The human CYP1A1 gene is regulated in a developmental and tissue-specific fashion in transgenic mice.

Regulation and expression of human CYP1A1 is demonstrated in transgenic mice. We have developed two transgenic mouse lines. One mouse strain (CYPLucR) carries a functional human CYP1A1 promoter (-1612 to +293)-luciferase reporter gene, and the other strain (CYP1A1N) expresses CYP1A1 under control of the full-length human CYP1A1 gene and 9 kb of flanking regulatory DNA. With CYPLucR(+/-) mice, 2,3,7,8-tetrachlordibenzo-p-dioxin (TCDD) and several other aryl hydrocarbon receptor ligands induced hepatocyte-specific luciferase activity. When other tissues were examined, TCDD induced luciferase activity in brain with limited induction in lung and no detectable luciferase activity in kidney. Treatment of CYP1A1N(+/-) mice with TCDD resulted in induction of human CYP1A1 in liver and lung, while mouse Cyp1a1 was induced in liver, lung, and kidney. Although induced CYP1A1/Cyp1a1 could not be detected by Western blot analysis in brains from CYP1A1N(+/-) mice, induction in brain was verified by detection of CYP1A1/Cyp1a1 RNA. The administration of TCDD to nursing mothers to examine the effect of lactational exposure via milk demonstrated prominent induction of luciferase activity in livers of CYPLucR(+/-) newborn pups with limited induction in brain. However, TCDD treatment of adult CYPLucR(+/-) mice led to a 7-10-fold induction of brain luciferase activity. Combined these results indicate that tissue-specific and developmental factors are controlling aryl hydrocarbon receptor-mediated induction of human CYP1A1.

The cytochrome P450 1A (CYP1A) 1 family of proteins is often linked through metabolism of environmental toxicants to the generation of chemical carcinogens (1). These genes are actively regulated by the aryl hydrocarbon (Ah) receptor (2)(3)(4). The Ah receptor is a member of the basic helix loop helix family of proteins (5,6) and is activated by halogenated hydrocarbons such as 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and polycyclic aromatic hydrocarbons as well as dietary substances such as bioflavonoids (3,7,8). The process of ligand association and Ah receptor activation culminates in nuclear translocation where it partners with the basic helix loop helix region of the aryl hydrocarbon nuclear translocator protein (9 -11). The Ah receptor-aryl hydrocarbon nuclear translocator complex serves as a transcriptional unit binding to highly conserved enhancer sequences termed xenobiotic response elements (XREs) (12)(13)(14). Those genes that carry conserved XREs in their regulatory region are susceptible to transcriptional activation. While the number of genes that are potential targets for the Ah receptor is extensive (15), the levels of transcriptional activation of Cyp1a1/CYP1A1 exceeds those of other target genes and often serves as a model to investigate the functional properties of the Ah receptor.
Although CYP isoforms have been monitored in human tissues and have been shown to be differentially expressed (16), very little is known regarding the mechanisms associated with Ah receptor control of the CYP1A1 gene in vivo. For the most part, studies undertaken in tissue culture or through animal models have been extrapolated to predict the response in humans. For example, the Ah receptor is functional in early embryogenesis as demonstrated by induction of hepatic rodent Cyp1a1 transplacentally at 15 days gestation following administration of 3-methylcholanthrene to pregnant mice (17,18). Analysis of Cyp1a1 in rodents has demonstrated that exposure to Ah receptor ligands is a prerequisite for protein expression. Undetectable or minimal Cyp1a1 is found in hepatic as well as most extrahepatic tissues (19 -21). These expression profiles change dramatically after the animals have been exposed to Ah receptor ligands as confirmed by transcriptional activation of Cyp1a1 and induction of mRNA and protein. The induced expression of Cyp1a1 is found in many extrahepatic tissues. Studies in various strains and substrains of mice that are genetically compromised in Ah receptor function (22,23) demonstrate that induction of Cyp1a1 is modulated by the Ah receptor. This has also been verified in mice where the Ah receptor gene has been rendered non-functional by targeted gene knock-out studies (24 -26). In addition, it is speculated that circulating humoral factors play an important role in Ah receptor function since the treatment of mice with agents known to alter circulating hormones influences Ah receptormediated expression of Cyp1a1 (19,27). Thus, to understand the regulatory properties associated with Ah receptor function and human CYP1A1 gene expression, it would be an advantage to examine expression patterns of the human gene in transgenic mice.
While the Ah receptor is conserved between rodents and humans, there are significant structural differences between the rodent Cyp1a1 and human CYP1A1 locus. The CYP1A cluster is located on chromosome 15q22-qter with CYP1A1 reading toward the telomere and CYP1A2 reading toward the centromere (28). The distance between the start of transcription of CYP1A1 and CYP1A2 is 23 kb (GenBank TM accession numbers AF253322 and NT 010194), and it has been speculated that active enhancer sequences that reside between the two promoters could contribute equally to the expression of both CYP1A1 and CYP1A2 (28). For example, there are 13 XRE core DNA sequences that may participate in CYP1A induction following activation of the Ah receptor. In comparison, the rodent Cyp1a1 and Cyp1a2 genes are organized similarly with Cyp1a1 transcription proceeding in the complement direction to Cyp1a2 transcription, but the promoters are separated by only 4.3 kb of DNA (GenBank TM accession number NT 039474). Since conservation within the regulatory regions of the human CYP1A and mouse Cyp1a locus has not been maintained, the regulatory regions flanking the human and mouse promoters may contribute significantly toward differences in expression patterns as determined by induction and through tissue-specific control.
In this study we describe the development of transgenic CYPLucR mice that express the CYP1A1-luciferase reporter gene. Expression of the CYP1A1-luciferase reporter gene in different tissues is dependent upon exposure to Ah receptor ligands. To complement these studies, we also developed CYP1A1N transgenic mice that carry a 15-kb fragment of human DNA encoding the entire CYP1A1 gene and 11 kb of regulatory DNA. Treatment of both CYPLucR and CYP1A1N mice with TCDD provides a novel opportunity to study the differential patterns of expression of human CYP1A1 as well as rodent Cyp1a1. As will be demonstrated, the induction of human CYP1A1 by Ah receptor ligands is dependent upon tissuespecific as well as developmental factors.
Generation of CYP1A1 Transgenic Mice-The human CYP1A1 gene was cloned from a human liver genomic library constructed in the replacement vector -EMBL-3 (29). The -EMBL-3 clone encoded bases 3975-19807 ( Fig. 1) as determined from alignment of DNA sequence with that deposited in GenBank TM (accession number AF253322). The -EMBL-3 DNA encoded exons 1-7 and ϳ11 kb of 5Ј flanking DNA. The CYP1A1 gene was removed from the -EMBL-3 clone by a single digest with SalI, and the purified DNA was used for microinjection into fertilized CB6F 1 (F 1 hybrid between BALB/c and C57BL/6N mice) mouse eggs and transplanted into the oviduct of pseudopregnant C57BL/6N mice for the production of transgenic mice. This service was provided by the UCSD Transgenic Mouse and Embryonic Stem Cell Shared Resource. For genotype analysis, DNA from tail clippings was screened by PCR using specific CYP1A1 primers that encoded a portion of exon 7. The oligonucleotides spanned bases 5558 -5539 (5Ј-GCCCACAGCCCA-GATAGCAA-3Ј) and bases 6551-6570 (5Ј-GGTCTGGCCAGGTCTAG-GCA-3Ј). For PCR analysis, DNA was amplified in a 30-l reaction containing 15 l of Qiagen Hotstart Master Mix (following the manufacturer's instructions), 2 l of template DNA, and a 0.6 M concentration of each primer. A 15-min activation of the polymerase was carried out at 95°C followed by 35 cycles from 95°C for 30 s to 57°C for 30 s and 72°C for 60 s followed by the final extension at 72°C for 10 min. Reactions were carried out in a PerkinElmer Life Sciences GeneAmp PCR 2400 system. From 10 litters and 48 offspring, six founders were CYP1A1-positive. Each of the founders was bred, and all of the F 1 -CYP1A1 mice produced inducible human hepatic CYP1A1 following a single treatment of 0.32 g/kg TCDD as determined by Western blot analysis. One founder, which we identified as CYP1A1N, was selected for breeding into C57BL/6N mice.
Generation of CYP1A1-Luciferase Transgenic Mice-A portion of the 5Ј flanking region of the human CYP1A1 gene from Ϫ1612 through ϩ292 was cloned upstream of the luciferase reporter gene as reported previously (30). The CYP1A1 promoter and luciferase gene were removed from the pL1A1N plasmid by EcoRI-BamHI digestion, and the fragment was purified from agarose gels. The purified DNA was microinjected into the pronucleus of fertilized CB6F 1 (F 1 hybrid between BALB/c and C57BL/6N mice) mouse eggs and transplanted into the oviduct of pseudopregnant C57BlL/6N mice. All procedures for the generation of transgenic mice were carried out at the University of California, San Diego Superfund Transgenic Core Facility. From 12 litters and 56 offspring, genomic DNA was extracted from tail clippings of 3-week-old mice using a Qiagen Dneasy tissue kit. The luciferase gene was identified using forward (5Ј-GGAGAGCAACTGCATAAGGC-3Ј, bases 70 -88) and reverse (5Ј-AATCTCACGCAGGCAGTTCT-3Ј, bases 641-659) primers. The 590-bp amplified luciferase fragment was identified in four founders. The CYPLuc mice were bred with C57BL/6N mice from Jackson Laboratory, and F 1 -CYPLuc mice were treated with a single intraperitoneal injection of TCDD (0.32 g/kg). One of the founders, identified as Rico, produced F 1 -CYPLucR-positive mice that expressed highly inducible hepatic luciferase activity. The F 1 -CYPLucR mice were bred for at least three generations.
Ethoxyresorufin O-Deethylase Assay-Ethoxyresorufin O-deethylase activity, a marker of CYP1A1 activity, was determined spectrofluorometrically in liver microsomes isolated from CYP1A1N ϩ/Ϫ and CYP1A1N Ϫ/Ϫ mice. The reaction was carried out in 100 mM potassium phosphate buffer (pH 7.4) containing 5 mM MgCl 2 , 8 M 7-ethoxyresorufin, and 1 mM NADPH. The reaction was initiated by the addition of 20 g of microsomal protein. The reaction mixture was incubated at 37°C for 15 min. Control samples were incubated in the absence of NADPH. The incubations were terminated by the addition of cold methanol. The formation of resorufin was measured fluorometrically with the excitation at 530 nm and emission at 590 nm and was normalized to a resorufin standard curve.
Reverse Transcription-PCR Analysis of Mouse Cyp1a1 and Human CYP1A1 RNA-Approximately 200 mg of tissue was homogenized, and total RNA was extracted using acidic phenol/guanidinium isothiocyanate solution (TRIzol, Invitrogen). Three micrograms of total RNA were denatured in the presence of 0.5 g of oligo(dT) primers at 70°C for 10 min, and the cDNA was synthesized as described previously (31). The PCR contained a 2 mM concentration of each primer for mouse Cyp1a1, human CYP1A1, firefly luciferase, and the actin gene. The forward Cyp1a1 primer was 5Ј-CTGGCTGTCACCGTATTCTGC-3Ј, and the reverse Cyp1a1 primer was 5Ј-GGTCACCCCACAGTTCCCGG-3Ј. The forward human CYP1A1 primer was 5Ј-CCGGCGCTATGACCACAAC-C-3Ј and encoded bases 7661-7680 (GenBank TM accession number AF253322), and the reverse primer was 5ЈCCTCCCAGCGGGCAATG-GTC-3Ј encoding bases 5822-5842. The forward mouse actin primer was 5Ј-ATGGCCACTGCCGCATCCTC-3Ј, and the reverse actin primer was 5Ј-GGGTACATGGTGGTACCACC-3Ј. The reaction was carried out in a 100-l volume containing 3 mM MgCl 2 , 50 mM KCl, 20 mM Tris-HCl, pH 8.4, a 0.2 mM concentration of each dNTP, and 5 units of VENT (exo Ϫ ) DNA polymerase. The reaction was initiated by a 5-min incubation at 95°C followed by elongation at 72°C for 7 min followed by 35 cycles at 94°C for 1 min, 58°C for 1 min, and 72°C for min.
Detection of Expressed Luciferase Activity-Tissues extracted from mice were homogenized in 200 l of 100 mM KPO 4 lysis buffer containing 4 mM EGTA, 4 mM EDTA, 1 mM dithiothreitol, and 0.7 mM phenylmethylsulfonyl fluoride. The homogenates were freeze-thawed three times and centrifuged at 14,000 ϫ g for 30 min. The supernatant was collected, and the protein concentrations were determined using a standardized Bradford kit purchased from Bio-Rad. For analysis of luciferase activity, 10 l of each supernatant was mixed with 300 l of 15 mM potassium phosphate buffer, pH 7.8, containing 15 mM MgSO 4 , 2 mM ATP, 4 mM EDTA, 25 mM Tricine, and 1 mM dithiothreitol. To start the reaction, 100 l of luciferin was added, and light output was measured for 10 s at 24°C using a Monolight 2001 luminometer (Analytical Luminescence Laboratories). The relative light units were normalized to the protein content.
Western Blot Analysis-Various tissues were homogenized in 4 volumes of 1.15% potassium chloride solution. The homogenate was centrifuged at 10,000 ϫ g at 4°C for 20 min in an Eppendorf Microfuge to obtain the supernatant fraction. The supernatant fraction was collected and then centrifuged at 100,000 ϫ g for 60 min at 4°C in a Beckman TL100 tabletop ultracentrifuge. The microsomal pellet was resuspended in 100 mM potassium phosphate buffer (pH 7.25) containing 0.1% glycerol, 1 mM EDTA, 20 M butylated hydroxytoluene, and 2 mM phenylmethylsulfonyl fluoride. Microsomes were stored in aliquots at Ϫ80°C prior to Western blot analysis. Protein concentrations were determined using Bradford solutions normalized to standard protein solutions. Twenty micrograms of microsomal protein were boiled for 5 min in loading buffer (6% SDS, 20% sucrose, 0.25 M dithiothreitol, and 0.1% bromphenol blue) and resolved on a 10% SDS-polyacrylamide Tris-glycine gel (Novex, San Diego, CA). The resolved protein was transferred onto a nitrocellulose membrane with an electrotransfer apparatus (Novex) for 90 min. Incubating the membrane with 5% nonfat dry milk in Tris-buffered saline solution for 1 h at room temperature blocked nonspecific protein binding sites. The membrane was then incubated with a rabbit anti-human CYP1A1 primary antibody (32) for 1 h followed by a wash and incubation with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Protein was visualized using chemiluminescence reagents according to the manufacturer's instructions (Renaissance Western blot chemiluminescence reagent, PerkinElmer Life Sciences) followed by exposure to the x-ray film.
Hepatocyte Cell Isolation-Primary hepatocytes were isolated from 8 -12-week-old mice. Mice were anesthetized by isoflurane inhalation. The inferior vena cava was cannulated and perfused at the rate of 4 ml/min. The portal vein was sectioned to allow flow through the liver. The liver was first perfused with Hanks' balanced salt solution containing 0.5 mM EGTA and 10 mM Hepes at pH 7.4 followed by a perfusion with 0.2% collagenase solution buffered with 1 mM sodium phosphate (pH 7.5), 130 mM NaCl, and 3 mM KCl for 3 min at 4 ml/min. The liver was dissected, and hepatocytes were isolated by mechanical dissection, filtered through a sterile 70-m filter, and washed twice by centrifugation at 50 ϫ g for 5 min. The hepatocytes were either lysed or cultured in 6-well plates in 3 ml of Williams' Medium E (Invitrogen) containing 1% L-glutamine and penicillin/streptomycin and supplemented with 10% fetal calf serum (v/v). Four hours after plating, the medium was replaced with fresh medium. Primary hepatocytes were treated 24 h after seeding and analyzed for luciferase activity as described above.

Characterization of Hepatic CYP1A1-Luciferase Expression
in CYPLucR ϩ/Ϫ Mice-The human CYP1A1 gene is regulated in part through activation of the Ah receptor, but other important transcriptional factor activators such as Sp1 (33), NF1, NF-Y B (34), and protein kinase C (35, 36) play important roles in modifying gene expression in the presence of Ah receptor ligands. In HepG2 cells, a Ϫ1612 bp fragment of the CYP1A1 promoter ( Fig. 1) containing these DNA binding sequences as well as the XRE sequences has been shown to serve as a template for transcriptional activation of luciferase reporter gene activity in the presence of Ah receptor ligands (30). To examine whether this CYP1A1-luciferase gene can function as a reporter gene in vivo, CYP1A1-luciferase (CYPlucR)-positive mice (see "Materials and Methods") were generated. Characterization of luciferase induction was examined in CYPLucR ϩ/Ϫ mice after 24 h following a single intraperitoneal injection of TCDD. Examining a range of TCDD treatments between 0.0125 and 16 g/kg, induction of luciferase activity was observed with a minimal dose of 0.32 g/kg (Fig. 2). Detection of microsomal Cyp1a1 by Western blot analysis was also confirmed with a minimal dose of 0.32 g/kg, demonstrating that the cellular and molecular properties that influence gene expression of the CYP1A1-luciferase and the endogenous Cyp1a1 gene are similar.
In tissue culture cells, the human CYP1A1-luciferase gene is responsive to halogenated hydrocarbons and polycyclic aromatic hydrocarbons (30,36) as well as to bioflavonoids such as chrysin (8) and certain substituted benzimidazoles such as omeprazole (38). Along with TCDD, 3-methylcholanthrene, benzo[a]pyrene, and benz[a]anthracene induced hepatic luciferase activity as well as microsomal Cyp1a1 in CYPLucR ϩ/Ϫ mice (Fig. 3). The compound 5,6-benzoflavone (␤-naphthoflavone) was a good inducer of Cyp1a1 in liver microsomes and also induced luciferase activity, but the induction of luciferase activity was less than that observed for 3-methylcholanthrene, benzo[a]pyrene, and benz[a]anthracene. Omeprazole and chrysin, which induce human CYP1A1 in human tissue culture cells as well as CYP1A1-luciferase in TV101 cells (8,38), were not able to induce luciferase activity in CYPLucR ϩ/Ϫ mice. This result would indicate that chrysin and omeprazole are being  (28) is shown on top with an outline of the CYP1A1 gene that was recovered in the human genomic EMBL-3 library. The SalI sites were conserved in the EMBL-3 vector and were used to isolate the human CYP1A1 gene. Based upon DNA sequence from GenBank TM accession number AF253322, the sequence spans bases 3975-19807, encompassing exons 1-7 and ϳ9 kb of flanking DNA. The 15.8-kb CYP1A1 gene fragment was used to generate the CYP1A1N mice. From this fragment of DNA, a portion of the regulatory region encoding the XRE sequences and the non-coding first exon (bases Ϫ1612 through ϩ292) was cloned into a luciferase reporter expression vector as described previously (30). The portion of DNA encoding the CYP1A1 promoter and the luciferase gene was used for the generation of the CYPLucR mice.
influenced by in vivo factors in a pattern that differs from their ability to induce CYP1A1 in tissue culture cells.
Microsomal P450 expression in liver is localized to the hepatocytes. To examine whether induced luciferase activity is expressed in hepatocytes, CYPLucR ϩ/Ϫ mice were treated for 24 h with 3.2 g/kg TCDD, and luciferase activity was assayed in isolated hepatocytes. When we compared luciferase activity in hepatocytes isolated from untreated mice, CYP1A1-luciferase activity was induced in hepatocytes from the TCDD-treated mice (Table I). A similar induction profile was observed when hepatocytes from untreated CYPLucR ϩ/Ϫ mice were first cultured for 24 h and then treated with 10 nM TCDD. These results suggest that integration of the CYP1A1-luciferase gene into the mouse genome can be regulated in a tissue-specific and inducible fashion comparable to what is observed for expression of the mouse Cyp1a1 gene.
Extrahepatic Expression of CYP1A1-Luciferase-Analysis of luciferase activity in different tissues both in the absence and presence of Ah receptor ligands would be considered a good indicator of the actions of ligand toward activation of the Ah receptor and induction of the CYP1A1 gene. To examine the expression of CYP1A1-luciferase in different tissues, 8-weekold CYPLucR ϩ/Ϫ mice were treated with 3.2 g/kg TCDD for 24 h, the tissues were isolated, and luciferase activity was monitored (Fig. 4). As expected, the most significant level of induced expression was noted in liver. No induction was noted in small or large intestine, which differs from previous reports on the induction of Cyp1a1 in rat and mouse gastrointestinal tract (39). To our surprise, there was no induction of luciferase activity in kidney and only limited induction in lung, two tissues that have been studied extensively and shown to elicit induction of Cyp1a1 (19), as shown in Fig. 5.
Interestingly significant levels of induced luciferase activity occurred in brain. This large increase in luciferase activity may in part be a result of basal luciferase activity that is nearly 100-fold higher than that observed in livers from untreated mice. Elevated levels of constitutive luciferase activity are in agreement with reports that Cyp1a1 is constitutively expressed in rodent brain (39,40). These results are also in agreement with previously published observations indicating that induced levels of Cyp1a1 accumulate in rat cerebral cortex and cerebellum (39). Localization of Cyp1a1 mRNA by in situ hybridization has also been identified in the choroid plexus and distributed in the lateral third and fourth ventricles. Since expression of CYP1A1-luciferase is Ah receptor-dependent, this finding suggests that one possibility of the higher basal levels of luciferase activity may result from exposure to brain-specific factors that serve as endogenous Ah receptor ligands.
Expression of the Human CYP1A1 Gene-To determine whether the hepatic and extrahepatic expression patterns of the CYP1A1-luciferase gene mimic expression patterns found with the full-length human CYP1A1 gene, CYP1A1N ϩ/Ϫ mice were developed that express the human CYP1A1 gene (Fig. 1). To examine expression of human CYP1A1 in CYP1A1N mice, Western blots were conducted with an anti-human CYP1A1 antibody (32) with specificity toward mouse Cyp1a1 and human CYP1A1. Using microsomes from wild type mice, it was seen that a dose of 3.2 g/kg TCDD induces liver Cyp1a1 (Fig.  5). In contrast, when TCDD was administered to CYP1A1N ϩ/Ϫ mice and liver microsomal proteins were identified by Western blot analysis, two prominent induced proteins were identified. One migrated as Cyp1a1, while the other migrated slightly faster with an R F value that is identical to TCDD-induced CYP1A1 from human HepG2 cells. TCDD-induced liver microsomal ethoxyresorufin O-deethylase activity in CYP1A1N ϩ/Ϫ mice was nearly twice the value detected in wild type mice from the same litter (Table II).  a Luciferase activity measured in cytosolic fractions from CYPLucR mice treated for 24 h with 3.2 g/kg TCDD. b CYPLucR mice were treated for 24 h with TCDD, and luciferase activity was measured in isolated hepatocytes.
c Hepatocytes were isolated from CYPLucR mice and placed into culture for 24 h. TCDD (10 nM) was then added to the isolated hepatocytes, and after a 24-h exposure luciferase activity was determined. d RLU, relative light units.
When we compared the induction profiles in extrahepatic tissues from TCDD-treated CYP1A1N ϩ/Ϫ mice, Cyp1a1 was detectable in lung and kidney (Fig. 4B) with minor identification in brain (data not shown). Unlike other reports, we did not observe any induced Cyp1a1 in small or large intestine. In contrast, human CYP1A1 was identified in lung, but there was no detectable induction in kidney or brain.
While we clearly identified TCDD-inducible CYP1A1-lucifer- FIG. 4. Tissue-specific induction of CYP1A1-luciferase activity in CYPLucR ؉/؊ mice by TCDD. Adult CYPLucR mice were treated with a single intraperitoneal injection of Me 2 SO (100 l) or TCDD (3.2 g/kg), and cytosolic fractions were prepared after 24 h. Tissues from three Me 2 SO-or TCDD-treated mice were pooled, and relative luciferase activity was determined. Each determination was carried out three times. RLU, relative light units. Statistically significant differences are noted by an asterisk.
FIG. 5. Induction of CYP1A1 and Cyp1a1 in CYP1A1N ؉/؊ and CYP1A1N ؊/؊ mice by TCDD. Western blots were prepared using microsomal preparations, and TCDD-induced CYP1A1 or Cyp1a1 was monitored by detection with the anti-human CYP1A1 antibody. A, three CYP1A1N-negative (F 1 -1A1 Ϫ/Ϫ ) and three CYP1A1N-positive (F 1 -1A1 ϩ/Ϫ ) mice derived from the same litter were treated with 3.2 g/kg TCDD (T) for 24 h, and microsomal preparations from liver were analyzed for induction of human CYP1A1 and mouse Cyp1a1. Microsomes from an Me 2 SO (D)-treated CYP1A1N-positive (F 1 -1A1 ϩ/Ϫ ) mouse is included. To monitor for induction of human CYP1A1, microsomal preparations from human HepG2 cells treated with 10 nM TCDD for 24 h were included. B, liver, lung, and kidney microsomes were prepared from adult CYP1A1N ϩ/Ϫ (1A1 ϩ/Ϫ ) and CYP1A1N Ϫ/Ϫ (1A1 Ϫ/Ϫ ) mice that were treated with either Me 2 SO (D) or TCDD (T). Tissues were pooled for two animals, and microsomal preparations were generated. Each lane contains 20 g of microsomal protein. C, RNA from liver and brain tissue from Me 2 SO (D)-treated and TCDD (T)-treated CYP1A1N ϩ/Ϫ (1A1 ϩ/Ϫ ) and CYP1A1N Ϫ/Ϫ (1A1 Ϫ/Ϫ ) mice was used in a reverse transcription-PCR. Primers specific for the detection of human CYP1A1, mouse Cyp1a1, and actin were included in the same reaction. Std, standard. ase activity in brain and this corresponds to previous reports that Cyp1a1 is inducible in brain, the inability of Western blot analysis to identify induced CYP1A1 in brain indicates that the expression patterns are very low in this tissue. However, when we analyzed CYP1A1/Cyp1a1 expression patterns in brain by reverse transcription-PCR, inducible RNA that encoded both Cyp1a1 and CYP1A1 was detected.
Effect of Lactational Exposure to TCDD on Neonatal Liver and Brain CYP1A1-Luciferase-To examine whether TCDD is capable of modifying neonatal gene expression patterns in CYPLucR ϩ/Ϫ mice, lactating maternal mice were given a single injection of TCDD, and activity in liver and brain from CYPLucR ϩ/Ϫ mice was monitored in feeding neonates. In this experiment, a single dose of 3.2 g/kg TCDD was injected into maternal mice the 1st day postpartum. Analysis of luciferase activity in liver and brain was assessed after 24 h in nursing neonatal CYPLucR ϩ/Ϫ mice.
In CYPLucR ϩ/Ϫ neonatal mice, luciferase activity in liver was barely detectable from feeding mice whose mothers received a single injection of Me 2 SO (Fig. 6). In liters where the maternal mouse received a single intraperitoneal injection of TCDD, luciferase activity in CYPLucR ϩ/Ϫ neonatal mice was induced. The levels of luciferase activity were more than double those observed in livers from 2-month-old mice that received a single intraperitoneal injection of TCDD.
In comparison to liver, brain CYP1A1-luciferase activity in neonatal mice that were breastfeeding from the TCDD-treated maternal mouse was only 2-fold that of constitutive luciferase activity. In contrast, brain CYP1A1-luciferase activity following TCDD treatment in adult mice was nearly 7-fold that observed from untreated adult mice. In addition, the constitutive levels of brain CYP1A1-luciferase activity in adult mice were nearly 3-fold higher than those in neonatal mice. The higher levels of constitutive luciferase activity as well as inducible activity in adult brain indicate that specific developmental factors are maturing in adult animals to aid in the expression of CYP1A1.

DISCUSSION
The development of transgenic animal models expressing human genes offers an opportunity to investigate the contribution of regulatory factors that modulate gene expression as well as structural determinants of the gene that may influence expression patterns in vivo. These studies provide an advantage over classical tissue culture model systems since expression patterns can now be linked to the role of circulating humoral factors as well as the contribution of tissue-specific regulatory factors. In this report, we have evaluated the tissuespecific expression patterns of a luciferase reporter gene under control of 1600 bases of the CYP1A1 promoter in CYPLucR mice and compared luciferase induction to expression patterns of human CYP1A1 in CYP1A1N mice that carry the entire human CYP1A1 gene. Both the CYPLucR and CYP1A1N genes were expressed in a similar fashion in response to TCDD induction in liver. Expression of CYPLucR in liver was localized to hepatocytes and corresponded with induction of mouse Cyp1a1 as well as induction of CYP1A1 in CYP1A1N mice.
Since both transgenes are induced in liver by TCDD, gene expression might be controlled entirely through activation of the Ah receptor. However, this does not appear to be the case. TCDD treatment induced mouse Cyp1a1 in lung and kidney tissue as previously reported (21,39). However, transgenic gene expression in kidney CYPLucR and CYP1A1N mice was not observed. Since CYP1A1 has been detected in human kidney (41), the dichotomy in expression of rodent Cyp1a1 and human CYP1A1 in kidney might indicate that the organization of the chromatin around the transgenic sequences is not conducive to activation by the Ah receptor. Alternatively the human CYP1A1 gene contains sequence elements that require tissue-specific regulatory factors that are only available in human kidney. Interestingly the human CYP1A1 gene contains a negative regulatory domain within Ϫ800 bases from the transcriptional start site that is not present in the mouse Cyp1a1. It is predicted that transcriptional factors associating with the negative regulatory domain sequences suppress transcriptional activation (34). Although Cyp1a1 was induced in mouse kidney, regulation of human CYP1A1 in this tissue may be tightly controlled through the negative regulatory domain sequence. We can only speculate that the differences in expression of Cyp1a1 and CYP1A1 in kidney result from evolutionary differences in the organization of the mouse and human regulatory regions flanking the Cyp1a1/CYP1A1 genes. However, it cannot be ruled out that positioning of the transgene at or near FIG. 6. Lactational exposure to TCDD and induction of liver and brain luciferase activity in CYPLucR mice. Maternal mice were injected with either 3.2 g/kg TCDD or Me 2 SO (DMSO) (100 l) 1 day after giving birth, and neonatal mice were allowed to nurse for 2 days. Liver and brain tissues from 3-dayold CYPLucR mice were collected, and luciferase activity was determined. For comparison, adult mice were treated with 3.2 g/kg TCDD, liver and brain tissue was isolated after 24 h, and luciferase activity was determined. Values are the means Ϯ S.D. of three separate experiments. The left-hand panel displays the luciferase activity in neonatal liver and brain from maternal mice treated with Me 2 SO or TCDD, while the right-hand panel displays luciferase activity in liver and brain tissue from adult mice. Neo, neonatal mice; RLU, relative light units. the rodent Cyp1a1 locus is necessary for appropriate expression in selective tissues. There are several dramatic differences in regulation of the transgenes in CYPLucR and CYP1A1N mice. Human CYP1A1 was induced in lung microsomes from CYP1A1N mice but was not observed in brain where dramatic inducibility of luciferase activity occurred in CYPLucR mice. While expression of CYP1A1 in CYP1A1N liver and lung paralleled closely the patterns observed for induction of Cyp1a1 and best mimicked the regulatory pattern expected for expression of CYP1A1 gene, the high levels of inducible luciferase activity in CYPLucR brain is unusual since the human CYP1A1 and mouse Cyp1a1 gene do not appear to be dynamically regulated in mouse brain. However, this may be a fortuitous finding by providing an animal model that will allow for analysis of Ah receptor ligands that are permeable to the blood-brain barrier.
The sensitive induction profiles in CYPLucR mice also served as an in vivo model system to examine developmental regulation of CYP1A1. Mouse Cyp1a1 is expressed constitutively and induced in adult brain (39,40), and we demonstrated that TCDD was capable of inducing CYP1A1-luciferase activity in adult brain from CYPLucR mice. In contrast, induction of CYP1A1-luciferase in brains from neonatal mice that were receiving TCDD through lactational exposure showed minimal induction. The lack of CYP1A1-luciferase induction in neonatal brain from CYPLucR mice was not observed in liver tissue where induction of luciferase activity in liver was nearly double that found in livers from adult CYPLucR mice. The progressive induction of CYP1A1-luciferase in brains from CYPLucR mice as a result of age indicates that late stage neonatal or adultspecific regulatory factors are necessary in brain tissue for the activated Ah receptor to participate in CYP1A1 induction. The induction pattern of CYP1A1-luciferase in adult brain was also concordant with induction of brain Cyp1a1 mRNA by TCDD, indicating that brain-specific regulatory factors work in concert with the Ah receptor to modulate transcriptional activation of the Cyp1a1 gene.
For the most part, agents that are capable of activating the mouse Ah receptor in tissue culture cells have been demonstrated to activate the human Ah receptor. However, species differences in Ah receptor response have been reported. For example, omeprazole, which is a substituted benzimidazole used clinically to suppress gastric acid secretion by inhibiting the H ϩ /K ϩ -ATPase in gastric parietal cells, transcriptionally activates CYP1A1 in human cells in an Ah receptor-dependent mechanism (38). However, the treatment of mouse hepatoma cells fails to initiate transcriptional activation of Cyp1a1 (42). This difference in response has been reported to result from a species-specific difference in the generation of an omeprazole metabolite that serves as a unique ligand for activation of the Ah receptor. Indeed, when we administer omeprazole to CYPLucR mice, there was no induction of CYP1A1-luciferase activity, supporting findings that unique metabolic profiles may influence the actions of Ah receptor activation in a speciesspecific fashion.
Chrysin, on the other hand, is a dietary flavonoid that has been shown to induce human CYP1A1 (8) as well as mouse Cyp1a1 in vitro (43). When chrysin was administered to CYPLucR mice, there was no detectable induction of hepatic CYP1A1-luciferase activity even after repeated administrations up to 72 h (data not shown). Although chrysin is an active inducer in mouse and human cells, the in vivo environment can dramatically influence the biological fate of specific substances. It might be anticipated that absorptive properties or extensive metabolism may alter the ability of chrysin to activate the Ah receptor in vivo. Thus, while analysis of Cyp1a1/CYP1A1 gene induction profiles in tissue culture is an excellent predictor that agents are potential Ah receptor ligands, the use of an animal model is more informative in predicting in vivo outcomes of events that regulate CYP1A1 gene transcription.
In developing a CYP1A1 transgenic model, we have attempted to clone a gene fragment that contains all of the regulatory elements. It has been emphasized by Corchero et al. (28) that regulatory elements on the CYP1A2 gene may participate in CYP1A1 gene expression. This is possible, and the lack of expression of CYP1A1 in kidney tissue may reflect an absence of an element that is needed for tissue-specific regulation in kidney tissue. However, our results do indicate that transgenic expression of the human CYP1A1 gene by Ah receptor agonists requires the involvement of unique tissue-specific and developmental factors, some of which may be independent of those required to induce expression of the mouse Cyp1a1 gene. This work differs slightly from previous reports describing the transgenic expression of the mouse Cyp1a1 (44) or rat CYP1A1 promoter (37) in that reporter gene expression of these constructs closely mimicked the expression patterns of the endogenous Cyp1a1 gene. It might be argued that the lack of expression of the CYPLuc and CYP1A1 transgenes in certain extrahepatic tissues results from a chromosomal integration site that prevents access of the activated Ah receptor to XRE sequences in these tissues. However, other founders carrying the CYP1A1 gene fragment generated similar induction profiles in liver, kidney and lung (data not shown). These data might indicate that induction of human CYP1A1 by Ah receptor ligands in vivo requires the presence of additional regulatory elements or transcriptional factors that are not essential for the induction of mouse Cyp1a1.