Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Cytochrome P (CYP) monooxygenases comprise a multigene family whose gene products are essential in the biotransformation of numerous endogenous and exogenous compounds. The P450 superfamily is divided into families and subfamilies. The cytochrome P (CYP) 2B1 and 2B2 genes are members of the P450 2B subfamily and are 98% similar in their mRNA and amino acid sequences (1, 2). This subfamily includes an estimated eight to 11 members at the genomic level, and contains P450 enzymes induced by phenobarbital (PB) or by hormones, as well as P450 genes not induced by PB (3). CYP2B1 is the dominating enzyme in rat lung (4). To date, altogether, six P450 enzymes, including CYP2A3, 2E1, 3A2, 4B1, and to a minor extent 1A1, have been detected in rat lung (4). In contrast, the CYP2B2 gene is not expressed in rat lung tissue (7), although it is expressed in rat liver. Rat pulmonary CYP2B1 expression at both the mRNA and the protein level has been reported at Day 7 after birth (8, 10). Thereafter, the pulmonary expression of CYP2B1 reaches its maximum from 4-6 wk after birth and remains at a high level. The expression of CYP2B1 in liver is low compared with that in lung tissue, and reaches a maximum during the first 4 wk after birth, followed by a decrease in level. These findings indicate different regulating mechanisms for CYP2B1 in lung and liver. However, it is most likely that CYP2B1 and CYP2B2 are to a large extent regulated in a similar way in liver cells. At a cellular level, pulmonary CYP2B1 expression is found mainly in alveolar type II cells, bronchiolar Clara cells, and nonciliated epithelial cells in trachea and larger bronchi (4, 11).

CYP2B1 and 2B2 are encoded by separate genetic loci, indicated to be nonallelic and closely linked to each other on the same autosome (17). The gene for CYP2B1 is 23 kb long and is separated into nine exons (18). This structure is very similar to that of the 2B2 gene, except for the first intron of CYP2B1, which is about 9 kb longer than that of the 2B2 gene (18). The greatest divergence in base composition between the two genes has been observed in exons 7 and 8. Close sequence homology between the promoter regions of the 2B1 and 2B2 genes has been found at up to 2,300 bp (19), with the exception of a repeated cis-acting (CA) sequence at position 255 (18), which was found to be significantly shorter in the 2B1 gene, and a 12-bp insert at position 706 in the 2B1 gene. Relatively little is known about the complex pattern of regulation involved in the tissue-specific and age-dependent expression of these genes. However, the CYP2B genes are strongly induced transcriptionally in liver cells by PB treatment (18, 20), and the involvement of a drug and heme-modulated transcription factor in liver has been indicated (21, 22). In contrast, pulmonary CYP2B1 is not induced by PB and is probably induced to only a small extent by other compounds, although one report describes the induction of pulmonary CYP2B1 by ozone (23).

Although several studies have indicated a role for genetic elements located close to the transcription initiation site of CYP2B2 in the PB-induction response (21, 23), other studies have demonstrated the importance of regulatory elements further upstream of the core promoter region of the CYP2B gene (20, 24, 31, 32). Recently, PB- enhancer activity has been associated with the function of DNA sequences at about 2.3 kbp in the rat CYP2B2 and mouse Cyp2b10 genes (32). A fully PB-inducible enhancer element was identified in this region of the Cyp2b10 gene, and was designated phenobarbital-responsive enhancer module (PBREM) (35).

The purpose of our work was to further elucidate tissue- and phenobarbital-dependent regulatory mechanisms of the CYP2B1 gene with a reporter gene construct introduced into transgenic mice. Our results suggested that sequences within 1.3 kb of the transcription start site in the CYP2B1 gene are important for regulation of pulmonary transcription of CYP2B1, and also for PB-induced transcription of CYP2B1 in liver cells.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Polymerase Chain Reaction Amplification of Rat Genomic DNA

Rat CYP2B1 and 2B2 5' flanking regions (including exon 1) were specifically amplified with the polymerase chain reaction (PCR) technique, using the GenAmp DNA amplification reagent kit (Perkin-Elmer, Roche Molecular Systems, Inc., Branchburg, NJ) essentially according to the manufacturer's instructions. DNA prepared from male Sprague-Dawley rat livers (Clontech Laboratories Inc., Palo Alto, CA) was used as a template, and the oligomer primers used in the reaction were synthesized to corresponding to the CYP2B1 and 2B2 5' upstream sense strand (primer 1: 5'-GCAAGCTTTTCCTCTAAGTGTCC-3', and primer 2: 5'-GCAAGCTTCAAACATAATCACATGTACCCA-3'), or to the downstream exon 1 antisense strand (primer 3: 5'-GCAAGCTTGAAGGAATTGAGGAGGCC-3'). The primers included an HindIII cloning site. The amplification reactions were performed with a hot-start procedure. Cycling conditions were one cycle at 50°C for 2 min, 30 cycles at 72°C for 3.5 min, and 95°C for 30 s; one cycle 72°C for 10 min; and a 4°C soak for 10 s. A 10-s extension of the 72°C incubation step was included in each cycle. The amplified products were analyzed through gel electrophoresis followed by ethidium bromide staining, yielding PCR products of 1,650 bp. Amplified 1,650-bp fragments were inserted into pGem 4z plasmids (Promega, Madison, WI), using the HindIII cloning sites. Three positive clones were selected and characterized by sequence analysis. Plasmid DNA was prepared according to Kraft and coworkers (36), and DNA sequencing was performed essentially according to the dideoxy chain termination method of Sanger and coworkers (37), using the Sequenase kit (USB Biochemicals, Cleveland, OH).

Northern Blotting

Total cytoplasmic RNA was prepared from PB-treated and control rat lungs and livers, using the guanidinium thiocyanate-phenol-chloroform extraction method (38), and 10 µg of each sample was separated on a formaldehyde-containing gel and transferred to a Hybond-N filter (Amersham, Buckinghamshire, UK) by capillary blotting. Filters were hybridized overnight at 42°C with probes corresponding to exon 1 (650-bp fragment cut with NcoI and PstI) and exons 7-9 (PR-17 complementary DNA [cDNA]) (8) labeled with (³²P)deoxycytosine triphosphate ([³²P]dCTP) (3,000 ci/mmol) with the multiprime labeling system (Amersham) (8), and subsequently washed. Films were developed after overnight exposure at 70°C with intensifying screens.

Construction of a Reporter Gene and Development of a Transgenic Animal Model

To identify parts of the CYP2B1 gene necessary for lung-specific expression, a 1,334-bp 5' fragment (released from the CYP2B1 clone by HindIII and SphI restriction enzymes) was coupled to the structural gene for LacZ, and the construct was transfected into primary hepatocytes from rats and into the human hepatoma cell line HepG2. Positive staining with 5-bromo-4-chloro-3-indolyl--D-galactopyranoside (X-gal) showed that the construct was able to drive expression in liver cells. A transgenic mouse model was thereafter developed at the Transgenie Core Facility of the Karolinska Institute, Novum, by injecting the construct into fertilized mouse oocytes. Integration of the construct was determined by PCR amplification of tail DNA, using LacZ primers (sense: GCATCGAGCTGGGTAATAAGCGTTGGCAAT; antisense: GACACCAGACCAACTGGTAATGGTAGCGAC) and cycling parameters described by Hanley and Merlie (39). The transgene copy number in the transgenic animals was estimated by adding serial dilutions of a known amount of LacZ copies to 12.5 ng of mouse liver DNA, prepared from either a transgenic animal or from a control mouse with the Wizard Genomic DNA purification kit (Promega). Amplification by PCR, using LacZ and rapsyn primers (internal control for equal loading), was performed for 28 cycles, essentially as previously described (39). The amounts of PCR products were measured by densitometry after separation by gel electrophoresis, and were plotted versus the number of plasmid template copies added of the transgene.

Animals and Treatments

C57/BI6xCBA mice were used for generation of transgenic animals. All animals were kept in rooms with a 12 h day/12 h night diurnal cycle, and were given Labfor R36 standard pelleted laboratory chow (Lactamin, Stockholm, Sweden) and water ad libitum. Transgenic offspring from two founders, T29-21 and T29-30, were used in the experiments. Offspring of T29-30 origin were further bred to a homozygous line through sister and brother matings before being used in experiments. The T29-21 offspring were analyzed in a heterozygous state. Homozygous transgenic mice derived from the T29-30 founder and control mice were treated with PB sodium salt in saline for 4 d (80 mg/ kg/d intraperitoneally) and were killed 24 h after the last injection, following the procedure described by Honkakoski and Lang (40).

Reverse Transcription-PCR

Using the method of Chomczynski and Sacchi (38), total cytoplasmic RNA was prepared from lung and liver tissue in transgenic mice (of both T29-21 and T29-30 origin) and control mice, and from Balb MK mouse keratinocytes stably transfected with a plasmid containing the LacZ gene coupled to the thymidine kinase (TK) promoter (positive control). Total RNA was also prepared from PB-treated mice (homozygous transgenic animals of T29-30 origin and control mice). Subsequently, the RNA was treated with deoxyribonuclease (DNase), using DNase I supplied in the total RNA kit from Scotlab (Scotlab Inc., Shelton, CT) according to the manufacturer's protocol, and was stored at 80°C until used in reverse transcription (RT)- PCR experiments.

From 1-2 µg of RNA was reverse transcribed using SUPERSCRIPT II RNase H-reverse transcriptase from Life Technologies (Gaithersburg, MD) according to the supplier's protocol with some modifications. Briefly, each reaction included 1 mM of deoxynucleotides, 1 µM of downstream primer, and the addition of 1.3 U ribonuclease inhibitor (RNasin) (Promega) in a total volume of 10 µl. The RT step was performed at 42°C for 50 min, followed by 15 min at 70°C. A negative control was obtained by omitting the enzyme. Five microliters of the solution from the RT step were subsequently added to the PCR reaction containing 1× PCR buffer, 2.5 mM MgCl₂, 250 µM deoxynucleotide triphosphates (dNTPs), 1 µM forward and reverse LacZ primers, and 2.5 U Amplitaq Gold DNA polymerase (Perkin-Elmer) in a total volume of 20 µl. The cycling parameters were as follows: a total of 75 cycles, including denaturing at 94°C for 1 min and a combined annealing and elongation step at 72°C for 4 min, with a 1-s extension to the second step at every cycle. The RT-PCR experiments were performed with a DNA Thermal Cycler 480 (Perkin-Elmer). PCR products were subsequently size-fractionated on 1.5% agarose gels, stained with ethidium bromide and visualized under UV light.

Histochemistry

Lung and liver tissue from transgenic mice (T29-21 and T29-30) and control mice were fixed for 10 min in 0.5% glutaraldehyde in PBS, rinsed twice for 15 min in phosphate-buffered saline (PBS) containing 1 mM MgCl₂, and incubated in 1 mg of X-gal/ml, 5 mM K₃Fe(CN)₆, 5 mM K₄Fe(CN)₆, and 1 mM MgCl₂ in PBS for 17 h at 37°C. The samples were postfixed overnight at 4°C in 4% paraformaldehyde/0.5% glutaraldehyde in PBS according to the method of Engelhardt and colleagues (41). This was followed by embedding in paraffin. The tissue was cut in 8-µm sections and viewed under the microscope. X-gal staining was also performed, using lung and liver tissue from PB-treated mice (T29-30 transgene and control mice).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Isolation and Sequencing of CYP2B1 and 2B2 5' Flanking Region

By amplification of rat genomic DNA, plasmid clones containing the 5' flanking sequence of the CYP2B1 (Figure 1) and 2B2 genes were isolated and sequenced from exon 1 to position 1,334. On the basis of a high degree of similarity with the previously reported 5' flanking sequence of the CYP2B1 and CYP2B2 genes (18, 19, 31, 42, 43), two clones were classified as CYP2B2 and one as CYP2B1 (data not shown).

View larger version (28K): [in this window] [in a new window]   Figure 1.   Nucleotide sequence of the 1,334-bp 5' fragment of the CYP2B1 gene used for generation of transgenic mice. Where differences exist, the corresponding CYP2B2 sequence (43) is given below the CYP2B1 sequence. A dot indicates the absence of a base in either the CYP2B1 or CYP2B2 sequence. The CYP2B1 start site of transcription is indicated in boldface type. The 12-bp insert (19) and CA repeats (31) are boxed and marked, as are the binding sites for nuclear receptors (NR), BTE, and C/EBP (50, 52).

The sequence of the 2B1 fragment was almost identical to previously reported 2B1 5' sequences (18, 19, 31). Comparison of the 2B2 and 2B1 sequences isolated in the study to position 1,334 demonstrated 33 single-base substitutions, and one single-base (703) and one two-base (271) insertion in 2B1. The previously reported deletion of 26 bases in the repeated CA region (272) was identified (19). The number of CA regions in the repeated CA sequence at position 255 was observed to be five for 2B1, which is in accordance with previously reported results (31). We also found that the repeated CA sequences varied in length among the different 2B2 clones (from 19-21 CA; data not shown). In resemblance to what was reported by Shaw and coworkers (19, 31), an insertion of 12 bases (CTAACCCAGAGA) at position 706 was observed in the P450 2B1 sequence.

RNA blots from PB-treated and control rats were probed with either an exon 1 probe or a cDNA (PR-17) corresponding to exons 7-9 (Figure 2). The PR-17 cDNA recognizes the two major PB-inducible rat liver P450 cytochromes, P450 2B1 and P450 2B2, and it was used to detect P450 2B1-specific mRNA in the lung, since the rat lung is known not to express any P450 2B2. Comparison of the two RNA blots revealed highly similar patterns, indicating identical expression of exon 1 and exons 7-9 in both lung and liver. This indicated the absence of an alternative splicing or transcription start site. In noninduced (control) liver, the signal is almost nondetectable. In confirmation of earlier reports, PB induced mrRNA expression in liver only.

View larger version (91K): [in this window] [in a new window]   Figure 2.   Northern blot analysis of lung and liver RNA from PB-treated or control (C) rats. Filters were hybridized overnight with probes corresponding to CYP2B1 exon 1 (650-bp fragment cut with NcoI and PstI) and exons 7-9 (PR-17 cDNA [8]).

Generation of Transgenic Mice

In order to investigate the presence of regulatory sequences in the 5' flanking region of the CYP2B1 gene, a reporter gene construct containing 1,334 bp of the 5' sequence linked to the bacterial LacZ gene was made. The construct was functional and was able to drive expression of LacZ in both transfected primary rat hepatocytes and in the human hepatoma cell line HepG2 (data not shown). Transgenic mice were generated by introducing the gene construct into fertilized mouse oocytes. Four founders (two male and two female) were generated. Transmission of the transgene from two of the founders (T29-21 and T29-30) was confirmed. Both founders had integrated about the same number of copies of the transgene (four and five copies for the T29-21 and T29-30 founders, respectively; data not shown).

Tissue-Specific and PB-Inducible Expression of the Transgene

Results of RT-PCR analysis demonstrated the presence of transcribed mRNA corresponding to the CYP2B1/LacZ construct in lung tissue from the transgenic animals. The mRNA was seen as an amplified band of the expected 822-bp molecular weight (Figure 3). In the T29-21 transgenic offspring, this band was markedly stronger than in mice from the homozygous T29-30 line, indicating higher levels of transcription of the gene construct in lung tissue from the T29-21 animals. Omitting the reverse transcriptase enzyme from the RT-PCR procedure resulted in the absence of an 822-bp band. Importantly, no band was detected in liver tissue from transgenic mice, thus demonstrating the lung-specific transcription of the gene construct. Neither was any 822-bp band amplified with the use of RNA prepared from lung or liver tissue from the control mice (data not shown). However, when the experiments were performed with total cytoplasmic liver RNA from PB-treated transgenic mice (T29-30-derived homozygous mice), a band appeared with the expected molecular weight (Figure 3).

View larger version (37K): [in this window] [in a new window]   Figure 3.   Identification of RT-PCR products corresponding to the transcribed gene construct from lung and liver total RNA in transgenic and control mice. Lanes 1-2: Lung RNA from a T29-21 heterozygous transgenic mouse. Lanes 3-4: Liver RNA from a T29-21 transgenic mouse. Lanes 5-6: Liver RNA from a T29-30 homozygous transgenic mouse. Lanes 7-8: RNA from the mouse BalbMK keratinocyte cell line stably transfected with a LacZ expression vector. Lane 9: LacZ DNA as positive control. M: 100-bp DNA ladder. Inclusion of reverse transcriptase (RT) and treatment with phenobarbital (PB) is indicated below each lane.

Results of the histochemical staining for LacZ activity of lung and liver tissue from heterozygous T29-21-derived transgenic mice demonstrated a few positive cells in bronchiolar structures (Figures 4a and 4b), whereas no positively staining liver cells were observed (Figure 4d). In neither lung (Figure 4c) nor in liver sections (with or without PB treatment) from control mice or T29-30-derived transgenic mice were any X-gal-staining cells observed (data not shown), in accord with the generally lower transgene expression levels in the T29-30-derived transgenic line.

View larger version (118K): [in this window] [in a new window]   Figure 4.   X-gal staining of lung and liver sections from transgenic and control mice. (a) Demonstration of X-gal staining in bronchiolar structures in a T29-21-derived transgenic mouse, indicating selective expression of the transgenic construct. (b) Higher magnification of a. (c) Lack of X-gal staining in bronchiolar structures from a control mouse. (d ) Lack of X-gal staining in liver tissue from a T29-21- derived transgenic mouse.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Recent results have reported the sequence of the rat CYP2B1 promoter region from +27 to 3,878 bp (19) and of the CYP2B2 promoter region to 2,345 bp (31). Shaw and coworkers compared their sequence of the 5' flanking region of the 2B1 gene with that of the 2B2 gene as reported by Hoffman and associates (43), and found the two sequences to be almost identical up to 2,300 bp from where the 2B1 sequence diverged significantly from the remaining published sequence of 2B2. Our results from sequencing of the rat 2B2 and 2B1 promoter regions up to 1,334 bp were in agreement with published data (18, 19, 31, 42), since we observed overall sequence homology between the 5' flanking region of the two genes. Both the differing length of the repeated CA sequence at 255 bp and the previously identified 12-bp insertion at position 706 (19, 31) were observed in our CYP2B1 genomic clone.

We wanted to further elucidate the role of the 5' flanking region of the rat CYB2B1 gene in lung-specific transcription of CYP2B1, as well as in hepatic induction by PB. The mode of regulation of tissue- or cell-specific expression of the CYP2B1 and 2B2 genes is largely unknown. Available data have indicated different regulatory mechanisms. A hierarchy of CA genomic control elements, each interacting with specific transacting factors, could be envisioned. In this way, one control element would respond to induction and another to different tissue- or cell-specific factors. For the murine surfactant protein-A (SP-A) gene, CA elements important for epithelial lung cell-specific transcription have been identified between positions 255 and 57 (45). This region contained four nucleotide sequences similar to thyroid transcription factor-1 (TTF-1) binding motifs (45). Interaction of TTF-1 with the DNA binding sites located in the 5'-flanking region of the murine surfactant protein (SP)-A gene, and enhancement of lung epithelial cell-specific expression in vitro, was indicated (45). However, the role of TTF-1 as a regulator of other lung cell-specific proteins is unclear. A recent study using a reporter gene construct containing a Clara cell secretory protein (CCSP) promoter region segment identified as giving cell-specific expression of CCSP in H441 cells demonstrated high-level expression of the reporter gene in primary cultures of rat lung Clara cells (46). However, although previous studies had shown that expression of TTF-1 affected the rat CCSP promoter in an H441-cell model (47), and had demonstrated TTF-1 expression in Clara cells (46), its expression pattern did not show any clear correlation with the expression of CCSP in these Clara cells. In our gene construct, three potential TTF-1 binding sites were observed (45, 48, 49), but since they are also found in the 2B2 gene, which is not expressed in rat lung, it is unlikely that TTF-1 plays any primary role in expression of the CYP2B1/LacZ transgene in lung cells. Interestingly, the existence of relatively few possible binding sites for liver-enriched transcription factors in the promoter region of the CYP2B1 and 2B2 genes has been reported (19). This could account for the low constitutive expression of these genes in the liver. However, possible binding sites for gene-regulatory proteins, such as activator protein-1 and nuclear factor-kB, as well as possible signal-transducer and activator-of-transcription (STAT) sites, have been identified upstream of 800 bp in the CYP2B1 promoter (19), although their role in the tissue-specific expression of CYP2B1 is unknown.

Recently, the promoter regions of rat CYP2B1 and 2B2 were analyzed to 2,345 bp, using transient transfection of a hepatic cell line (31). The highest reporter gene activity was found with a deletion construct including only the proximal 177 bp of the 5' flanking region of the CYP2B1 gene, thus indicating the presence of upstream sequences with capacity to repress transcriptional activity of the CYP2B1 gene. This was consistent with the results of Park and colleagues, indicating a role for the proximal promoter region in transcription of the CYP2B1 gene (50). In the work of Park and colleagues, a deletion in the region between 110 bp and 57 bp in the CYP2B gene strongly reduced promoter activity, indicating strong CA elements in this region. The presence of sequences similar to a basal transcription element (BTE) between 82 bp and 67 bp (50) as well as a functional CCAAT/enhancer binding protein (C/EBP) site between 64 bp and 45 bp, were described. Mutation of these sequences in the CYP2B1 promoter region showed that these elements were involved in the regulation of transcriptional activity. A role for C/EBP in tissue-specific gene expression was previously demonstrated in hepatocytes (51), and a cooperative role for C/EBP and the transcription factor Sp1 in positive regulation of CYP2D5 has been suggested. If an Sp1-like protein binds to the BTE site of CYP2B1, it is possible that a similar interaction between Sp1 and C/EBP family members would be important for expression of CYP2B1 as well. However, although these BTE and C/EBP elements may be important for basal levels of transcription, it is again highly unlikely that they play a role in lung-specific expression of the CYP2B1 gene, since they are present in both the 2B1 and 2B2 promoter regions.

Previous results obtained with CYP2B2 transgenes to 800 bp and 2,300 bp revealed no pulmonary expression (20), whereas RT-PCR analysis revealed that our mice transgenic for 2B1 to 1,334 bp of the 5' flanking region exhibited lung expression of the gene and that hepatic tissue from PB-treated animals also expressed it. Thus, tissue-specific elements were present in our construct, but were lacking in the previously analyzed 2B2 regions. An interesting candidate element would be the 12-bp insertion at 706 bp (Figure 1). This element is lacking in the 2B2 promoter region but is relatively well conserved in the corresponding mouse 2b10 gene (eight of 12 bases), which is expressed in mouse lung (52). Furthermore, our results in the histochemical staining of lung and liver tissue indicated that pulmonary transcription of our CYP2B1 gene construct was restricted to only a few cells in the bronchioles of the transgenic mice. In these experiments we observed that the transgenic offspring of one of our founders, T29-21, apparently transcribed the gene construct at higher levels in their lungs than did the offspring of the other founder. However, a variegated pattern of transgene expression is a relatively common finding, and could be explained by surrounding chromatin structure or possibly by methylation (53, 54). On the basis of the staining pattern observed with X-gal, we suggest that transcription occurred in specific lung cells such as Clara cells. This is consistent with the relatively high level of expression of CYP2B1 found in rat and mouse Clara cells (53), although our results do not exclude the possibility that both mRNA and protein are expressed in other cells but at levels below the detection limits of the methods used.

The nature of PB-dependent regulatory elements in mammalian CYP genes is so far unresolved, although a role for CA sequences (Barbie-box-like sequences) close to the transcription start site has been implicated (24). However, other results have questioned the role of these sequences in the PB-induction response. With rat CYP2B2 transgenes (20), it has been shown that although CA elements in the proximal region interacted with PB-modulated proteins and could be involved in the PB-induction process (21, 29), these elements were not sufficient to mediate the induction response. However, these proximal elements could be part of the core promoter region, and not specific PB-induction enhancers or activators. Consistent with this, no major role of Barbie-box-like sequences in basal or PB-induced mouse Cyp2b10 gene transcription was found in primary hepatocytes (52), nor did mutation of similar sequences in the proximal CYP2B2 promoter affect the PB-inducibility of a reporter gene construct transiently transfected into rat liver cells (32). It has been proposed that the 2B2 gene is normally repressed by negative regulatory factor(s) localized between 800 bp and 20 kb, and that these factors could be modified or released by PB (20).

The presence of two PB-responsive regions in the CYP genes (33, 52) has been suggested. In the mouse 2b10 gene, these were located at approximately 2.3 kb and 1.0 kbp. In both the rat CYP2B2 and mouse Cyp2b10 genes, the former region includes a PB response enhancer module element (PBREM) (32). In the Cyp2b10 gene, a fully PB-inducible, 51-bp core enhancer element was identified within this region (35).

The upstream enhancer region at approximately 2.3 kbp contains an AGGTCA binding site for nuclear receptors and a TGGN-CCA region binding NF1/CEBP (55). These elements were found to be conserved in the upstream enhancer region of both the rat 2B1 gene and the rat 2B2 gene (52). In all three genes (2B1, 2B2, and 2b10), the proximal enhancer region at approximately 1.0 kbp contains an AGGTCA element but not the NF1/CEBP  binding sequence. Thus, our construct contains this proximal PB-responsive region and is PB responsive, indicating that the proximal enhancer also has a role in the PB-induction response. This is consistent with the fact that the previously reported 800 bp 2B2 transgene is without PB responsiveness (20). Regarding the two PB-responsive regions in the CYP genes, it is interesting to observe that the upstream but not the proximal enhancer is functional with a heterologous promoter in primary hepatocytes (52). It is tempting to speculate that the upstream enhancer is functionally independent because of the presence of both an AGGTCA and a CEBP binding site, whereas the proximal enhancer may require interaction with the CEBP site close to the promoter, and is therefore not PB inducible when linked to a heterologous promoter.

In conclusion, our results show that the region up to 1.3 kb in the 5' flanking region of the rat CYP2B1 gene includes sequences important for lung-specific transcription of this gene, as well as hepatic induction of CYP2B1 transcription by PB. However, the expression levels of the gene were low, indicating that additional elements, present further upstream or downstream, are necessary for efficient transcription.