift assay (EMSA). Protein concentrations were determined using the bicinchoninic acid protein determination assay (Sigma).
Synthetic DRE Oligonucleotides.
Complementary pairs of synthetic DNA fragments corresponding to the AhR/ARNT binding site of mouse DRE3 (Denison and Yao, 1991) to two putative DRE sites in the mouse Ig 3′α-enhancer and to the κB site from the 3′α-hs4 enhancer were synthesized using an Applied Biosystems DNA synthesizer. The DRE oligonucleotides were purified by HPLC (Macromolecular Structure Facility, Michigan State University), annealed, and end labeled using T4 polynucleotide kinase (Boehringer Mannheim) and [γ-32P]ATP (DuPont NEN, Boston, MA). The oligonucleotide sequences (consensus nucleotides are underlined) for the 3′α-enhancer and κB site are as follows: 3′αE(hs1,2), TAGGGGTCTATTAACTCACCACGCTAGGCCATCATGGAGAG, positions 1096 to 1136, GenBank accession no. X62778 (Dariavach et al., 1991); 3′α-hs4, AGCAGAGG GGGGGACTGGCGTGGAAAGCCCCATTCACCCAT, position 319 to 360, GenBank accession no. L39932 (Michaelson et al., 1995); and hs4-κB, GATCTCTCTGGAAAGCCCCTCTGA, GenBank accession no.L39932 (Michaelson et al., 1995).
EMSA.
Nuclear protein preparations were used in the EMSA as previously described (Reyes et al., 1992; Probst et al., 1993) with a few modifications. Briefly, nuclear extracts (10 μg of protein) were incubated with poly(dI-dC) (Boehringer Mannheim) at room temperature for 15 min. Radiolabeled DRE oligomer was added (40,000 cpm) and incubated at room temperature for another 30 min. The binding of protein to the DNA was resolved by a 4.0% nondenaturing polyacrylamide gel electrophoresis (PAGE), dried on 3-mm filter paper (Whatman, Hillsboro, OR), and autoradiographed. Final reaction concentrations were as follows: 25 mM HEPES (pH 7.5), 1 mM EDTA, 2 mM DTT, 10% glycerol, 108 mM KCl, and 1.0 μg poly(dI-dC). Where indicated, a 100-fold excess of unlabeled DRE3, 3′αE(hs1,2), 3′α-hs4, or hs4-κB oligomer was added to the reaction.
EMSA-Western Analysis.
EMSA analysis was conducted as described above but included samples containing 10 pmol of unlabeled DRE3, 3′αE(hs1,2), or 3′α-hs4 instead of radiolabeled oligomers. The nonradiolabeled portion of the EMSA gel was separated from the radiolabeled portion and transferred to nitrocellulose (Amersham, Arlington Heights, IL). Protein-DNA blots were blocked in Blotto buffer (1% low fat dry milk in 0.1% Tween 20 Tris-buffered saline) for 1 to 2 h at 22°C. Primary antibody to the AhR (17-10B), previously characterized by Pollenz et al. (1994), was a generous gift of Dr. Richard S. Pollenz (Medical University of South Carolina, Charleston, SC). The ARNT antibody (NB 100-110) was purchased from Novus Biological (Littleton, CO). p65 (RelA), p50, RelB, and c-rel antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Immunochemical staining was performed as previously described (Sulentic et al., 1998). Optical density for the protein of interest was measured by densitometry using a model 700 imaging system (Bio-Rad).
Statistical Analysis of Data.
Mean ± S.E. was determined for each treatment group of a given experiment. Statistical difference between treatment groups and the vehicle controls was determined by Dunnett's two-tailed t test. For EC50 and IC50 generation, a complete concentration-response curve for TCDD, HxCDD, and TriCDD was obtained for a given endpoint. Concentration-response curves were fit by a four-parameter logistic concentration-response equation given as Y = [A1 − A2)/[1 + (X − X0)P]] + A2. The derived parameter, EC50 or IC50(X0, concentration generating half-maximal response) was expressed as nanomolar (nM) concentration (mean ± S.E.). Statistical difference between EC50 and IC50 means was determined by a one-way ANOVA followed by a least-significant difference test. P < .05 was considered statistically significant.
Results
Polychlorinated Dibenzo-p-dioxin (PCDD)-Mediated Inhibition of LPS-Induced IgM Secretion in CH12.LX B-Cells Follows an SAR That Is Concordant with AhR Ligand-Binding Affinity and CYP1A1 Induction.
Transcriptional regulation through an AhR/DRE mechanism of TCDD-induced CYP1A1 expression is well established (Whitlock, 1990). For the following specific PCDD congeners, previous reports have determined the rank order potency for AhR-binding affinity and AhR-dependent induction of CYP1A1 as TCDD > HxCDD > TriCDD ≫ MCDD, with the MCDD congener having no affinity for the AhR and being unable to induce ethoxyresorufin O-deethylation activity (Poland and Glover, 1976; Poland et al., 1979). The binding affinities (Kd) for TCDD, HxCDD, and TriCDD are 0.27, 0.77, and 1.92 nM, respectively (Poland and Glover, 1976; Dr. Alan Poland, personal communication, Oct. 22, 1998). Likewise, induction of CYP1A1 in the CH12.LX cells after a 24-h incubation with the PCDD congeners is concentration-dependent, as determined by quantitative RT-PCR, and correlated with AhR-binding affinity (Fig.1). Although there was no statistical difference observed between the EC50 values for the HxCDD and TriCDD congeners, the general trend for the rank order potency was TCDD > HxCDD > TriCDD ≫ MCDD; MCDD had no effect on CYP1A1 expression (Table 1). To further characterize the AhR dependence of TCDD-induced IgM secretion, the effect of these congeners on LPS-induced IgM protein secretion from the CH12.LX B-cell line was examined by ELISA. Inhibition of IgM secretion was concentration-dependent and correlated with AhR-binding affinity (Fig. 2). Similar to the results for CYP1A1 induction, the rank order potency was TCDD > HxCDD > TriCDD ≫ MCDD; again, MCDD had no affect on IgM secretion (Table 1).
Specific PCDD Congeners Have No Affect on CYP1A1 Expression or LPS-Induced IgM Secretion from the AhR-Deficient BCL-1 B Cells.
In agreement with a lack of AhR expression, TCDD, HxCDD, TriCDD, and MCDD had no affect on CYP1A1 expression in BCL-1 cells as determined by qualitative RT-PCR (Fig. 3). To confirm the AhR dependence of congener-induced inhibition of IgM secretion, the effect of specific PCDD congeners on LPS-induced IgM secretion from the AhR-deficient BCL-1 cell line was evaluated by ELISA. TCDD, TriCDD, and MCDD had no affect on IgM secretion (Fig.4). In contrast, the HxCDD congener significantly enhanced IgM secretion (Fig. 4); however, this stimulatory effect was not consistently observed.
LPS-Induced μ Expression in CH12.LX Cells Is Inhibited by PCDD Congeners and Follows an SAR for AhR Binding.
Because IgM is composed of two heavy chains and two light chains, the genes encoding these proteins are potential transcriptional targets modulated by TCDD. To determine whether transcriptional regulation of the μ gene underlies the inhibition of IgM secretion by TCDD, expression of the μ gene in LPS-stimulated CH12.LX cells was analyzed by quantitative RT-PCR analysis after a 24-h treatment with TCDD, HxCDD, and TriCDD at various concentrations. Although TCDD and HxCDD inhibited μ gene expression in a concentration-dependent manner, the TriCDD congener that induces CYP1A1 and inhibits IgM secretion did not affect μ gene expression (Fig. 5; Table 1). In addition the effect of HxCDD exhibited a rather flat concentration response that is in contrast with its effect on CYP1A1 induction (compare Figs. 1 and5). The lack of effect by TriCDD and the blunted response of HxCDD may be due to slower kinetics of AhR transformation by lower-affinity AhR ligands. This potential effect on kinetics might be more pronounced in the shorter, 24-h μ expression assay as opposed to the longer, 72-h IgM protein secretion assay. In contrast to μ expression, CYP1A1 induction, although a 24-h assay, is extremely sensitive to TCDD, HxCDD, and TriCDD (Fig. 1). This sensitivity is likely due to the presence of six DREs in the CYP1A1 promoter; five of which are capable of positively regulating transcription as demonstrated by reporter gene assays (Lusska et al., 1993). Furthermore, the presence and functionality of DREs within critical regulatory regions of the μ gene are unclear. To explore the possibility of slower kinetics for the effects of TriCDD and HxCDD on μ expression, LPS-stimulated CH12.LX cells were incubated with the PCDD congeners for 48 h followed by quantitative RT-PCR analysis. In contrast to the 24-h results, μ expression was inhibited in a concentration-dependent manner by TriCDD and the concentration response for HxCDD was sigmoidal (Fig.6). Although there was no statistical difference between the IC50 values for the TCDD and HxCDD congeners, the general trend for the rank order potency was TCDD > HxCDD > TriCDD ≫ MCDD; MCDD had no affect on μ expression (Table 1; Fig. 6). The effect of the PCDD congeners on LPS-induced μ expression in the AhR-deficient BCL-1 cells was also analyzed at 48 h by quantitative RT-PCR analysis. All of the congeners had no effect on μ gene expression from the BCL-1 cells (Fig. 4).
Inhibition of IgM Protein Secretion and μ Expression Is AhR-Dependent.
IC50 and EC50 values were generated from extensive concentration-response curves (i.e., at least nine concentrations per congener) for each congener (Table 1). An abbreviated version of these curves is represented in Figs. 1 to 6. For a given congener, statistical comparisons of the IC50 values for μ expression (48 h) and IgM protein secretion and the EC50 for induction of CYP1A1 expression were not significantly different with the exception of a slight difference between CYP1A1 induction and μ (48 h) inhibition with the TriCDD congener (Table 1). These results suggest a common mechanism of action and because induction of CYP1A1 is an established AhR-mediated event, these results continue to support AhR-mediated inhibition of μ expression and IgM protein secretion. In addition, the IC50 and EC50 values for a given endpoint among the PCDD congeners tended toward an SAR that was concordant with the AhR-binding affinity for the respective PCDD congeners, again supporting AhR mediation of these three responses (Table 1).
TCDD Induces AhR Binding, with the CH12.LX Cells, to a DRE-Like Site Located within the 3′αE(hs1,2) and 3′α-hs4 Enhancers of the Ig Heavy Chain.
We have identified several DRE-like sequences in the 3′α-enhancer of the mouse Ig heavy chain gene. Our studies focused on two of the DRE-like sites; one of which is located in the 3′αE(hs1,2) enhancer and the other in the 3′α-hs4 enhancer. Of the four 3′α-enhancers, these two enhancers have much stronger transcriptional activity than the Cα3′E and hs3 enhancers (Chauveau et al., 1998). In the CH12.LX cells, EMSA analysis demonstrated TCDD-inducible binding, which migrated similarly to the DRE3 positive control, to both the 3′αE(hs1,2) and the 3′α-hs4 oligomers (Figs.7A and 8A, lanes 2, 4, and 5). Binding to these oligomers was also reduced with the addition of unlabeled DRE3, although not as effectively as with the unlabeled oligomers themselves (Figs. 7A and 8A, lanes 5–7). These results suggest that the AhR nuclear complex binds to both DRE-like sites identified within the 3′αE(hs1,2) and 3′α-hs4 oligomers, which was confirmed by EMSA-Western analysis. Antibodies specific for the AhR and ARNT identified these proteins as components of the TCDD-inducible complex in both the 3′αE(hs1,2) and 3′α-hs4 oligomers, as well as in the DRE3 positive control (compare Figs. 7 and8, B and C, lanes 2 and 4). The AhR and ARNT migrated identically among the oligomers in the EMSA-Western analysis (Figs. 7 and 8, B and C, lanes 2 and 4), as well as with the TCDD-inducible protein complexes formed with both oligomers in the EMSA (Figs. 7 and 8, compare lane 2 and 5 of A to lanes 2 and 4 of B and C). However, it is notable that in the EMSA, the TCDD-inducible complex formed with the 3′αE(hs1,2) oligomer is rather broad and diffuse, suggesting multiple protein-DNA complexes (Fig. 7A, lane 5). In contrast, the EMSA-Western analysis identified a sharp band containing both the AhR and ARNT that was part of the TCDD-inducible band detected in the EMSA (compare Fig. 7A, lane 5, and B and C, lane 4). TCDD also induced the binding of a second protein complex to the 3′α-hs4 oligomer that does not contain the AhR (compare Fig. 8A, lane 5, and B, lane 4). Nuclear proteins other than the AhR nuclear complex appear to be induced by TCDD to bind to both the 3′αE(hs1,2) and 3′α-hs4 oligomers.
TCDD Does Not Induce AhR Binding to DRE-Like Sites within the 3′αE(hs1,2) and 3′α-hs4 Enhancers with the BCL-1 Cells.
EMSA-Western analysis of nuclear protein isolated from the AhR-deficient BCL-1 cells was also performed to determine whether protein-DNA complexes are formed with the DRE3, 3′αE(hs1,2), and 3′α-hs4 oligomers. With the DRE3 oligomer, faint TCDD-inducible complexes are formed; however, these complexes do not contain the AhR or ARNT (Figs. 7 and 8, compare lane 2 of A–C). Similar to the CH12.LX cells, a broad TCDD-inducible complex was formed with the 3′αE(hs1,2) oligomer (Fig. 7A, lane 5), again suggesting multiple protein-DNA complexes. In contrast, the TCDD-inducible 3′αE(hs1,2) protein-binding complex identified in the BCL-1 cells did not contain the AhR or ARNT (Fig. 7A, lane 5, compare with B and C, lane 4) and is of a lower molecular weight as evidenced by its greater migration compared with that observed in the CH12.LX cells (Fig. 7A, compare lanes 5). Interestingly, the unlabeled DRE3 oligomer competed for TCDD-inducible binding to the labeled 3′αE(hs1,2) oligomer (Fig. 7A, compare lanes 5 and 7), perhaps suggesting that proteins other than the AhR and ARNT are capable of binding to the DRE. TCDD treatment also induced binding of a protein complex to the 3′α-hs4 oligomer that migrated similarly to the lower complex observed with the CH12.LX cells (Fig. 8A, compare lanes 4 and 5 of CH12.LX and BCL-1). Neither of these complexes contained the AhR or ARNT (Fig. 8, B and C, lanes 4). Collectively, these results suggest that TCDD at concentrations that induce AhR activation also induce additional protein-binding complexes that recognize both of these oligomers in an AhR-independent manner.
TCDD Induces Binding to a κB Site Located within the 3′α-hs4 Enhancer with the CH12.LX Cells.
Our EMSA-Western results suggest that nuclear proteins distinct from the AhR nuclear complex are induced by TCDD to bind the 3′αE(hs1,2) and 3′α-hs4 oligomers in both the CH12.LX and BCL-1 cells. Interestingly, the 3′α-hs4 oligomer possesses a full κB motif that overlaps the DRE site. The potential significance concerning the overlap of these two motifs is 3-fold. First, it is well established that NF-κB/Rel proteins are important regulators of the 3′α-enhancer (Michaelson et al., 1996; Chauveau et al., 1998), suggesting that the NF-κB/Rel proteins are likely candidates for the additional TCDD-induced protein-3′α-hs4 complexes being observed. Second, an increase in κB binding after TCDD treatment has been recently demonstrated by EMSA (Ashida and Matsumura, 1998; Barnes et al., 1999). Third, the AhR/ARNT complex may compete with NF-κB/Rel proteins for binding. Therefore, antibodies specific for the NF-κB/Rel family members, p65, p50, c-Rel, and RelB, were used to probe EMSA-Western blots for 3′α-hs4 in an attempt to identify the binding proteins induced by TCDD in the CH12.LX and BCL-1 cells. With the CH12.LX cells, analysis revealed TCDD-induced binding to the 3′α-hs4 oligomer of p65, RelB, p50, and c-Rel, all of which were identified in the lower protein-DNA complex (Fig.9, compare lanes 2 of A and E with lanes 2 of C, D, G, and H). This complex does not contain the AhR or ARNT but may be composed of several protein-DNA complexes of various hetero- and homodimers of p65, RelB, p50, and c-Rel. For instance, there are modest differences in migration between the NF-κB/Rel proteins (i.e., p65 and RelB) and, in some instances, between vehicle and TCDD treatment (i.e., p65 and c-Rel); however, all NF-κB/Rel proteins identified in the EMSA-Western analysis migrated within the TCDD-inducible band detected in the EMSA (Fig. 9, lane 2, compare A and E with C, D, G, and H). Interestingly, a faint upper band was also identified with the RelB antibody (Fig. 9D, lanes 1 and 2) that migrated slightly higher than the AhR (Fig. 9, lane 2, compare B and D) and therefore, is not likely to be a component of the AhR/ARNT complex. Because RelB does not form homodimers and because none of the other κB proteins exhibit a similar migration pattern to RelB, its binding partner is unclear. It is notable that NF-κB/Rel family members can form “cross-family” dimers that might explain our inability to detect other NF-κB/Rel family members in association with the slower migrating RelB-DNA complex. Interestingly, an unlabeled κB oligomer containing the identical κB sequence from the 3′α-hs4 enhancer completely abrogated binding of the lower complex, suggesting specificity of this sequence for κB binding (Fig. 9A, lane 5). Unlabeled κB also decreased TCDD-induced binding of the AhR/ARNT complex although not completely. This effect may be related to the much stronger binding of the NF-κB/Rel proteins as well as the possibility that the excess unlabeled κB oligomer may nonspecifically affect DRE binding. Because the κB binding site overlaps the flanking region and one nucleotide of the core region of the DRE-binding site and because the unlabeled κB oligomer does not contain the complete DRE site, it is unlikely that the AhR nuclear complex would bind directly to the unlabeled κB oligomer.
TCDD Induces Binding to a κB Site Located within the 3′α-hs4 Enhancer with the BCL-1 Cells.
For the BCL-1 cells the κB binding to the 3′α-hs4 oligomer was modestly different compared with the CH12.LX cells. Unlike the CH12.LX cells, there was relatively little basal κB binding as demonstrated by EMSA analysis (Fig. 9A, lane 1, compare with Fig. 10A, lane 1). The EMSA-Western analysis indicated that c-Rel was the only NF-κB/Rel protein with detectable binding in the vehicle treatment group (Fig.10H, lane 1). TCDD treatment of the BCL-1 cells induced 3′α-hs4 binding of p65, RelB, and c-Rel; these proteins migrated similarly to the single TCDD-inducible protein binding complex (Fig. 10, lane 2, compare A and E with C, D, and H). In contrast to the CH12.LX cells, p50 was not induced by TCDD (Fig. 10G). Slight differences in migration between the NF-κB/Rel proteins (i.e., p65 and RelB) were identified (Fig. 10) as demonstrated with the CH12.LX cells. Additionally, binding to the 3′α-hs4 oligomer was also completely abrogated by the unlabeled κB oligomer (Fig. 10A, lane 5). Induction of DNA binding by NF-κB/Rel proteins after TCDD treatment in the AhR-deficient BCL-1 cells suggests an AhR-independent activation of p65, RelB, and c-Rel.
Discussion
The toxicity of TCDD is thought to be mediated transcriptionally through an interaction of the AhR/ARNT nuclear complex with DREs in regulatory regions of TCDD-sensitive genes. Putative DREs have been identified in regulatory regions of several nonmetabolic genes, supporting a potential for aberrant gene expression induced by TCDD through DRE-dependent transcriptional regulation (Lai et al., 1996); nevertheless, this type of regulation has been most extensively characterized for the induction of metabolic enzymes and has not been directly correlated with TCDD-induced toxicity, including immune suppression. However, considerable evidence does exist that supports an essential role for the AhR in TCDD-mediated immune suppression (Vecchi et al., 1983; Davis and Safe, 1988; Kerkvliet et al., 1990; Sulentic et al., 1998). In addition, several laboratories have demonstrated novel protein-protein associations involving the AhR and several proteins, other than ARNT, including NF-κB, Sp1, transcription factor IIB, retinoblastoma protein, and the Src family kinase c-src (Enan and Matsumura, 1996; Kobayashi et al., 1996; Ge and Elferink, 1998; Swanson and Yang, 1998; Tian et al., 1999). This suggests a possible interaction of the AhR with different signaling pathways and thus, a potential for DRE-independent mediation of some of TCDD's effects.
In the present studies, we have demonstrated an SAR in the AhR-expressing CH12.LX cells between AhR-binding affinity and three different endpoints: CYP1A1 induction, IgM protein secretion, and μ gene expression. In addition, the PCDD congeners had no effect on these endpoints in the AhR-deficient BCL-1 cells. These results extend our previous observation that the AhR is obligatory for the effects of TCDD on IgM modulation (Sulentic et al., 1998) but more importantly, a comparison of the EC50 and IC50 values of each congener for a particular endpoint provides pharmacological evidence for or against a common receptor-mediated mechanism. Interestingly, the EC50 and IC50 values for each endpoint were not significantly different for each congener, suggesting that inhibition of IgM protein secretion is a result of an AhR-dependent mechanism that involves at least in part an inhibition of μ expression.
The difference in potency between TCDD and HxCDD for CYP1A1 induction (0.11) and inhibition of IgM secretion (0.19) is very similar; however, the difference in potency for inhibition of μ expression (0.4) is less than that for the other two endpoints. This may be attributed to differences in the affinity of the AhR-nuclear complex for DRE-like sites within various target genes; the number of DRE sites within these genes may also be an important factor in the magnitude of transcriptional regulation mediated through these response elements. Additionally, AhR-mediated effects may occur independently of binding to the DRE, perhaps, as mentioned above, through an interaction with other cellular signaling pathways. In addition, Poland and Glover (1976) discovered that the AhR-binding affinity of TriCDD did not correlate with its biological potency. AhR binding (0.14 of TCDD) was much greater than the relative ability of TriCDD to induce hepatic hydroxylase activity in the chick embryo (0.0006 of TCDD). Poland and Glover (1976) did not know the reason for this discrepancy but suggested a potential metabolic inactivation of TriCDD in vivo. This is in contrast to our results with the CH12.LX cells, in which the difference in potency between TCDD and TriCDD for CYP1A1 induction and inhibition of IgM secretion and μ expression was 0.07, 0.05, and 0.06, respectively. This discrepancy in potency of the TriCDD congener may be due to differing kinetics of these responses as shown for μ expression and/or to differences in metabolism, as suggested by Poland and Glover (1976), which is consistent with the low drug-metabolizing capability of lymphocytes. In any case, our results demonstrate a good correlation in the difference of potency between TCDD and TriCDD for each endpoint.
The identification of several putative DREs within the 3′α-enhancer of the mouse Ig heavy chain gene supports a potential role for DRE-dependent transcriptional regulation of the μ gene. The 3′α-enhancer is composed of four enhancer domains [Cα3′E, 3′αE(hs1,2), hs3, and 3′α-hs4], which form a locus control region that appears to regulate high-level Ig production and Ig heavy chain class switching (Pettersson et al., 1997). In addition, Dariavach et al. (1991) demonstrated an approximately 2-fold increase in μ expression, as measured by a ribonuclease protection assay, following transfection of a plasmacytoma with only the 3′αE(hs1,2) enhancer region. Addition of the other enhancer domains might result in a more profound effect on μ expression that would be consistent with the results of Chauveau et al. (1998), demonstrating the greatest transcriptional activity in a gene reporter assay with the entire 3′α-enhancer. We have identified a TCDD-inducible protein-DNA complex by EMSA analysis with the 3′αE(hs1,2) and 3′α-hs4 oligomers; each of these oligomers contains a DRE-like site. Addition of unlabeled DRE3 competed for the TCDD-inducible complex, suggesting that the protein complex is composed of AhR and ARNT. Further analysis by EMSA-Western analysis demonstrated that the AhR and ARNT have affinity for the 3′αE(hs1,2) and 3′α-hs4 DREs and lends further support for the premise of DRE-dependent regulation of μ gene expression. DRE binding of the AhR and ARNT may directly inhibit μ expression or may inappropriately activate the 3′α-enhancer. For example, the 3′α-enhancer appears to regulate isotype class switching and previous studies have demonstrated a decrease in the IgG2a and general IgG response to specific antigens after TCDD treatment (Harper et al., 1994; Kerkvliet et al., 1996). TCDD-induced DRE binding to the 3′α-hs4 and 3′αE(hs1,2) enhancers may induce an incomplete or inappropriate stimulus toward class switching (i.e., DNA recombination without replication or transcription), resulting in decreased μ gene expression.
Michaelson et al. (1996) have demonstrated the regulation of both the 3′αE(hs1,2) and 3′α-hs4 enhancers by κB-binding proteins. Interestingly, the 3′α-hs4 oligomer used in our studies contains a full κB binding site that overlaps the DRE-binding motif. EMSA-Western analysis has identified TCDD-inducible protein binding, which does not include the AhR, to this κB site. Specifically, the second TCDD-inducible binding complex contained four NF-κB/Rel family members, p65, p50, c-Rel, and RelB. Although the mechanism responsible for the TCDD-induced increase in κB binding is unknown, this increase is in general agreement with similar observations made by other laboratories (Yao et al., 1995; Gollapudi et al., 1996; Ashida and Matsumura, 1998; Barnes et al., 1999). Puga and coworkers have suggested that the increase in κB binding is due to TCDD-mediated oxidative stress, which is a known inducer of NF-κB/Rel proteins (Yao et al., 1995; Ashida and Matsumura, 1998). In contrast, Tian et al. (1999) demonstrated a decrease in κB binding with TCDD treatment and also observed an association between the AhR and p65. The reason for this discrepancy is unclear but may be related to a difference in cell culture systems. Tian et al. (1999) conducted their studies with the mouse hepatoma cell line Hepa1c1c7 and NF-κB/Rel proteins would presumably have different regulatory roles in a hepatoma cell line, which does not express Ig genes versus a B-cell line. Interestingly, κB binding was also induced by TCDD with the AhR-deficient BCL-1 cells, suggesting an AhR-independent induction of at least p65, RelB, and c-Rel. Unlike the CH12.LX cells, p50 was not induced with the BCL-1 cells. However, the EMSA and EMSA-Western analyses were conducted using unstimulated cells and the effect of LPS stimulation in the presence or absence of TCDD may be very similar to the TCDD effect seen in these BCL-1 experiments. In addition, the inhibition of IgM secretion by TCDD may require the induction of both DRE and κB binding. Because these sites overlap in the DRE-flanking region, binding to either the DRE or κB motif may facilitate or stabilize binding to the other motif. Another possibility is that there may be competition between the AhR/ARNT and NF-κB/Rel protein complexes for binding to these overlapping motifs that would not be ascertained in these experiments because the labeled oligomer is in excess to the nuclear protein. Furthermore, because NF-κB/Rel proteins bind DNA as dimers, they are capable of complexing in a variety of combinations and each combination may have distinct transcriptional functions. TCDD may affect the formation of these dimers differently in the CH12.LX and BCL-1 cells (as seen with p50 and Rel B binding), perhaps resulting in an altered sensitivity of IgM expression between these cell lines. Alternatively, induction of NF-κB/Rel proteins by TCDD in the BCL-1 cells and normal LPS-induced IgM secretion in the presence of TCDD might suggest that activation of these proteins is not critical to the inhibition of IgM expression.
We have also identified a κB binding site within proximity to the DRE site within the 3′αE(hs1,2) enhancer. However, the 3′αE(hs1,2) oligomer used in the present experiments only contains half of this κB motif. EMSA analysis did detect a rather diffuse TCDD-inducible band; although, the AhR was only present in a discrete region of this band. Therefore, it is presently unclear whether this banding pattern is a result of NF-κB/Rel proteins binding to the κB half-site or due to the binding of other proteins to other regions of the 3′αE(hs1,2) oligomer. Future studies using an oligomer containing the entire κB response element will be required to adequately characterize whether NF-κB proteins are a component of these TCDD-inducible protein-DNA complexes.
The present studies demonstrate an essential role by the AhR in the inhibition of IgM secretion by PCDDs. However, more importantly, these results demonstrate TCDD-inducible binding of the AhR and κB proteins to sites within the 3′α-enhancer, thus providing the first direct, putative link between TCDD-mediated transcriptional regulation and TCDD-induced inhibition of μ gene expression. In addition, the magnitude of inhibition for μ expression (70–80%) is concordant with the magnitude of inhibition for IgM protein secretion (approximately 60–70%). It will be interesting to characterize the interaction of TCDD-induced DRE and κB binding and to determine their effect on transcriptional regulation of the 3′α-enhancer, especially because the induction of κB binding by TCDD appears to be AhR independent, which may imply that κB binding is not critical to the inhibition of μ expression by TCDD. However, an induction of κB binding may have profound effects on the vast number of genes that are regulated by NF-κB/Rel proteins.
Acknowledgments
We thank Dr. Richard Pollenz for providing the anti-AhR antibody, Dr. Kathryn Brooks for the BCL-1 cells, Dr. Geoffrey Haughton for the gift of the CH12.LX cells, Drs. Kurunthacha Kannan and John Giesy for analyzing the purity of our initial PCDD congeners, and Drs. Gregory Fink and Ronald Johnson for statistical help and expertise.
Footnotes
-
Send reprint requests to: Norbert E. Kaminski, Department of Pharmacology and Toxicology, 315 Food Safety and Toxicology, Michigan State University, East Lansing, MI 48824. E-mail:kamins11{at}msu.edu
-
↵1 This work was supported in part by funds from the National Institute of Environmental Health Sciences Grant ES02520 and Superfund Grant P01 P42ES04911.
- Abbreviations:
- TCDD
- 2,3,7,8-tetrachlorodibenzo-p-dioxin
- AhR
- aryl hydrocarbon receptor
- ARNT
- aryl hydrocarbon receptor nuclear translocator
- DRE
- dioxin-responsive enhancer
- CYP1A1
- cytochrome P450 1A1
- NF-κB
- nuclear factor-κB
- SAR
- structure-activity relationship
- HxCDD
- 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin
- TriCDD
- 2,3,7-trichlorodibenzo-p-dioxin
- MCDD
- 1-monochlorodibenzo-p-dioxin
- DMSO
- dimethyl sulfoxide
- LPS
- lipopolysaccharide
- RT-PCR
- reverse transcriptase-polymerase chain reaction
- IS
- internal standard
- ELISA
- enzyme-linked immunosorbent assay
- DTT
- dithiothreitol
- PMSF
- phenylmethylsulfonyl fluoride
- EMSA
- electrophoretic mobility shift assay
- PAGE
- polyacrylamide gel electrophoresis
- PCDD
- polychlorinated dibenzo-p-dioxin
- Received May 1, 2000.
- Accepted July 13, 2000.
- The American Society for Pharmacology and Experimental Therapeutics