Isolation and characterization of rat CYP11B genes involved in late steps of mineralo- and glucocorticoid syntheses.

We isolated and characterized four forms of rat CYP11B genes, which were tentatively named CYP11B1, -B2, -B3, and -B4. Genomic Southern analyses indicated that the members of the rat CYP11B gene subfamily were confined to these four genes; among them, CYP11B1 and -B2 encoded steroid 11 beta-hydroxylase and aldosterone synthase, respectively, while CYP11B3 was a gene highly homologous to CYP11B1 without a known expression product. By being devoid of a region spanning two exons conserved in the other three, CYP11B4 was presumably a pseudogene. In the nucleotide sequences, CYP11B1, -B3, and -B4 showed 95-96 and 93-100% identities in the coding and 0.5-kilobase 5'-flanking regions, respectively. However, the homology between the nucleotide sequences of one of the three and CYP11B2 was rather low, about 90 and 50% in the coding and 0.5-kilobase 5'-flanking regions, respectively. As a whole, CYP11B2 rather than CYP11B1, -B3, or -B4 was more homologous to CYP11B genes of other animals such as cow and human. In transient transfection experiments using mouse adrenocortical Y1 cells and chloramphenicol acetyltransferase gene constructs, the 0.5-kilobase 5'-flanking region of CYP11B1 had a 4- and 10-fold higher promoter activity than the corresponding regions of CYP11B2 and -B3, respectively. The possible presence of a suppressive element(s) was noted in the upstream of the 0.5-kilobase region of CYP11B1. Although a variant of cAMP-responsive element, which was present in rat CYP11B2 and all known CYP11B genes of other animals, was modified in rat CYP11B1 and -B3 genes, dibutyryl cAMP stimulated all the promoter activities of the 5'-flanking regions of the rat genes by 3-fold.


Isolation and Characterization of Rat CYP11B Genes Involved in Late
Steps of Mineralo-and Glucocorticoid Syntheses* (Received for publication, October 6, 1992) Kuniaki Mukai, Michiyo Imai, Hideo Shimada, and Yuzuru Ishimurat From the Department of Biochemistry, School of Medicine, Keio University, 35 Shinanomachi, Shinjuku-ku, Tokyo 160, Japan We isolated and characterized four forms of rat CYPllB genes, which were tentatively named CYP11B1, -B2, -B3, and -B4. Genomic Southern analyses indicated that the members of the rat CYPllB gene subfamily were confined to these four genes; among them, C Y P l l B l and -B2 encoded steroid 11shydroxylase and aldosterone synthase, respectively, while CYPllB3 was a gene highly homologous to C Y P l l B l without a known expression product. By being devoid of a region spanning two exons conserved in the other three, CYPllB4 was presumably a pseudogene. In the nucleotide sequences, C Y P l l B l , -B3, and -B4 showed [95][96] and 93-100% identities in the coding and 0.5-kilobase 5"flanking regions, respectively. However, the homology between the nucleotide sequences of one of the three and CYPllB2 was rather low, about 90 and 50% in the coding and 0.5-kilobase 5"flanking regions, respectively. As a whole, CYPllBP rather than CYP11B1, -B3, or -B4 was more homologous to CYPllB genes of other animals such as cow and human.
In transient transfection experiments using mouse adrenocortical Y 1 cells and chloramphenicol acetyltransferase gene constructs, the 0.5-kilobase 5"flanking region of C Y P l l B l had a 4and 10-fold higher promoter activity than the corresponding regions of CYPllB2 and -B3, respectively. The possible presence of a suppressive element(s) was noted in the upstream of the 0.5-kilobase region of CYP11B1. Although a variant of CAMP-responsive element, which was present in rat CYP11B2 and all known CYPllB genes of other animals, was modified in rat CYPllBl and -B3 genes, dibutyryl CAMP stimulated all the promoter activities of the 5"flanking regions of the rat genes by 3-fold.
Biosynthesis of mineralo-and glucocorticoids in the adrenal cortex are regulated by many factors including adrenocorti-* This work was supported in part by grants-in-aid for scientific research on priority areas (63635005 and 032411031, for general scientific research (03670148), and for encouragement of young scientists (04780210) from the Ministry of Education, Science and Culture of Japan, and by grants from the Takeda Science Foundation and from Keio University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertlsement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleotide sequence(s) reported in thispaper h been submitted to the GenBankTM/EMBL Data Bank with accession number(s)
$ T o whom correspondence should be addressed. Tel.: 011-81-3-014086-014108. 3355-2827;Fax.: 011-81-3-3358-8138. cotropic hormone (ACTH)' and angiotensin I1 (1)(2)(3). The syntheses are also zone-specific; glucocorticoids are mainly produced in the zonae fasciculata-reticularis, while mineralocorticoids are formed in the zona glomerulosa. In rat, production of corticosterone, the primary glucocorticoid in this animal species, is controlled by ACTH, while that of aldosterone, the major mineralocorticoid, is regulated by a variety of factors such as the renin-angiotensin system, extracellular potassium, and ACTH. Both corticosteroids are synthesized from a common precursor, deoxycorticosterone, in respective zones. Such zone specificities were recently ascribed to the localization of two distinct homologues of cytochrome P450 (P450) enzymes, P450118 and P450aLd,, which are responsible for the formation of corticosterone and aldosterone, respectively (4)(5)(6)(7)(8)(9)(10)(11)(12)(13). Similarly, two enzymes were also found in mouse and human (14)(15)(16)(17)(18)(19), but the situation appears to be different in cow and pig (20)(21)(22)(23)(24)(25)(26). In the latter animals, only one kind of P450 enzyme, commonly referred to as P45Oll8, exists throughout the three zones and exerts a catalytic activity specific to each zone. The P45011a produces aldosterone from deoxycorticosterone in the zona glomerulosa, while the same enzyme in the zonae fasciculata-reticularis catalyzes the conversion of deoxycorticosterone to corticosterone but not to aldosterone. Thus, two modes exist for the corticosteroidogenesis in mammals, one is the rat type and the other is the bovine type (6).
In an attempt to clarify the regulation mechanisms of corticosteroidogenesis, we have isolated and characterized from rat the genes for P450118 and P450ald, (CYP11B1 and -B2, respectively, according to the standard nomenclature (27)) together with two additional genes, termed here CYP11B3 and -B4. CYPllB3 was a gene whose transcript is unknown, while CYP11B4 was likely to be a pseudogene. Nucleotide sequences of the 5"flanking regions of CYPllBl and -B2 were significantly different from each other, suggesting differences in their transcriptional regulation. A preliminary account of a portion of this work has appeared (28).
Transient Transfection and CAT Assay-5 X lo6 Y1 cells in a subconfluent state were mixed with 20 pg of plasmid DNA in 0.5 ml of a buffered saline solution (32). After preincubation on ice for 10 min, the DNA was introduced by an electric pulse (31) at a capacitance of 500 microfarads at 400 V in a cuvette with an electrode distance of 0.4 cm using a Bio-Rad electroporation apparatus. The cells were then incubated on ice for 10 min and were diluted with 10 ml of the cell culture medium. They were cultured in a 100-mm dish for 66 h at 37 "C and harvested for preparation of cell extract. When necessary, dibutyryl CAMP (1 mM), porcine ACTH (1 pM), human angiotensin I1 (1 p~) , or [Sari-Ala'langiotensin 11 (1 p M ) was added 24 h after the transfection. All these reagents were purchased from Sigma. CAT assay was performed using [14C]chloramphenicol (53 mCi/mmol, Amersham Corp.) as described elsewhere (33). Transfection experiments were performed at least 4 times using different preparations of plasmid DNA. Protein concentration of cell extracts was determined by the Bradford method (34).

Four Forms of CYPllB
Genes on Rat Genome-A rat P450,ldo cDNA containing the whole protein coding sequence (1.5 kb) was used as the probe for the isolation of genomic clones containing CYP11B genes from a rat genomic library. This probe was expected to hybridize with both P45Oald,, and P450110 genes, because their cDNA sequences had 88% identity with each other (7)(8)(9). Positive clones were analyzed by restriction mapping and also by hybridization with an oligonucleotide probe specific to either P45o,ld, or P45OIl0 gene. Then, four clones with different restriction maps were selected from the positive clones: one containing the CYPllB2 gene, which was hybridized with the P450ald, gene-specific probe, and the other three containing CYP11B1, -B3, and -B4 genes, respectively, which were hybridized with the P45OIl0 genespecific probe (Fig. 1). To determine whether there were more genes closely related to CYPllB than these four, genomic Southern blot hybridization was performed (Fig. 2). Two probes, which recognized different regions on the CYPllB genes, were used to hybridize genomic fragments in EcoRI, BamHI, or HindIII digest. One probe contained a rat P450,1d, cDNA fragment from the 5'-untranslated region to Sau3AI site a t 162, and the other contained a region of the cDNA from PstI site at position 1241 to BamHI site at 1820 in the 3"untranslated region (8,9). The results in combination with the four restriction maps of the isolated genomic clones in Fig. 1 indicated that all hybridizing fragments in EcoRI, BarnHI, and HindIII digests were assignable to restriction fragments predicted from the four genomic clones (see legend for Fig. 2). We, therefore, conclude that there are no additional genes closely related to CYPllB in the rat genome.
DNA Sequence Analysis of CYPIlB Genes-Comparison of the genomic DNA sequences with those of cDNA for P45OI1, and P450alda revealed that the coding sequences of CYPllBl and -B2, respectively, were identical to those of the cDNA for P4501,, and P450.ldO. This finding together with the results of Southern blot analyses (Fig. 2) established that the genomic Genomic DNA (10 pg) prepared from liver was digested with EcoRI (lanes E ) , BamHI (lanes B), or HindIII (lanes H). Two cDNA fragments for rat P450ald, that corresponded to CYPllB2 were used as probes: one, a cDNA fragment containing a portion of the exon 1 (lejt panel), and the other, another region of the cDNA consisted of the exons 8  clones we isolated contain the genes for P450110 and P450.ldO. The third one, CYP11B3, had a similar structure to those of CYPllBl and -B2. All of them contained nine exons, and the locations of all introns corresponded exactly to each other. Furthermore, the exonic sequences of CYP11B3 were found to contain no in-phase termination codon except for the conserved TAG codon, which terminates translation for CYPllBl and -B2 gene products. However, no CYP11B3 . gene product(s) including the transcript has hitherto been reported. The reason for this phenomena will be discussed later. On the other hand, the CYP11B4 gene lacked a sequence corresponding to the region encompassing exon 3 and a part of exon 4 present in the other three genes. Thus CYP11B4 is likely to be a pseudogene. The coding sequences of CYP11B1, -B2, and -B3 as well as the sequences in -B4 corresponding to the coding regions of the others are shown in Fig. 3.
The coding sequences of CYPllBl and -B2 were found to be homologous by 88%, while the homologies between CYP11B2 and -B3 and between CYPllBl and -B3 were 89 and 96%, respectively. When the deleted coding sequences in CYP11B4 were not taken into calculation, the coding sequences had 95, 91, and 95% nucleotide identities with those of the CYP11B1, -B2, and -B3 genes, respectively. Especially in exon 5, the homology of CYPllBl and -B2 was 65%, being the lowest among the nine exons, whereas those of CYPllB3 and -B4, respectively, were identical and homologous by 97% to that of CYP11B1. All exon-intron boundaries were consistent with the consensus splicing sequence except that the 5'-end of the third intron of CYPllB3 was GC instead of the GT consensus dinucleotide (Table I). In CYP11B4, a 589nucleotide sequence between the end of exon 2 and the position corresponding to the 41st nucleotide of exon 4 contained no homologous sequence to exons 3 and 4 of the other three genes. At least 180 nucleotides from the 3'-end of exon 2 of CYP11B4 were identical to the corresponding intronic sequence of CYP11B1. The cysteine residues thought to be the fifth ligand to heme were conserved in the four genes at the nucleotide position of 1336 for CYPllBl and the corresponding positions for the other three genes. Fig. 4 shows the sequences of 5"flanking regions of the four CYPllB genes. The homology of the 0.5-kb 5"flanking regions of CYPllBl and -B2 was as low as approximately 50%. In further upstream regions up to 1.7 kb, the homology was approximately 45%. The low homologies seemed to be consistent with the observed differences in gene regulation of the two genes in vivo (1)(2)(3). The 490-nucleotide sequence of the 5'-flanking region of CYP11B3 and the 492-nucleotide sequence of that of CYP11B4, respectively, were 93% homologous and identical to the corresponding region of CYP11B1. All four 5"flanking regions contained a common sequence, GATAAAA, a derivative of the ,TATA box, defining the transcription initiation sites in CYPllBl and -B2 as described in our preliminary report (28). Also, we have reported that CYPllBl had minor transcription initiation sites in at least the 200-nucleotide upstream region of the major transcription initiation site (28). Neither typical CCAAT box nor GC box was found in the four 5'-flanking regions.

Upper and lower case letters represent exon and intron sequences, respectively. Identical nucleotides of CYP11B2, -B3, and -B4 with that of -B1 in the alignments are indicated by dots. Introns which were completely sequenced are marked by an asterisk. CYPllB4 lacked exon 3 and a part of exon 4 as indicated by dashes.
Gene Exon no.

3'-Exon junction
Intron size 5'-Exon junction Exon no.     .... ....  ogies to those of CYPllB genes of other animal species than those of rat CYP11B1, -B3, and -B4 genes. A modified CAMPresponsive element (CRE) or Adl, TGACGTGA, which was a cis-regulatory element, has been identified originally for the mouse CYPllB2 gene (35, 36) and a bovine CYPllB gene (37-39). The element was conserved in rat CYP11B2 and human CYPllBl and -B2 genes (17, 40) but not in rat CYP11B1, -B3, and -B4 genes. The conserved sequence was replaced with TGACATTA in CYPllBl and -B4 and with TGGCATTG in CYP11B3. A promoter element AGGTCA was reported to be shared in the three mouse steroidogenic P450 genes (41, 42), and a novel CRE or Ad4, CCAAGGTC, was defined in the bovine gene (43, 44). As indicated by the underlines, these two sequences overlap with each other. Variants of these elements were found in all of the 5'-flanking regions of rat CYPllB genes; CCAAGGTC at -325 in CYPllB2, and CAAGGTT at -296 and AAGGCTC at -235 in both CYPllBl and -B4. The latter two sequences were conserved in corresponding positions of CYP11B3. In further upstream regions of CYPllBl and -B2, AGGTCA-like sequences were found at -1421- , -1218- , -781, -621, and -480 in CYPllBl and at -1604- , -1420- , -1157- , -1001 in CYPllB2. In CYP11B1, a pair of inverted repeat sequences was present from -1058 to -921 and from -232 to -366, though its functional significance in transcriptional regulation has not been known.

CACCAT g t g a g c a c c t (250) t g t g t t c c a g ATGCTG
Promoter Activity of the C Y P l l B l , -B2, and -B3 Genes-To explore the promoter activities of the 5"flanking regions of the CYPllB genes, these regions were placed immediately upstream of the CAT gene and were introduced into mouse adrenocortical tumor Y1 cells; each 5'-flanking region of CYPllBl (1.6 kb), -B2 (1.5 kb), and -B3 (1.7 kb) was inserted into a promoterless vector, pCAT-Basic, to produce the plasmids, pBl(l.G)-CATB, pB2(1.5)-CATB, andpB3(1.7)-CATB. When their activities were examined, the cells transfected with pB2(1.5)-CATB or pB3(1.7)-CATB produced low levels of CAT (Fig. 5, lanes 9 and 10, and lanes 13 and 1 4 ) , while CAT activity was almost undetectable in the case of pBl(1.6)-CATB (Fig. 5, lanes 5 and 6).
The low promoter activity of these plasmids, especially that of pBl(l.G)-CATB described above, suggested to us the possibility that the 1.5-1.7-kb 5'-flanking regions of CYP11B1, -B2, and -B3 contained suppressive DNA sequence(s). Accordingly, we constructed another set of CAT reporter plasmids, pB1(0.5)-CATB, pB2(0.5)-CATB, and pB3(0.5)-CATB, which carried the 0.5-kb 5"flanking regions of the genes. When their CAT activities were tested, those of both pB2(0.5)-CATB and pB3(0.5)-CATB were greater than those of pB2(1.5)-CATB and pB3(1.7)-CATB by approximately 3fold (Fig. 5, lanes 11 and 12 and lanes 15 and 16). Interestingly, pB1(0.5)-CATB showed the highest level of CAT activity among the fusion gene constructs carrying the 0.5-kb 5'flanking regions as seen in Fig. 5 (lanes 7 and 8 ) and Table   11. The result suggested that the upstream region from -493 to -1602 relative to the transcription initiation site in CYPllBl contained a nucleotide sequence(s) suppressing the promoter activity within the 492-nucleotide 5'-flanking re-   1 and 2 ) and pCAT- Control (lanes 3 and 4 ) were used as a negative and a positive control, respectively. The levels of CAT expression were assayed using ["C]chloramphenicol(33). The reaction products in the assays were separated by thin layer chromatography and were detected by autoradiography. Upper tuo spots represent acetylated forms of chloramphenicol.
gion. I t should also be noted that the promoter activity of the 0.5-kb region of CYP11B3 was only one-tenth of that of C Y P l l B l despite the homology of 93% between their nucleotide sequences in the 0.5-kb region. The results are summarized in Table 11.
Steroid synthesis in Y1 cells has been known to be stimulated by ACTH, which exerts its effects through CAMP-  The CAT activity for pCAT-Control in the absence of BWAMP is taken as 100%. Transfection experiments were repeated at least 4 times to assure the reproducibility. The values were taken from one of the experiments with typical results.
ND, not determined. mediated processes (30). As seen in Fig. 5, addition of dibutyryl CAMP to the culture medium stimulated the promoter activities of all the fusion gene constructs by approximately 3-fold. In the adrenal cortex, on the other hand, aldosterone synthesis is regulated by angiotensin I1 and ACTH (2,3,45). The effects of ACTH, angiotensin 11, and an angiotensin I1 antagonist, [Sarl-Ala"]angiotensin 11, on the CAT expression were, therefore, examined. Addition of ACTH induced an increase in the expression reproducibly by 2-2.5-fold for all the genes except for the case of pBl(l.G)-CATB. Both angiotensin I1 and its antagonist did not affect the levels of CAT activities of any fusion gene under our experimental conditions (data not shown). Next, we produced another set of CAT plasmids where the six fragments containing the 5"flanking regions used in the experiments shown in Fig. 5 were inserted into a promoterless but SV40 enhancer-carrying plasmid, pCAT-Enhancer. These plasmid constructs were introduced into Y1 cells, and the effects of SV40 enhancer on these promoter activities were examined. As shown in Table 11, the SV40 enhancer conferred a strong promoter activity on the 1.6-kb 5"flanking region of CYP11B1, suggesting that the enhancer overrode the action of the suppressing DNA sequence(s) in the upstream region from -1.6 to -0.5 kb. The promoter activity of the 0.5-kb 5'flanking region of CYPllBl was stimulated 10-fold by the enhancer, comparable to that of the 1.6-kb region in the presence of the enhancer. Promoter activities of the 1.5-and 0.5-kb regions of CYPllBZ increased by 10-fold upon introduction of the enhancer, while the ratios of induction by dibutyryl cAMP were not affected. The enhancer stimulated the promoter activities of the 1.7-and 0.5-kb regions of CYP11B3 by 20-fold. It should be noted, however, that the enhanced promoter activities of the 0.5-kb regions of CYPllB2 and -B3 were only one-third and one-fourth of that of CYP11B1, respectively. The CAT activities of pBZ(1.5)-CATE and pB3(1.7)-CATE in the absence of dibutyryl cAMP were lower than that of the promoterless parent pCAT-Enhancer plasmid. This suggested that the -1.5 to -0.5-kb region of CYPllBZ and the -1.7 to -0.5-kb region of CYP11B3 repressed the nonspecific transcription brought about by SV40 enhancer.

DISCUSSION
The four CYPllB genes described in this paper had a similar structural organization except for the deleted region in CYP11B4, suggesting that they have arisen from a common ancestor by gene duplication. In accordance with this interpretation, an unequal crossing over observed between human CYPllBl and -B2 (46) exemplifies gene duplication to generate multiple genes. Among rat CYPllB genes, the overall nucleotide sequences of CYP11B1, -B3, and -B4 have higher homologies than those between one of the three and CYP11B2. Thus, CYP11B1, -B3, and -B4 form a subgroup in rat CYPllB genes. The structures of the genes in this subgroup are unique among known CYPllB genes of mouse, rat, bovine, and human. The nucleotide sequences of both coding and 5"flanking regions of rat CYP11B1, -B3, and -B4 showed lower homologies to those of mouse CYPllB2 (14, 35), bovine CYPllB (38,39), and human CYPllBl and -B2 (15,40) than the homologies of rat CYPllB2 to the genes of the other species. In particular, the structure of human CYP11B1, whose gene product corresponds functionally to that of rat CYPllBl (18,19,40), resembles that of rat CYP11B2. Although no information is available for the structure of the 5"flanking region of mouse CYPllBl at present, the corresponding genes in rat and mouse may be structurally and functionally homologous to each other (14).
CYPllBl and -B2 encode Ilp-hydroxylase P450 and aldosterone synthase P450, respectively. Isolation of the genomic clones indicated the presence of two additional genes, CYPllB3 and -B4. Of these, CYP11B3 had the same number of exons and the same location of introns as those of CYPllBl and -B2. The exonic sequences of CYP11B3 contained no inphase termination codon (Fig. 3). Also found was the replacement of the consensus dinucleotide GT with GC found at the 5'-end of the third intron of CYP11B3 (Table I). Similar replacements have been reported for rat pyruvate kinase M (47), mouse adenine phosphoryltransferase (48), and chicken a-globin (49) genes. The GC dinucleotide could be used as a splice site. As already mentioned earlier, however, the existence of the third homologue of either aldosterone synthase or llp-hydroxylase P450 has not been reported. In the transient transfection assays, the CYP11B3 0.5-kb 5"flanking region exhibited a far lower promoter activity than that of the corresponding region of CYP11B1. This could be attributed to the nucleotide mutations from the B1 gene, although the difference between the sequences was small. The mutations were at only 34 positions out of 492 nucleotides. Considering that the Y1 cells resemble the zonae fasciculata-reticularis cells, which express CYP11B1, the transfection experiments using the Y1 cells possibly reflect gene regulation in the zonae fasciculata-reticularis cells. The weak promoter activity of CYP11B3, therefore, suggests that the gene is expressed in a very small amount, if transcribed, at least in adult rat adrenals. On the other hand, CYP11B4 is considered to be a pseudogene because a sequence corresponding to a region containing the exon 3 and a part of exon 4 was missing. However, a possibility remains that this gene is transcribed in vivo because the 492-nucleotide sequence of the 5"flanking region of the CYP11B4 is identical to that of CYP11B1.
Transient transfection assays demonstrated that the CYPllBl 5"flanking region exhibited a promoter activity in Y l cells. This promoter activity was suppressed by an upstream DNA sequence(s) someplace in the region spanning from -1602 to -493 (Fig. 5). The suppression was abolished by a potent enhancer such as SV40 enhancer (Table 11). Further examinations are required to demonstrate whether the suppressive element(s) functions or does not function in gene regulation in uiuo. Steroid synthesis in Y1 cells is stimulated by ACTH and also by membrane-permeable cAMP derivatives (30). This characteristic resembles that of the zonae fasciculata-reticularis cells of adrenal cortex, in which CYPllBl is expressed to synthesize glucocorticoids. In the 5"flanking region, a CRE-like sequence, TGACATEA, was present with nucleotide changes at two positions from the conserved sequence, TGACGTGA (36). The nucleotide changes did not abolish the promoter activity and the inducibility by CAMP, suggesting that this CRE-like sequence was active. Alternatively, the novel CRE-like sequences, CAAGGTT present at -295 or AAGGCT at -235, could function in the inducibility by dibutyryl CAMP. Thus, although the 5"flanking region of rat CYPllBl had lower homologies to those of rat CYP11B2 and CYPllB genes of other animal species, the promoter activity and its inducibility by dibutyryl cAMP in Y1 cells reflected the expression and regulation of this gene in the zonae fasciculata-reticularis cells.
Aldosterone synthesis is regulated largely by the reninangiotensin system and, to a minor extent, by ACTH. The promoter activity of the 5"flanking regions of CYPllBZ in Y1 cells was affected by neither angiotensin I1 nor an angiotensin I1 antagonist, whereas dibutyryl cAMP stimulated the activity. The induction by the cAMP derivative is consistent with the reported regulatory effect of ACTH (2,3,45). The result that the promoter activity of the 0.5-kb 5"flanking region of CYP11B2 was one-fourth that of CYPllBl sug-gested that different sets of cis-regulatory element(s) and trans-acting protein(s) were involved in transcriptional regulation. The insensitivity of the promoter activity to angiotensin I1 and its antagonist can be ascribed to at least three possibilities. First, Y1 cells do not respond to angiotensin I1 or lack a signal transduction pathway(s) leading to transcription of CYP11B2. Second, the 1.5-kb 5'-flanking region cloned into the promoterless CAT plasmid contains no "angiotensin 11-responsive element(s)." Third, transcription efficiency of CYPllBZ is not regulated by angiotensin I1 in uiuo, i.e. CYPllBZ is transcribed constitutively in the zona glomerulosa cells and is not regulated at the transcription reaction as discussed below.
Stimulation of aldosterone synthesis by angiotensin I1 in the zona glomerulosa cells is known to be observed in short term (minutes) and in long term (hours to days) regulation (2,3). The short term regulation stimulates conversion of cholesterol to pregnenolone, whereas it is not known whether the conversion of deoxycorticosterone to aldosterone is affected. The observation that the long term regulation by angiotensin I1 stimulates conversion of deoxycorticosterone to aldosterone can be attributed to enhancement of P450.1d0 biosynthesis in the zona glomerulosa cells and/or proliferation of the cells expressing the enzyme. Although the levels of transcript and protein product from CYPllB2 in adrenal glands increased by angiotensin I1 during 1 or 2 weeks (4-6, 11, 50, 51), it has not been demonstrated that angiotensin I1 enhanced transcription efficiency of the CYP11BZ gene within the zona glomerulosa cells. Recently, we demonstrated that the long term effect of angiotensin I1 was exerted, at least in part, through proliferation of cells expressing the CYPllB2 gene product (13). One of the most important, unanswered questions in corticosteroid synthesis is the mechanism(s) underlying the long term regulation of aldosterone synthesis in the zona glomerulosa cells. To answer the question, mechanisms for zone-specific gene regulation of CYPllBl and -B2 as well as mechanisms for the proliferation of adrenocortical cells in each zone should be elucidated.