Transcriptional synergism between vitamin D-responsive elements in the rat 25-hydroxyvitamin D3 24-hydroxylase (CYP24) promoter.

Transcription of the CYP24 gene is induced by 1,25-(OH)2D3 through a vitamin D receptor-dependent process. The functional activities of three possible vitamin D response elements (VDREs), located on the antisense strand of the rat CYP24 promoter, were investigated by transient expression of native and mutant promoter constructs in COS-1, JTC-12, and ROS 17/2.8 cells. A putative VDRE with a half-site spacing of 6 base pairs at −249/−232 (VDRE-3) did not contribute to 1,25-(OH)2D3 induced expression in the native promoter, although activity has been reported when the element was fused to the heterologous thymidine kinase promoter. Two VDREs with half-site spacings of 3 base pairs at −150/−136 and −258/−244 (VDRE-1 and VDRE-2, respectively), showed transcriptional synergism in COS-1 cells when treated with 1,25-(OH)2D3 (10−7 to 10−11 M). The contribution of both VDREs was hormone-concentration dependent from 10−10 to 10−12 M, with VDRE-1 demonstrating greatest sensitivity to 1,25-(OH)2D3. Transactivation by VDRE-1 was always greater than VDRE-2, but the converse was observed for the binding of vitamin D receptor-retinoid X receptor complex by each VDRE in gel mobility shift assays. The synergy observed between VDRE-1 and VDRE-2 may have important implications in cellular responses to different circulating levels of 1,25-(OH)2D3.

as either a homodimer or a heterodimeric complex to a DNA sequence known as the vitamin D-responsive element (VDRE) present in the promoter of target genes (11)(12)(13). Heterodimers consisting of VDR and retinoid X receptor (RXR) are widely documented (12), although VDR heterodimeric complexes have also been demonstrated for both the retinoic acid receptor (14,15) and the thyroid hormone receptor (16). Vitamin D-responsive elements generally display a binding motif that consists of two imperfect, direct repeat hexameric sequences (i.e. halfsites) that are separated by 3 bp or, more rarely, by 6 bp; these VDREs are referred to as DR-3 and DR-6, respectively (12).
Molecular regulatory studies of CYP24 gene expression by 1,25-(OH) 2 D 3 are in progress and promoter analysis data for rat (14, 30 -33) and human (34) have been reported. In the rat CYP24 gene promoter, three VDREs on the antisense strand have been identified. We have previously defined the proximal VDRE (30) in its native promoter context, while the two more distal VDREs have been tested by fusing to a heterologous promoter (14,31). To date, however, the functionality of the VDREs has not been verified in the context of the native CYP24 promoter, and there is no direct information available regarding the contribution of each VDRE to vitamin D induction or whether there is a cooperative interaction between the response elements. These issues are addressed in the current investigation, in which mutagenic constructs of the rat CYP24 promoter have been used in transient gene expression and gel mobility shift analysis. mutagenesis. The fragment was cloned into the HincII site of pBluescript KS ϩ in the antisense orientation to generate pBKS-WT, and single-stranded DNA purified as described by Kunkel et al. (35). Briefly, Escherichia coli CJ236 (dut ϩ , ung Ϫ ) was transformed with pBKS-WT and superinfected with helper phage M13K07. Single-stranded DNA was purified by NaCl/PEG precipitation and used as template in the mutagenesis reactions. The primers employed (P1, P2, P4, and P5) contain mutations (shown as bold letters) in VDRE-1, VDRE-2, VDRE-2, and VDRE-3, and VDRE-3, respectively: P1, 5Ј-CGC CCT CAC TCA AGC TTC TGA CTC CAT CCT CTT CCC ACA CCC-3Ј; P2, 5Ј-GTG CTC GCA GCG CAT GCG CTG AAC CCT GGG CTC GAC CCG CCT-3Ј; P4, 5Ј-GTG CTC GCA GCG CAT GCG CAC TAC CCT GGG CTC GAC CCG CCT-3Ј; P5, 5Ј-GTG CTC GCA GCG CAC CCG CTG AAC CCT GGG CTA TAC CCG CCT-3Ј. The positions of these mutations are shown in Fig. 1. After annealing each mutant oligonucleotide to the template, the second DNA strand was completed using T4 DNA polymerase. Positive clones were isolated by colony screen hybridization using tetramethylammonium chloride washes (36). Mutant clones were verified by sequence and restriction enzyme analysis. In VDRE-1, a HindIII site was introduced to produce the mutated nucleotides.

Materials-Hoffmann
Generation of Promoter Constructs-The CYP24 mutant promoter fragments were excised from pBKS-WT by digestion with XhoI/EcoRV and cloned into XhoI/SmaI-digested pGL3-Basic containing the firefly luciferase reporter gene (Promega). Plasmid DNA was prepared by alkaline lysis and CsCl/ethidium bromide equilibrium density gradients (37). All plasmid DNA was quantified by spectrophometry, and supercoil form was estimated by 1% agarose gel analysis to ensure experimental consistency.
Maintenance and Transfection of Tissue Culture Cells-COS-1 cells and monkey kidney proximal tubular JTC-12 cells (kindly supplied by Dr T. Matsumoto, University of Tokyo School of Medicine, Japan; Ref. 38) were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum (FCS). Rat osteosarcomaderived ROS 17/2.8 cells were maintained in DMEM/Ham's F-12 (1:1) supplemented with 10% FCS. In preparation for electroporation, cells were grown to 80 -90% confluence, removed by trypsinization, washed once in phosphate-buffered saline and resuspended at (6 ϫ 10 6 cells/ml) in 20 mM Hepes (pH 7.05) containing 137 mM NaCl, 5 mM KCl, 0.7 mM Na 2 HPO 4 , 6 mM dextrose (39), and 500 g/ml sheared salmon sperm DNA. COS-1 and JTC-12 cells (3 ϫ 10 6 ) and 1 pmol of construct DNA were electroporated at 280 V and 960 microfarads, while ROS 17/2.8 cells were electroporated at 200 V and 960 microfarads using a Bio-Rad Gene Pulser. COS-1 cells were also co-transfected with VDR expression clone pRSV-hVDR, generated by cloning the human VDR cDNA sequence (kindly supplied by Nigel Morrison, Garvan Institute for Medical Research Sydney, Australia) downstream of the Rous sarcoma virus promoter. Following electroporation, the samples were placed on ice for 10 min and divided (1.5 ϫ 10 6 cells) into two wells of a six-well plate containing DMEM, 10% FCS (for COS-1 and JTC-12 cells) or DMEM/ Ham's-F-12, 10% FCS (for ROS 17/2.8 cells). Cells were allowed to recover for 20 h and then the media replaced with RPMI medium (without phenol red) supplemented with 12% charcoal-stripped FCS. Ethanol carrier or 1,25-(OH) 2 D 3 was added at the indicated concentration and the cells incubated for 24 h prior to harvesting.
Luciferase Assay-Cells were washed once with phosphate-buffered saline and treated with 200 l of Cell Culture Lysis Reagent (Promega) for 10 min at room temperature. The cells were harvested with a rubber policeman, frozen on dry ice, thawed on ice, and vortexed vigorously. After centrifugation, the supernatant was removed and assayed for protein concentration. Luciferase activity was determined using a luciferase assay system (Promega) in 25 g of lysate protein and measurements made with a Berthold model LB 9502 Luminometer.
Gel Mobility Shift Assays-Double-stranded oligonucleotide probes were synthesized that encompassed the VDRE-1, VDRE-2, or VDRE-3 sites in the rat CYP24 promoter. An oligonucleotide that encompassed a known VDRE located at Ϫ758/Ϫ740 in the mouse osteopontin gene promoter (40) was employed as control (mSpp1-VDRE). Each doublestranded oligonucleotide was designed with SalI and XhoI overhangs at the 5Ј or 3Ј ends, as shown below.
Each double-stranded oligonucleotide was labeled by end-filling with [␣-32 P]dCTP using Klenow enzyme and purified by 12% polyacrylamide gel electrophoresis. Nuclear extracts from 1,25-(OH) 2 D 3 -treated COS-1 cells were prepared with or without co-transfection of pRSV-hVDR expression vector, as described previously (30). Binding reactions for each assay contained 5 g of nuclear protein, 1 g of poly(dI-dC) to a final volume of 12 l in binding buffer (25 mM Tris-HCl, pH 7.6, 100 mM KCl, 0.5 mM dithiothreitol, 5 mM MgCl 2 , 0.5 mM EDTA, and 10% glycerol) and were incubated on ice for 10 min. Radiolabeled probe (200,000 cpm) was added and samples incubated on ice for another 20 min. For gel shift inhibition assays, a VDR monoclonal antibody designated IgG2b (Affinity BioReagents Inc., Neshanic Station, NJ) was employed. For supershift assays, an RXR monoclonal antibody (4RX-1D12, Ref. 41), kindly provided by Dr Pierre Chambon (Strasbourg Cedex, France), was employed. These antibodies were included in the binding reactions and incubated on ice for 10 min prior to addition of probe. Gel shift competition assays were performed with unlabeled competitor oligonucleotide at molar excess concentrations in the binding reaction. Retarded DNA nuclear protein complexes were resolved on a 4% nondenaturing polyacrylamide gel in a low ionic strength running buffer (0.5 ϫ TBE) at 4°C. The gel was dried and exposed to Kodak X-Omat AR film with an intensifying screen at Ϫ70°C. In the absence of added 1,25-(OH) 2 D 3 , basal expression of each of the mutant constructs in the different cell lines was the same as the wild-type construct, demonstrating that these VDREs do not contribute to basal expression (results not shown). Transient transfection experiments were performed initially in COS-1 cells that were co-transfected with CYP24 promoter/ luciferase constructs and VDR expression vector (pRSV-hVDR) to compensate for the deficiency of VDR in the COS-1 cell line. In response to 1,25-(OH) 2 D 3 treatment, the wild-type construct (pGL3-WT) gave a 17.8-fold level of induction in these cells (Fig. 2). When VDRE-1 alone was mutated (pGL3-M1), leaving the distal VDRE-2 and VDRE-3 intact, the level of 1,25-(OH) 2 D 3 induction was reduced to 2.8-fold, which demonstrated the substantial contribution of VDRE-1 to promoter activity. A 6.6-fold level of induction was observed with the pGL3-M2 construct in which VDRE-2 was mutated, and essentially the same level was observed when both VDRE-2 and VDRE-3 were mutated in pGL3-M3 (Fig. 2). These findings established that both VDRE-1 and VDRE-2 are functional and suggested that VDRE-3 does not significantly contribute to induction even when the overlapping VDRE-2 is inactivated. To confirm this latter finding, the expression of pGL3-M5, in which the proximal half-site in VDRE-3 was altered, was investigated. The hormone induction observed with this construct was similar to the wild type. The possible role of VDRE-3 was examined further by mutating both VDRE-1 and VDRE-2 and leaving VDRE-3 intact. This construct (pGL3-M6) was inactive (Fig. 2). There was no response to 1,25-(OH) 2 D 3 by the construct pGL3-M4, in which all three VDREs were altered. Thus, when evaluated in all three cell types, VDRE-1 and VDRE-2 were responsible for the 1,25-(OH) 2 D 3 -mediated induction of the wild-type promoter region. The induction seen for the wild-type construct (17.8-fold) was greater than the sum of the individual contributions of VDRE-1 (6.6-fold) and VDRE-2 (2.8-fold), thus demonstrating transcriptional synergism.

Transient Expression of Promoter Constructs in COS
Cell specificity for the synergism between VDRE-1 and VDRE-2 was also evaluated in JTC-12 and ROS 17/2.8 cells.
Both cell lines express endogenous VDR and respond to 1,25-(OH) 2 D 3 -mediated gene induction (38,42). Expression of the wild-type and mutant CYP24 promoter constructs in JTC-12 and ROS 17/2.8 cells produced transactivation results that were similar to those obtained in COS-1 cells (Fig. 2). A nearly 2-fold synergistic action between VDRE-1 and VDRE-2 was demonstrated in all three cell lines, in which a greater contribution was observed for VDRE-1 compared with VDRE-2 (Fig. 2).
VDREs Bind Nuclear Protein Complexes-Binding of nuclear proteins to the VDREs was investigated by gel mobility shift analysis using extracts from COS-1 cells transfected with pRSV-hVDR expression vector and treated with 10 Ϫ7 M 1,25-(OH) 2 D 3 (30). An oligonucleotide encompassing the mouse osteopontin VDRE (i.e. mSpp1-VDRE) was employed as a control probe and contained a functionally active VDRE known to bind strongly the VDR-RXR complex (40). A major protein complex of the same mobility as the mSpp1-VDRE was detected with VDRE-1 and VDRE-2, but the bands were not evident when probes were incubated with nuclear extract prepared from COS-1 cells not transfected with pRSV-hVDR (Fig. 3A). This finding indicated that VDR is present in the protein complex that binds to VDRE-1 and VDRE-2. Competition experiments using a 10-fold molar excess of mSpp1-VDRE completely prevented formation of the protein complex observed with either VDRE-1 or VDRE-2 (results not shown). To further characterize the protein complex, we employed both a monoclonal antibody to VDR that interfered with DNA binding of VDR and a supershifting RXR monoclonal antibody. The VDR monoclonal antibody prevented formation of the major protein complex detected when radiolabeled VDRE-1, VDRE-2, or mSpp1-VDRE were used as probes (Fig. 3B). The RXR monoclonal antibody supershifted the VDR-containing complex obtained with each of these probes (Fig. 3C). It can be concluded from these data that the major protein complex that binds to VDRE-1 and VDRE-2 contains both VDR and RXR. Other gel mobility shift experiments with mutant oligonucleotides established that the mutations introduced into VDRE-1 and VDRE-2 inhibited completely the binding of the VDR⅐RXR complex (data not shown).
Based upon the intensity of the protein-VDRE complexes, it appeared that VDRE-2 had a greater binding affinity than VDRE-1 while the mSpp1-VDRE had a higher binding affinity than either VDRE-1 or VDRE-2 (Fig. 3). To investigate the binding affinity of VDRE-1 and VDRE-2 more directly, competition of the radiolabeled mSpp1-VDRE probe was undertaken with either unlabeled VDRE-1 or VDRE-2 at 10-, 25-, 50-, 100-, and 200-fold molar excess. The results showed that the protein complex was efficiently competed by a 10-fold excess of the competitor VDRE-2, but a 50-fold excess of VDRE-1 was required for significant competition (Fig. 4A). These results confirmed that the complex containing VDR and RXR has a stronger binding affinity for VDRE-2 than VDRE-1. Competition experiments were also carried out using radiolabeled VDRE-1 or VDRE-2 as probes and each of these oligonucleotides as competitors. With self-competition, binding of the VDR⅐RXR protein complex to VDRE-1 and VDRE-2 was totally abolished with the corresponding oligonucleotides, confirming the specificity of protein binding (Fig. 4B). With VDRE-1 as probe, a 10-fold molar excess of VDRE-2 effectively prevented protein complex formation, while in the reverse experiment, a 50-fold molar excess of VDRE-1 was required. These experiments further demonstrated the greater binding affinity of VDRE-2.
Using COS-1 nuclear extracts, we have detected a very faint retarded protein complex with the VDRE-3 probe in gel shift assays and antibody experiments (data not shown), which demonstrated that this complex contained VDR and RXR. The biological significance of VDR⅐RXR binding to VDRE-3 is not clear in view of the inactivity of the sequence in the CYP24 promoter. In other studies, DR-6 type VDREs, and in particular VDRE-3, have been reported to preferentially bind VDR homodimers and retinoic acid receptor-VDR heterodimers rather than VDR⅐RXR (12,14).  (Fig. 5A). However, when cells were treated with 10 Ϫ11 M 1,25-(OH) 2 D 3 , the induction was reduced to 8.3-fold and was further lowered to 3.7-fold with 1,25-(OH) 2 D 3 at 10 Ϫ12 M (Fig. 5A). Mutant constructs of VDRE-1 (pGL3-M1) and VDRE-2 (pGL3-M2) were active at 10 Ϫ11 M 1,25-(OH) 2 D 3 , giving 2.1-fold and 3.3-fold levels of induction, and they retained their transcriptional synergism. At the lowest level of 1,25-(OH) 2 D 3 tested (10 Ϫ12 M), only VDRE-1 was found to be responsive (Fig. 5A). Separate experiments confirmed that VDRE-3 did not contribute to the induction response (when VDRE-1 and VDRE-2 were functional) under the decreasing concentrations of 1,25-(OH) 2 D 3 employed (data not shown).

Studies in COS-1 Cells Treated with Different Amounts of 1,25-(OH) 2 D 3 or VDR-Having
The possible impact of cellular VDR concentration on the synergistic interaction between VDRE-1 and VDRE-2 was evaluated in COS-1 cells receiving one-tenth the amount of the transfected pRSV-hVDR. Under these conditions, the wild-type promoter construct gave an 8.0-fold level of induction compared with 16.4-fold when cells were co-transfected with the full amount of pRSV-hVDR (Fig. 5B). Studies with the mutant constructs pGL3-M1 and pGL3-M2 also showed a decreased level of induction, but the cooperative interaction between VDRE-1 and VDRE-2 was retained. Upon comparing results at different 1,25-(OH) 2 D 3 and VDR levels, it was noted that VDRE-1 makes a greater contribution to the hormone's cellular induction than VDRE-2. Such results are diametric to the VDR⅐RXR binding affinities for the two VDREs, as measured by gel mobility shift analysis.

DISCUSSION
Transcription of the rat cytochrome P450 24 (CYP24) gene is induced by 1,25-(OH) 2 D 3 (19 -23). Three VDREs in the first 298 bp of the 5Ј-flanking sequence have been reported to participate in regulation of the rat CYP24 gene (14, 30 -33). To date, only the proximal VDRE (VDRE-1) has been evaluated for function in the context of its native promoter (30). The other putative VDREs (VDRE-2 and VDRE-3) were characterized when linked to the heterologous thymidine kinase promoter (14,31), which does not reflect native promoter architecture and function. Using constructs containing the first 298 bp of the CYP24 gene promoter, we have now characterized VDRE-1 and VDRE-2 as functional hormone response elements. Upregulation of the CYP24 promoter regulation in three different cell lines is consistent with the operation of a general modulatory loop that functions in 1,25-(OH) 2 D 3 -responsive cells expressing the CYP24 gene. A very similar regulatory pathway also appears to be present in the human CYP24 promoter, which has been verified to contain two VDREs of the DR-3 type at about the same position as the VDRE-1 and VDRE-2 in the rat (34).
VDRE-3 with a 6-bp spacing was not responsive to 1,25-(OH) 2 D 3 when tested in three cell lines following selected mutation of VDRE half-sites. The data from these cellular promoter expression studies demonstrate convincingly that VDRE-3 is not functional. Our findings with VDRE-3 are in contrast to those of Kahlen and Carlberg (14), who reported that this sequence responds to 1,25-(OH) 2 D 3 (4 -6-fold) when fused to the thymidine kinase promoter and expressed transiently in ROS 17/2.8 cells. We conclude that the environment of the VDRE-3 in the rat CYP24 promoter is not conducive to 1,25-(OH) 2 D 3 responsiveness, at least in the cell types examined, and our study emphasizes the importance of investigating the functional role of a VDRE in the context of its native promoter architecture. Nevertheless, it is possible that VDRE-3 may be active in an appropriate cellular environment, since the sequence has been shown in gel mobility shift assays to bind VDR homodimers, heterodimers of VDR, and retinoic acid receptors (14) and, in the present work, to bind weakly a complex of VDR⅐RXR (data not shown). In the human CYP24 promoter (34), the sequence 5Ј-ATGCGAacgccgGAGTTC-3Ј corresponds to the rat VDRE-3, but its function has not been investigated. Other DR-6 type VDREs have been identified in FIG. 3. Gel shift analysis using oligomers to VDRE-1 and  VDRE-2. A, double-stranded oligomers to VDRE-1, VDRE-2, and mouse osteopontin VDRE (mSpp1-VDRE) were labeled by end-filling with [␣-32 P]dCTP and incubated with nuclear extracts of COS-1 cells either mock-transfected (Ϫ) or transfected with 5 g of pRSV-hVDR (ϩ). The major retarded complex (arrowed) is unique to pRSV-hVDRtransfected cells. B, for neutralization assays, nuclear extracts (from COS-1 cells transfected with pRSV-hVDR) were incubated with (ϩ) or without (Ϫ) VDR monoclonal antibody (VDR mAb) prior to addition of labeled probes for VDRE-1, VDRE-2, and mSpp1-VDRE. C, for supershift assays, as in B but using RXR monoclonal antibody (RXR mAb). The upper arrow (*) indicates the supershifted complex.
the promoters for osteocalcin and fibronectin (12). The human osteocalcin DR-6 motif confers 1,25-(OH) 2 D 3 inducibility when fused to the thymidine kinase promoter (13). However, as far as we are aware, it has not been established whether the osteocalcin DR-6 or fibronectin DR-6 motifs contribute to 1,25-(OH) 2 D 3 responsiveness in their natural promoters.
Gel mobility shift analysis was used to investigate the protein complexes in COS-1 cell nuclear extracts that bind to VDRE-1 and VDRE-2. A major retarded band was identified with each VDRE, and antibody studies established that this complex contained both VDR and RXR. This result agrees with other studies, where different DR-3 type VDRE sequences have been shown to bind this heterodimer (12). Gel shift competition data indicated that the VDR⅐RXR protein complex bound more tightly to VDRE-2 than VDRE-1, a somewhat unexpected result, since the contribution of VDRE-1 to the 1,25-(OH) 2 D 3 inductive response was always greater than that of VDRE-2. Liu and Freedman (43) have demonstrated substantial tran- scriptional synergism between VDR and various classes of nonreceptor transcription factors when the DNA binding sites for these proteins are positioned closely in a reporter plasmid. Such a situation may explain the greater transactivation of VDRE-1, in which its affinity for VDR⅐RXR (compared with VDRE-2) is enhanced through a cooperative interaction with nearby, and as yet unidentified, promoter-bound transcription factor(s). It is also possible that the low 1,25-(OH) 2 D 3 transactivation contributed by VDRE-2 arises from the greater distance of this VDRE from the transcriptional machinery. This seems unlikely, however, since deleting the promoter region between the two VDREs does not alter transactivation by 1,25-(OH) 2 D 3 (31).
A significant finding in the present work is the transcriptional synergistic response between VDRE-1 and VDRE-2, which resulted in an 18-fold level of induction in COS-1 cells. Transient studies of CYP24 promoter constructs in COS-1 cells under non-limiting conditions for VDR or 1,25-(OH) 2 D 3 showed that the transactivation was about twice the sum of the hormonedependent transactivations for VDRE-1 and VDRE-2 when evaluated separately. A similar level of synergism between these VDREs was also seen in JTC-12 and ROS 17/2.8 cells in which VDR is expressed endogenously. The mutations introduced into VDRE-1 or VDRE-2 inhibited completely the response of each site to 1,25-(OH) 2 D 3 induction. In previous work (30), mutated VDRE-1 prevented 1,25-(OH) 2 D 3 induction from a CYP24 promoter that contained this VDRE but not VDRE-2. Combined mutagenesis of VDRE-1 and VDRE-2, in the current work, prevented induction and demonstrated the effectiveness of both mutations. Hence, the coordinated interaction between the two VDREs appears to be the sole basis for the observed synergism.
Synergistic induction of the rat CYP24 promoter constitutes one of the highest stimulations observed so far for any 1,25-(OH) 2 D 3 -responsive promoter and would appear to be the only regulatory region in which two VDREs have been shown to be functional in their native promoter. The synergism observed between the two VDREs may be at the level of enhanced cooperative DNA binding of the VDR⅐RXR protein complexes. Liu and Freedman (43) have demonstrated cooperative binding of dimeric VDR complexes to two VDRE sites separated by 50 bp in an artificial promoter. However, Zierold et al. (31) were unable to show cooperative binding between protein complexes bound to sequences encompassing VDRE-1 and VDRE-2 in gel shift experiments. Another possible mechanism could involve independent binding of the protein complexes to each VDRE, with synergism resulting from an interaction of these complexes with the basal transcription machinery (43,44). Of possible relevance to this latter mechanism is the demonstration that VDR can interact with the general transcription factor TFIIB (45).
Maximal induction of the CYP24 promoter in COS-1 cells was observed over a 1000-fold range of 1,25-(OH) 2 D 3 concentration (i.e. 10 Ϫ7 M to 10 Ϫ10 M). Therefore, it would appear that the VDREs can function maximally in a synergistic manner at physiological concentrations of 1,25-(OH) 2 D 3 . Continued decline in hormone concentration to 10 Ϫ11 M resulted in VDRE-1 being the dominant response element, although synergism was retained between the two VDREs. The same effect was observed with similar 1,25-(OH) 2 D 3 concentrations in ROS 17/2.8 cells, but a 10-fold higher concentration of hormone was required in JTC-12 cells (results not shown). The reason for the reduced sensitivity in these cells is under investigation. Nevertheless, the collective results establish synergism occurrence over a physiological range of 1,25-(OH) 2 D 3 concentration in which VDRE-1 is always the major contributor to induction, particularly at low hormone concentrations.
The transcriptional synergism between the two VDREs in the promoter of CYP24 may have important biological consequences. The induction of CYP24 by 1,25-(OH) 2 D 3 constitutes an interesting feedback mechanism whereby 1,25-(OH) 2 D 3 acts to regulate its metabolic clearance rate and, thereby, influence its ambient concentration. Increased levels of CYP24 activity will result in elevated side-chain oxidation of 1,25-(OH) 2 D 3 and ultimately conversion to the water-soluble calcitroic acid and subsequent excretion (17,18). In the normal situation, CYP24 is expressed predominantly in the kidney but can be induced by 1,25-(OH) 2 D 3 in this tissue, in the intestine (19 -23), and also in a variety of other cell types (24 -29). It seems probable, therefore, that CYP24 expression in different tissues not only protects the tissue from the effects of excessive 1,25-(OH) 2 D 3 but also regulates serum hormone levels by increasing its metabolic clearance rate. A synergistic response would ensure rapid removal of hormone when levels are sufficiently high to cause hypercalcemia and accelerated bone resorption (46). It can be predicted that VDRE-1 is preferentially utilized at lower 1,25-(OH) 2 D 3 levels, but at higher hormone levels both VDREs are activated and transcriptional synergism facilitates the efficient inactivation of 1,25-(OH) 2 D 3 .