Induction of the Cyp1a-1 dioxin-responsive enhancer in transgenic mice.

Cyp1a-1, whose product, aryl hydrocarbon hydroxylase, assists in detoxification of polycyclic aromatic hydrocarbons, is the best characterized of the murine cytochrome P450 genes. The Cyp1a-1 dioxin-responsive enhancer region has been previously analyzed in vitro and found to induce expression of heterologous genes upon exposure of transfected cells to various aromatic hydrocarbons. A 2.58 kbp DNA fragment containing the Cyp1a-1 enhancer elements and promoter region was coupled to the chloramphenicol acetyltransferase (CAT) reporter gene and used to create transgenic mice. CAT assays were performed on tissues harvested from three different lines of transgenic mice following mock-induction or induction using the aromatic hydrocarbon, 3-methylcholanthrene. Basal levels of expression were detected in the spleen and small bowel of non-induced mice, with little or no expression detected in the liver. Treatment with 3-methylcholanthrene increased hepatic expression levels by as much as 10,000-fold. More modest levels of induction were also recorded in the spleen, small bowel, kidney, and lung. The results indicate that the dioxin-responsive enhancer region functions as a strongly inducible promoter in vivo. Differences in the response to induction between male and female mice suggest that Cyp1a-1 expression may be governed in a gender related manner.     

The potential role of DNA methylation in the response to 2,3,7,8-tetrachlorodibenzo-p-dioxin

The Ah receptor is an intracellular protein that binds the environmental contaminant 2,3,7,8-tetrachlorodibenzo-p-dioxin. The liganded receptor interacts with a specific DNA recognition motif located within a dioxin-responsive enhancer upstream of the CYP1A1 gene. Methylation protection and methylation interference studies indicate that the liganded receptor contacts both DNA strands, at 4 guanine residues contained within the recognition motif. These findings imply that the liganded receptor interacts with its cognate enhancer within the major groove of the DNA helix. Cytosine methylation of the recognition motif at CpG dinucleotides diminishes the protein-DNA interaction, as measured by gel retardation. Furthermore, methylation at cytosine inhibits the enhancer function of the DNA. These findings imply that DNA methylation can diminish the response to dioxin by impeding the Ah receptor-enhancer interaction.     

Guanine 6-O-methylation pattern within the dioxin responsive element of the CYP1A1 enhancer shows two critical guanines for AhR/ARNT binding

The core-recognition motif for TCDD-liganded AhR/ARNT complex of the dioxin-responsive element (DRE) contains four guanine residues, three on the antisense (5'-T(T)/(A)GCGTG-3') and one on the sense (5'-CACGC(A)/(T)A-3') strand. It has been reported that, in methylation-protection and methylation-interference assays, the TCDD-liganded AhR/ARNT contacts all four guanine residues. On the other hand, it is known that some anticancer drugs, and various environmental and workplace chemicals, including strongly human carcinogenic nitrosamines, lead to the highly miscoding 6-O-methylation of guanine. In the present study, we have investigated whether specific methylation of guanine at the 6-O-position interferes with the binding of TCDD-liganded AhR/ARNT complex to its recognition motif in the CYP1A1 enhancer, and how individual 6-O-methylated guanines contribute to this interference. We found that only two 6-O-methylguanine residues are critical: the closest to the 5'-end within the three-nucleotide sequence (5'-GTG-3'), identical to a half-site E-box element, on the antisense strand, and the only guanine on the sense strand. In contrast, the 6-O-methylguanine closest to the 5'-end (well) and the one closest to the 3'-end (to a lesser extent), both on the antisense strand, were still able to bind the TCDD-liganded AhR/ARNT complex. This shows that the 6-O-methylation of guanine in the core sequences of CYP1A1 enhancer interferes with the binding of the ligand-activated AhR/ARNT complex in a differentially selective manner; it fully impedes binding of this complex to DRE (the prerequisite of most of the toxic effects of TCDD) only when one of the two particular guanines is methylated at the 6-O-position.     

Characterization of multiple forms of the Ah receptor: recognition of a dioxin-responsive enhancer involves heteromer formation.

We have employed a combination of gel retardation, protein-DNA cross-linking, and protein-protein cross-linking techniques to further examine the 2,3,7,8-tetrachlorodibenzo-p- dioxin-(TCDD-) dependent changes in the Ah receptor that result in a DNA-binding conformation. Gel retardation analysis of DNA-Sepharose chromatographic fractions of rat hepatic cytosol indicated that TCDD-dependent and sequence-specific DNA binding coeluted with a 200-kDa form of the Ah receptor (peak 2) previously characterized as being multimeric and having high affinity for calf thymus DNA. The TCDD-bound, 100-kDa form of the receptor (peak 1) bound weakly to the DNA recognition motif. These results indicated that the DNA-binding form of the Ah receptor is a multimer. SDS-polyacrylamide gel electrophoresis of peak 2 cross-linked to a bromodeoxyuridine-substituted DNA recognition motif indicated that this form of the receptor present in rat hepatic cytosol is composed of at least two DNA-binding proteins of approximately 100 and 110 kDa. Using the chemical cross-linking agent dimethyl pimelimidate, we further established that the 100-kDa form of the receptor (peak 1) associates with a different protein to generate the receptor form (peak 2) that binds to the dioxin-responsive enhancer. Photoaffinity-labeling studies indicated that only the 100-kDa protein (peak 1), and not the 110-kDa protein, binds ligand. Together, these observations imply that the DNA-binding form of the Ah receptor exists as a heteromer.     

Assessment system for dioxin absorption in the small intestine and prevention of its absorption by food factors

It has been reported that 90% of the amount of dioxin in the whole body is absorbed orally with food. However, a concise and simple system to assess dioxin absorption in the small intestine has not yet been established. The present study reports a new in vitro assessment system for this purpose. A stable dioxin-responsive cell line was established by introducing a plasmid that incorporates a xenobiotic-responsive element upstream of the luciferase gene into human hepatic HepG2 genomic DNA. Dioxin was added to the apical side of differentiated human intestinal epithelial Caco-2 cell monolayers that had been cultured on a semipermeable membrane, and the basal medium was recovered after an appropriate incubation time. To the recovered medium was added dioxin-responsive HepG2, and a luciferase assay was performed. The established stable cell line clearly showed dose-and time-dependent response to dioxin. When a food factor such as chlorophyll, which has been reported to increase dioxin excretion in in vivo studies, was added with dioxin, a significant decrease in dioxin permeability to the Caco-2 monolayer was observed. This assessment system would be useful to search for those food factors that could prevent dioxin absorption in the small intestine.     

A novel BK virus-based episomal vector for expression of foreign genes in mammalian cells.

A composite mammalian cell-E. coli shuttle vector was developed based on the human papova virus BK and pSV-neo. The vector contains a dioxin-responsive enhancer (DRE) controlling a mouse mammary tumor virus (MMTV) promoter for the inducible expression of inserted genes. In human cells the vector replicates episomally, presumably utilizing the BKV rather than the SV40 origin, and expresses the BK T/t antigens. A deletion in the late BK region precludes the expression of the core/capsid proteins VP1, VP2, and VP3, thereby preventing the infectious lytic cycle. HeLa cells which were transfected with this vector and selected for resistance to the antibiotic G418 maintained the construct primarily in episomal form during more than one year of continuous culture, with little or no integration into the host genome. Transformed cells cultured in higher concentrations of G418 contained higher copy numbers of the vector. This permits one to vary the dosage of an inserted gene easily and reversibly without the need of conventional amplification techniques and clonal analysis. Using a chloramphenicol acetyl transferase (CAT) reporter gene inserted downstream of the MMTV promoter, we found that CAT expression was greater in clones with higher vector copy number. CAT expression was inducible with 2,3,7,8-tetrachlorodibenzo-p-dioxin, but inducibility was found to be inversely proportional to the copy number. Transformation of bacteria with plasmid molecules retrieved from the mammalian host was efficient, making this vector well adapted for the screening of cDNA libraries for the ability to express a phenotype in mammalian cells. Moreover, DNA sequences were stable during long-term passage in mammalian cells; vector passaged continuously for more than one year retained fully functional bacterial genes for resistance to chloramphenicol and ampicillin.     

Lack of an absolute requirement for the native aryl hydrocarbon receptor (AhR) and AhR nuclear translocator transactivation domains in protein kinase C-mediated modulation of the AhR pathway.

Protein kinase C (PKC)-mediated modulation of the aryl hydrocarbon receptor (AhR) pathway was examined in CHOK1-derived L10.I cells stably transfected with the pGUDLUC6.1 reporter; pGUDLUC6.1 is solely controlled by four dioxin-responsive enhancer elements. Co treatment of L10.I cells with 10 nM 2,3,7, 8-tetrachlorodibenzo-p-dioxin (TCDD) and 81 nM phorbol 12-myristate 13-acetate (PMA), an activator of sn-1,2-diacylglyerol binding PKCs, enhanced transactivation of the reporter construct several-fold relative to cells treated with a saturating 10 nM TCDD dose alone; this effect was dubbed the "PMA effect." A domain swapping and deletional analysis of the native AhR and AhR nuclear translocator (ARNT) protein transactivation domains (TADs) was performed to determine if these domains are absolutely required for the AhR x ARNT dimer-mediated PMA effect in the L10.I model system; controls demonstrate the suitability of the L10.I model for these analyses and that endogenous AhR and ARNT levels are extremely low in this model. Transient coexpression of the AhR and ARNT-474-FLAG, an ARNT protein lacking the native ARNT TAD, in L10.I cells reveals the native ARNT TAD is not absolutely required for the AhR x ARNT-474-FLAG dimer to mediate the PMA effect. Transient coexpression of AhRDeltaCVP, a chimeric AhR protein in which the native AhR TAD has been replaced with the VP16 (herpes simplex virus protein 16) TAD (which control experiments demonstrate is unaffected by PMA), and ARNT in L10.I cells indicates that the native AhR TAD is not absolutely required for this AhRDeltaCVP x ARNT dimer to mediate the PMA effect. These observations strongly suggest that PKC-mediated modulation of the AhR pathway is not absolutely dependent on coactivators recruited to the AhR. ARNT dimer by the native TADs of the AhR and its heterodimerization partner ARNT.     

An SV40 "enhancer trap" incorporates exogenous enhancers or generates enhancers from its own sequences

We have transfected monkey CV-1 cells with non-infectious, linear SV40 DNA, lacking the 72 bp repeat enhancer region. Infectious virus was recovered from this "enhancer trap" upon cotransfection with enhancer DNA segments from various viruses, notably a truncated polyoma enhancer that was integrated as a dimer. Cotransfection of the "enhancer trap" with fragmented DNA of mouse, monkey, or human origin yielded no recombinant virus with integrated cellular sequences, with one possible exception. In some transfection experiments without added viral enhancer DNA, SV40 variants were generated that have a segment of their flanking "late" DNA duplicated to substitute for the deleted 72 bp repeat. In one variant, an 88 bp duplication creates a strong enhancer from this nonenhancing DNA region. Both the polyoma enhancer fragment and the spontaneously created enhancers lack the alternating purines-pyrimidines or "CACA box" suggested to be characteristic for enhancer elements and show only limited homology to the "GTGG(AAATTT)G box."