Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

Hepatic cytochromes P450 belonging to the CYP2B subfamily metabolize a wide variety of xenobiotics as well as endogenous substances such as steroids and fatty acids (Guengerich, 1987; Oguri et al., 1994). Of members in the CYP2B subfamily, rat CYP2B1 and CYP2B2 are known to be highly inducible by many compounds that have divergent chemical structures (Waxman and Azaroff, 1992) and have been studied extensively both toxicologically and biologically. The mechanism by which PB induces the CYP2B subfamily P450 is not completely understood (for recent reviews, see Savas et al., 1999; Waxman, 1999). However, recent results from Negishi and colleagues (Honkakoski et al., 1998) have demonstrated that a heterodimer of nuclear receptors CAR and RXR plays a crucial role in the regulation of the CYP2B subfamily and that PB recruits the CAR from cytosol to nucleus by mechanism(s) that may involve a protein dephosphorylation event (Kawamoto et al., 1999). The transcriptional activation process likely involves CAR/RXR interaction with a specific region of 5'-flanking gene sequence designated as the PBREM. A recent report using a transgenic mouse model demonstrated that mutation of the central NF1 motif within the CYP2B2 PBRU had no major effect on the PB induction process (Ramsden et al., 1999). However, the role of potential protein-protein interactions within the PBRU/PBREM still needs to be explored in more detail. [A specific region located in 5'-upstream of the rat CYP2B2 gene that is suggested to play a role in PB-mediated induction was designated the PBRE (Trottier et al., 1995) or PBRU (Stoltz et al., 1998). The analogous region found in mouse Cyp2b10 and human CYP2B6 genes was designated the PBREM (Honkakoski and Negishi, 1997). In this article, we refer to this region in the CYP2B2 gene as the PBRU.]

Hashimoto et al. (1988) have reported that SD (Qdj:SD) rats being bred in the Kyushu University animal colony were an abnormal strain because CYP2B2 was not induced after PB treatment, although CYP2B1 was normally PB-responsive. These investigators have also observed that the lack of CYP2B2 mRNA elevation by PB pretreatment suggested a transcriptional defect, that the phenotype lacking in response to PB-mediated induction appeared to be recessive trait caused by a single gene mutation, and that there was no mutation and/or deletion in the proximate 0.8 kbp of the gene 5'-flanking sequence. This animal model may prove to be a valuable tool for clarifying mechanisms underlying PB-mediated induction. The mechanisms accounting for the lack of CYP2B2 induction in these animals remain unknown. One possibility is that the CYP2B2 gene in Qdj:SD rats has a defect(s) in its exon structure, resulting in the production of abnormal protein. Therefore, a primary objective of this study focused on this latter issue, and we have completed sequencing of the exonic regions. Because the PBRU is likely to be important for PB responsiveness, we also determined the sequence of the CYP2B2 gene's upstream region containing this response element.

CYP2B1 and CYP2B2 are closely related enzymes with similar catalytic function (Guengerich, 1987; Funae and Imaoka, 1993). Because many PB-like inducers reported thus far increase both P450s, the specific contribution of each isoform to the drug-metabolizing process has not been fully clarified. The Qdj:SD rats appeared to represent an appropriate animal model for addressing this issue. Therefore, in the current investigation we also estimated the function of hepatic CYP2B P450s in PB-treated Qdj:SD rats. The results suggest that CYP2B1 synergizes with CYP2B2 in eliciting its maximal function.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. The chemicals described below were purchased from the sources indicated: sodium PB (Tokyo Chemical Industry, Co. Ltd., Tokyo, Japan), 4-hydroxybiphenyl (Wako Pure Chemical Industries, Co., Ltd., Osaka, Japan), and morphine hydrochloride (Takeda Chemical Industries, Ltd., Osaka, Japan). Standards of hydroxytestosterones were kindly donated by Dr. T. Baba, Shionogi Pharmaceutical Co. (Osaka, Japan). Rabbit anti-CYP2B1/2 antibody was prepared in this laboratory. Anti-CYP3A antibody, which cross-reacts with CYP3A1 and CYP3A2, was purchased from Daiichi Pure Chemicals, Co. (Tokyo, Japan). [It is well established that a major glucocorticoid-inducible isoform of the CYP3A subfamily is CYP3A1. Later it was suggested to designate this form CYP3A23 (Komori and Oda, 1994). However, to avoid confusion, we refer to the PB-inducible isoform using the more conventional term CYP3A1 in this article, consistent with the terminology used in a large number of previous reports.]

Animals, Treatment, and Preparation of Microsomes and Genomic DNA. Male Qdj:SD rats weighing 120 to 150 g were obtained from the animal facility at Kyushu University. This strain was being bred by crossing female and male parental rats randomly selected from a colony. Reference animals, male Crj:SD rats, with the same body weight as the Qdj:SD rats were purchased from Charles River Japan (Kanagawa, Japan). Rats were treated (i.p.) once a day with sodium PB at a dose of 80 mg/kg/ml saline for 4 consecutive days. Control rats were given saline alone. The animals were fasted overnight after the last treatment, and the livers were removed. A portion (1 g) of each liver was stored at 80°C, and genomic DNA was extracted using established methods (Maniatis et al., 1989). Microsomes were prepared from the remaining liver using methods described elsewhere (Yamada et al., 1993) and stored at 80°C.

Immunoblotting. Sodium dodecyl sulfate-polyacrylamide gel (7%) electrophoresis was performed according to Laemmli's procedures (1970). In all experiments, hepatic microsomes consisting of 9 µg of protein were loaded and electrophoresed. The proteins in the gel were transferred electrically to a polyvinylidene difluoride membrane using reported methods (Towbin et al., 1979) and then reacted with rabbit anti-CYP2B1/2 antibody. The band associated with the antibody was visualized using the method of Blake et al. (1984), using alkaline phosphatase-conjugated anti-rabbit IgG antibody.

Assays and Statistical Analysis. Hepatic microsomal activities of testosterone metabolism (Shinohara et al., 1997), 4-hydroxybiphenyl glucuronidation (Bock et al., 1979), and morphine-3-glucuronidation (Hanioka et al., 1990) were determined using the methods indicated. Protein contents of the microsomal fractions were determined according to established methods (Lowry et al., 1951). Statistical differences were calculated using ANOVA with the post hoc test (Fisher's protected least-significant difference method).

PCR Amplification of CYP2B1/2 Genes. Amplification of the CYP2B1 gene was performed by dividing the gene into three segments (2B1/12 kbp, 2B1/6 kbp, and 2B1/7 kbp; Fig. 1A). Similarly, four segments of the CYP2B2 gene (2B2/PBRU, 2B2/12 kbp, 2B2/4 kbp, and 2B2/0.6 kbp; Fig. 1B) were separately amplified. Primer pairs used in these PCRs are listed in Table 1. In a final volume of 50 µl, the reaction mixture consisted of 500 ng of genomic DNA; 200 nM each DNA primer; 2.5 mM MgCl₂; 0.4 mM each deoxynucleotide triphosphate; 2.5 U of Taq-polymerase (TaKaRa LA Taq; TAKARA Shuzo Co., Ltd., Shiga, Japan); and 5 µl of reaction buffer (10 × LA PCR Buffer II) supplied with the Taq-polymerase kit. The reaction mixture was added to a 500-µl sterilized plastic tube with a cap, and a few drops of mineral oil (Sigma Chemical Co., St. Louis, MO) were layered onto the solution followed by cycling in a PCR tempcycler (ASTEC PC-700, ASTEC, Fukuoka, Japan). The conditions of PCR were as follows: 2B1/12 kbp, 2B1/6 kbp, 2B1/7 kbp, 2B2/12 kbp, and 2B2/4 kbp, 98°C for 1 min (94°C for 30 s and 64°C for 20 min) for 35 cycles and 72°C for 10 min with a hold at 4°C; 2B2/0.6 kbp, 98°C for 1 min (94°C for 30 s, 58°C for 1 min, and 65°C for 5 min) for 30 cycles and 72°C for 10 min with a hold at 4°C; and 2B2/PBRU, 98°C for 1 min (94°C for 30 s, 56°C for 1 min, and 65°C for 5 min) for 30 cycles and 72°C for 10 min with a hold at 4°C. In the experiment for nucleotide sequencing, the PCR product was purified using a commercial column (GFX PCR DNA and Gel Band Purification kit; Amersham Pharmacia Biotech Inc., Piscataway, NJ).

View larger version (20K): [in this window] [in a new window]   Fig. 1.   PCR amplification of the CYP2B1 (A) and CYP2B2 (B) genes. One of each of the primer pairs is specific either to the CYP2B1 or to the CYP2B2 gene and selectively amplifies the target gene. See Table 1 for additional information and the sequences of the primers. All PCR products except "CYP2B1/12 kbp" were confirmed as the desired products by sequence analysis of the gene exons that contain diagnostic nucleotide substitutions distinguishing CYP2B1 and CYP2B2 (Suwa et al., 1985). There is a major difference in the length of first intron between CYP2B1 and CYP2B2 (Suwa et al., 1985). The 12-kbp intron 1 region of CYP2B1 identified amplified products of CYP2B1 gene.

Nucleotide Sequencing. Exonic sequences of CYP2B1/2 genes and a region containing the PBRU of the CYP2B2 gene were determined by sequencing the PCR products shown in Fig. 1 and Table 1. The intronic sequences of the CYP2B1/2 genes have not been determined previously. In parallel with the current study, we sequenced the CYP2B2 gene cloned from an SD rat genomic library (pWE'-39E) (Ramsden et al., 1993) and determined the area containing 7 kbp of 5'-upstream region together with the exons/introns (S. Matsumoto, M. Yamamoto, H. Yamada, C. J. Omiecinski, and K. Oguri, unpublished data). Some of the sequencing primers used in this investigation were synthesized based on this information (see legends to Tables 1 and 2). To minimize possible PCR artifacts, the PCRs listed in Table 1 were performed at least four times for each region, and these products were subjected separately to DNA sequencing.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Induction Pattern in Qdj:SD Rats and Phenotyping. The induction pattern of PB-inducible P450s and UGT in Qdj:SD rats was compared with those in Crj:SD rats, a reference animal, by determining hepatic microsomal activity of metabolism of testosterone, 4-hydroxybiphenyl, and morphine (Table 3). In a reference animal (Crj:SD rats), PB treatment enhanced testosterone 6- and 16-hydroxylase activities, the markers for the CYP3A and -2B isoforms, respectively. Testosterone 6-hydroxylase activity was also increased by PB treatment in all Qdj:SD rats, whereas for the 16-hydroxylase, only a minor increase was observed in one (rat 3) of four individual rats. The activities of 4-hydroxybiphenyl and morphine glucuronidation were increased with PB in both strains of rats including rat 3 of the Qdj:SD rats. These results suggest that some individual Qdj:SD rats lack the ability to induce the CYP2B isoform in response to PB treatment as reported previously (Hashimoto et al., 1988), that the whole colony of these rats is not genetically homogeneous, and that impaired induction is limited to the CYP2B subfamily.

View larger version (47K): [in this window] [in a new window]   Fig. 2.   Interindividual difference in the expression of CYP2B1/2 (A) and CYP3A (B) subfamilies in Qdj:SD rats pretreated with phenobarbital. Ten rats were randomly chosen from a breeding colony and treated with PB, and their contents of hepatic P450s were examined by immunoblotting. In A, hepatic microsomal activity (nanomoles per minute per milligram of protein) of testosterone 16-hydroxylase was also shown. Each bar represents the mean of two determinations. Based on the difference in the intensity of the CYP2B2 immunoblot band and testosterone hydroxylase activity, 10 rats were phenotyped as shown at the top of the figure (see text for details).

View larger version (67K): [in this window] [in a new window]   Fig. 3.   Hepatic microsomal activity of testosterone and 4-hydroxybiphenyl metabolism in Qdj:SD (+/+), (+/), and (/) rats pretreated with phenobarbital. Hatched bars [(+/+), rats 1, 3, 7, and 10], open bars [(+/), rats 2, 5, and 9], and half-closed bars [(/), rats 4, 6, and 8] are the mean ± S.E. of three or four individual rats indicated. See Fig. 2A for phenotyping. *, significantly different from +/+ (P < .05). , significantly different from +/ (P < .05).

Structural Analyses of CYP2B1/2 Genes. The size and sequence of CYP2B2 and -2B1 genes in Qdj:SD rats were compared with those in Crj:SD rats. To this end, the CYP2B1/2 genes were selectively amplified by PCR, by using specific primers for each gene (Table 1). Because the sequence homology between CYP2B2 and -2B1 genes decreases markedly in the upstream region far from 2.3 kbp, primer G is highly specific for the amplification of "2B2/PBRU". A number of other primers were designed by choosing regions in which differences exist between CYP2B1 and -2B2 genes (see the legend to Table 1). Two PCR products from the CYP2B2 gene, 2B2/12 kbp and 2B2/4 kbp (Fig. 1), were the same length in both Qdj:SD (/) and Crj:SD rats (Fig. 4). The identity of these products as CYP2B2, and not as CYP2B1, was confirmed by subjecting the samples to sequencing of the exonic regions (data not shown). This observation indicates that the CYP2B2 gene of Qdj:SD (/) rats does not have any large deletion and/or insertion. To further probe potential mechanisms accounting for the absence of the CYP2B2 gene product in Qdj:SD (/) rats after PB treatment, the sequences of the upstream regulatory regions as well as all exonic sequences were determined (Table 4). Interestingly, the sequence spanning 2319 to 1640 bp and containing the PBRU (²³¹⁵GATCGTGGA ... CTGGGTGATC²¹⁵³) (Trottier et al., 1995; Stoltz et al., 1998) was identical between Qdj:SD (/) and Crj:SD rats. As for the exons, seven nucleotide alterations were detected in comparison with the sequences reported previously. However, there were no sequence differences between mutant-type and wild-type Qdj:SD rats. None of seven base changes coded for stop-codons, and four were silent mutations. Three mutations coded for amino acid substitutions, with two (mutations at third and ninth exons) located in substrate recognition sites 1 and 6, respectively. The GT-AG rule was confirmed in each exon-intron junction (data not shown).

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Agarose gel electrophoresis of the PCR product amplified for the CYP2B2 gene present in Qdj:SD (/) and Crj:SD rats. See Fig. 1 and Table 1 for the region amplified. A portion of the PCR reaction mixture was loaded, electrophoresed in a 0.6% agarose gel, and stained with ethidium bromide.

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

Since its discovery, the abnormal PB-CYP2B phenotype has been assumed to be present in the whole colony of Qdj:SD rats at Kyushu University. However, during breeding for the past 12 years, mating between mutant-type rats and rats with normal character would have occurred. Breeding experiments by the discoverers of this phenomenon indicated that the phenotype of the offspring (filial generation 1, or F₁) generated by mating Qdj:SD rats with normal SD rats was normal (Hashimoto et al., 1988). Furthermore, they reported only two phenotypes in the F₂ generation. From these observations, the mutant character in the Qdj:SD rats had been suggested to be a recessive trait. However, an intermediate phenotype was observed in the colony of Qdj:SD rats analyzed in this study. The reason for the inconsistency between these studies is unknown, but it might be attributable to the difference in the sensitivity and/or accuracy of the phenotyping methods; e.g., the earlier researchers phenotyped rats by determining CYP2B2 mRNA levels after PB treatment using assays that might not have been highly quantitative.

The coding regions of the CYP2B2 gene in the mutant-type Qdj:SD rats were sequenced in this study, but no defects were discovered that would be anticipated to have any deleterious consequences. In addition, conservation of the GT-AG rule was detected in all exon-intron junctions. In general, the terminal base of the exon is G, which links to the GT of the intron. The structure of the CYP2B2 gene reported agrees with this rule (Suwa et al., 1985). At the end of fifth exon of the CYP2B2 gene in Qdj:SD (/) rats, G was mutated to A (⁸²²G  A; nucleotide number is counted from the protein coding site of cDNA) (see Table 4). However, genes not adhering to the ... G-GT... junction structure have also been observed (Breathnach and Chambon, 1981). Furthermore, the G/A mutation at the terminal base of the fifth exon was also observed in the Qdj:SD (+/+) rats in which CYP2B2 is normally induced using PB. The observation made here strongly suggests that the structural gene in Qdj:SD (/) rats codes an expressible and functional CYP2B2. A previous report suggesting that small amounts of CYP2B2 protein exist in the liver of PB-pretreated Qdj:SD rats supports this view (Hashimoto et al., 1988). Mutant rats other than the Qdj:SD strain having an abnormality in CYP2B2 expression have already been reported (Rampersaud and Walz, 1987). A large deletion in the structural gene of CYP2B2 has been suggested as the reason for the impaired expression in these rats (Omiecinski et al., 1992). Therefore, the mechanism causing a lack in PB-mediated induction of CYP2B2 is quite different between Qdj:SD rats and other mutant rats reported. Sequencing experiments also failed to detect mutations in the PBRU of Qdj:SD (/) rats. From the evidence presented here, the impaired induction of the CYP2B2 in Qdj:SD (/) rats is considered likely to involve one of the following mechanisms: 1) mutations within 5'-upstream gene regions not examined here or mutations within introns; or 2) absence or lowered expression of trans-acting factor(s) necessary for gene regulation. If the former is the case, then elements other than the PBRU would in effect be needed in PB-mediated induction. An alternative consideration is whether trans-factors bound to the PBRU may interact with other trans-factors associated with DNA at more distal sites. Perhaps different trans-factor(s) are involved in the induction of CYP2B1 versus CYP2B2, because Qdj:SD (/) rats exhibit normal PB-mediated induction of the CYP2B1 gene. It has been demonstrated that the RXR-pregnane X receptor complex plays an important role in the induction of the CYP3A isoforms (Kliewer et al., 1998; Pascussi et al., 1999). As shown in this investigation, CYP3A1 and CYP3A2 are normally induced using PB treatment in Qdj:SD (/) rats. Thus, this observation does not support another possible mechanism, that defects in RXR function are involved in the impaired induction of CYP2B2 in these rats.

Abnormal expression and induction was rather specific to the CYP2B2 in Qdj:SD (/) rats because other PB-inducible enzymes including CYP2A and -3A subfamilies and UGT(s) were shown to be expressed normally. However, in the progress of this study, we noted a possibility that CYP2B1 induced with PB does not act effectively or has a defect in its function in Qdj:SD (/) rats; that is, the activity of testosterone 16-hydroxylase in PB-treated Qdj:SD (/) rats was shown to be quite low even though the CYP2B1 protein level was increased markedly using PB treatment. Sequencing experiments detected some nucleotide substitutions in the CYP2B1 exons of these rats. However, it is unlikely that these substitutions are responsible for the impaired function. Of the two nonsynonymous mutations detected, one (²⁸²Glu  Val) was located near substrate recognition sites 4 (Goto, 1992). However, this mutation is common in both Qdj:SD (/) and (+/+) rats, and liver microsomes from the (+/+) rats exhibit high levels of testosterone 16-hydroxylation, as high as Crj:SD rats. Based on the evidence from sequencing, the CYP2B1 protein expressed in Qdj:SD rats is suggested to have no defect in its primary structure. Our results may suggest that the CYP2B1-CYP2B2 interaction is needed for the occurrence of maximal catalytic function of the CYP2B1. Although this is speculative, the reason for the low activity of testosterone 16-hydroxylase in Qdj:SD (/) rats might be attributable to a decreased interaction between CYP2B1 and CYP2B2, resulting from a lowered expression of the latter enzyme. Early studies reported that purified CYP2B1 reconstituted in simple liposome exerted the catalytic function in the absence of the CYP2B2 (Ryan et al., 1979). This observation seems to disagree with the above assumption. On the other hand, it has been reported that the cytosolic domain of P450 is pulled into liposome by some kinds of phospholipid such as phosphatidic acid (Ahn et al., 1998). The secondary structures and the function of the P450 enzyme deeply embedded into membrane are shown to be different or enhanced from those in liposome consisting of one kind of phospholipid (Ahn et al., 1998). Thus, it seems likely that 1) the CYP2B1 in endoplasmic reticulum membrane differs from that in artificial liposome in its structure and function, and 2) the function of the CYP2B1 is altered by interaction with the partner proteins as well as by the composition of the phospholipid in membrane. To our knowledge, functional cooperation of CYP2B1 and CYP2B2 has not been reported. However, different electrophoretic mobility of CYP2B1 between (/) and (+/+) Qdj:SD rats may suggest the possible interaction of 2B1 and 2B2; that is, CYP2B1 accompanied by the coexpression of CYP2B2 in the (+/+) phenotype migrated faster than the CYP2B1 in (/) rats that lacked CYP2B2 (see Fig. 2A). It has been suggested that CYP2B4 and CYP1A2 interact with each other, affecting the function of the partner (Backes et al., 1998). In the latter case, CYP1A2-NADPH cytochrome P450 reductase interaction is facilitated by the binding of CYP2B4 to CYP1A2 (Backes et al., 1998). A substrate-induced change in CYP2A6-CYP2E1 interaction has been reported (Tan et al., 1997). Furthermore, Yamazaki et al. (1997) demonstrated the activation of the catalytic function of CYP3A4 by CYP1A2. These data may support the possibility of the CYP2B1-CYP2B2 interaction described above.

In conclusion, this study could not detect any serious mutations in the exons of CYP2B2 gene, suggesting that this gene codes expressible protein, nor were any mutations detected at the PBRU, which has been demonstrated to play a critical role in the PB induction process. Although not conclusive, the available evidence suggests that the Qdj:SD rats lack a PB-mediated induction of CYP2B2 may be attributable either to mutations within a regulatory region of the gene different from the PBRU or to the absence/lowered expression of trans-acting factor(s) cooperatively involved in the induction process.