TltE OF BIOUXlcaL CHEMISTRY Vol. 269, No. 14, Issue April ...TltE JOliRNAL OF BIOUXlcaL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 269,

TltE JOliRNAL OF BIOUXlcaL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 269, No. 14, Issue of April 10545-10550, 1994 Printed in U.S.A.

Identification of a Vitamin D-responsive Element in the 5”Flanking Region of the Rat 25-Hydroxyvitamin D3 24-Hydroxylase Gene*

(Received for publication, December 3, 1993, and in revised form, January 18, 1994)

Yoshihiko Ohyama$§$ Keiichi Ozonoll, Motoyuki Uchida**, Toshimasa ShinkiSf, Shigeaki KatoOO, Tatsuo SudaSS, Osamu YamamotoS, Mitsuhide Noshiros, and Yukio Katos From the $Graduate Department of Gene Science, Faculty of Science, Hiroshima University, 1-3-1 Kagamiyama, Higashi-Hiroshima 724, the IlOsaka Medical Center for Maternal and Child Health, 840 Murodo-cho, Zzumi, Osaka 590-02, **Bio-Medical Research Laboratories, Kureha Chemical Industry Co. Ltd., 3-26-2 Hyakunin-cho, Shinjuku-ku, Tokyo 169, the *$Department of Biochemistry, School of Dentistry, Showa University, 1-5-8 Hatanodai, Shinagawa-ku, Tokyo 142, the @Department of Agricultural Chemistry, Faculty of Agriculture, Tokyo University of Agriculture, 1-1-1 Sakuragaoka, Setagaya-ku, Tokyo 156, and the $Department of Biochemistry, Hiroshima University School of Dentistry, 1-2-3 Kasumi, Minami-ku, Hiroshima 734, Japan

The 5”flanking region of the rat vitamin D, 24-hydroxylase (P45Occ24) gene was examined and a vitamin D-responsive element (VDRE) responsible for the lq25- dihydroxyvitamin D, (1,25-(OH),D,) enhancement was identified. Unidirectional deletion analyses of the 5‘- flanking region indicated that the region [-167/-1021 is involved in vitamin D responsiveness. Further functional analyses showed that the segment [-204-1291 conferred the hormone responsiveness in an orientation-independent manner when it was placed upstream to the heterologous thymidine kinase promoter or the rabbit j3-globin promoter. The segment [-2W-1291 contained two direct repeat motifs homologous to other VDREs found in the osteocalcin and osteopontin genes. Synthetic oligonucleotides containing the putative VDRE were used for functional analyses and gel mobility shift assays. The proximal [-151/-1371, but not the distal [-169/-1551 direct repeat activated the transcription in response to 1,25-(OH),D, through the P-globin promoter. Furthermore, the proximal direct repeat formed a complex with the vitamin D receptor and a nuclear accessory factor(s) from COS cells (or retinoid X receptor) in the presence of l,BS-(OH),D,. These results indicate that a direct repeat motif, AGGTGAgt- gAGGGCG, located at -151 base pairs upstream in the antisense strand binds to a heterologous dimer consist- ing of the VDR occupied with 1,25-(OH),D, and the nuclear accessory factor and that it plays a critical role in mediating the vitamin D enhancement of the rat P45Occ24 gene expression.

Vitamin D, is first hydroxylated at carbon 25 to yield 25- hydroxyvitamin D, (25-(OH)D,)’ in the liver, then in the kidney

* This study was supported by Grant-in-Aid 04771451 for Scientific Research from the Ministry of Science, Education and Culture of Japan and Naito Foundation Grant 92-110 (to Y. 0.). A preliminary report of portions of this work was presented at 15th Annual Meeting of The American Society for Bone and Mineral Research, Tampa, FL, Septem- ber 18-22,1993. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertisernent” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequencefsl reported in this paper has been submitted

to the GenBankTM/EMBL Data Bank with accession numberis) 017792.

Gene Science, Faculty of Science, Hiroshima University, 1-3-1 ll To whom all correspondence should be addressed: Graduate Dept. of

Kagamiyama, Higashi-Hiroshima 724, Japan. ”el.: 81-824-24-7458;

The abbreviations used are: 25-(OH)D,, 25-hydroxyvitamin D,; FS: 81-824-24-0733.

at the a-position of carbon 1 to generate 1,25-(OH),D,, the active form of vitamin D. The resulting metabolite exerts various biological effects, including the maintenance of calcium homeostasis, regulation of bone remodeling, and modulation of cell growth and differentiation (DeLuca, 1988; Haussler et al., 1988; Minghetti and Norman, 1988; Suda et al., 1990; Pike, 1991).

Another important enzyme involving in vitamin D metabo- lism is 25-hydroxyvitamin D, 24-hydroxylase (P45Occ24). This enzyme is responsible for the conversion of 25-(OH)D, and 1,25- (OH),D, to 24,25-dihydroxyvitamin D, (24,25-(OH),D,) and 1,24,25-trihydroxyvitamin D,, respectively (Ohyama and Okuda, 1991). These metabolites are thought to be inactive forms of vitamin D (Brommage and DeLuca, 1985; Haussler et al., 1988), although some investigators have suggested that massive doses of 24,25-(OH),D, stimulate bone formation with- out inducing hypercalcemia (Nakamura et al., 1992a, 1992b).

The presence of 1,25-(OH),D, induces 24-hydroxylase activity in its target cells (Tanaka and DeLuca, 1974), which may play a crucial role in eliminating the hormonal activity of vitamin D compounds (Brommage and DeLuca, 1985). Examina- tion of the 24-hydroxylase induction mechanism at the molecular level contributes to the understanding of the vitamin D endocrine system. We purified rat 24-hydroxylase (Ohyama et al., 1989; Ohyama and Okuda, 1991), cloned its cDNA (Ohyama et al., 1991), and characterized the gene structure (Ohyama et al., 1993). These studies revealed that 24-hydroxylase mRNA levels were dramatically increased by injecting of vitamin D in vivo (Shinki et al., 1992). The expression level of 24-hydroxylase mRNA differed between the kidney and the intestine (Shinki et al., 1992). However, the molecular mechanism by which 1,25-(OH),D, up-regulates the gene expression of P45Occ24 is not known.

Recent studies have shown that the enhancer termed vitamin D-responsive element (VDRE) mediates the effect of 1,25- (OH),D,. Several VDREs have been identified in the genes of human (Ozono et al., 1990) and rat osteocalcin (Demay et al., 1990; Markose et al., 1990; MacDonald et al., 1991) and mouse osteopontin (Noda et al., 1990). A comparison of these VDREs has suggested that the VDRE consists of a direct repeat structure of two six-nucleotide half-sites with a three-nucleotide spacer. The VDRE, like the thyroid hormone and retinoic acid response elements, belongs to the subgroup of hormone re-

VDRE, vitamin D-responsive element; VDR, vitamin D receptor; RXR, retinoid X receptor; 1,25-(OH),D,, lu,25-dihydroxyvitamin D,; 24,25- (OH),D,, 24,25-dihydroxyvitamin D,; P45Occ24, 25-hydroxyvitamin D, 24-hydroxylase; CAT, chloramphenicol acetyltransferase.

10545

10546 Vitamin D-responsive Element in the Vitamin D 24-Hydroxylase Gene

sponse elements which has direct repeats of a half-site (Ume- sono et al., 1991; Ozono et al., 1991). The VDRE is the binding site for the ligand-saturated vitamin D receptor (VDR) (Liao et al., 1990). The VDR requires a nuclear factor before it can bind to the VDRE (Liao et al., 1990; Sone et al., 1991a, 1991b), and retinoid X receptor (RXR) meets this requirement (Yu et al., 1991).

We describe here the identification and characterization of VDRE in the 5"upstream region of the P45Occ24 gene. This element, which is homologous to other identified VDREs, lies between -151 and -137 nucleotides from the transcription initiation site, binds to VDR and confers vitamin D-dependent transcriptional activity.

EXPERIMENTAL PROCEDURES

All enzymatic manipulations proceeded according to the standard procedures (Sambrook et al., 1989).

Materials-Restriction and modifying enzymes were purchased from Toyobo Inc. (Osaka, Japan) and Takara Shuzo (Kyoto, Japan). [ W ~ ~ P I ~ C T P (111 TBq/mmol), [y-32PlATP (111 TBq/mmol), [14Clchloram- phenicol(1.85 GBq/mmol), and [35S]dCTP (50 TBq/mmol) were obtained from DuPont-NEN.

Measurement of25-(0H)D3 24-Hydroxylase Activity-The mitochon- drial fraction from 25-(OH)D,-treated COS7 cells was solubilized with 0.8% cholate and centrifuged a t 8000 x g for 15 min. The enzyme activity of the supernatant was assayed in the presence of NADPH, adrenodoxin, and NADPH-adrenodoxin reductase as an electron trans- porting system. After a 1-h incubation, products were extracted and analyzed by high performance liquid chromatography as described previously (Ohyama et al., 1989).

Sequencing of the 5'-Flanking Region-The 5"flanking region of P45Occ24 was isolated and subcloned into pUC19 vector designated p304.5 (Ohyama et al., 1993). Deletion mutants (Yanisch-Perron et al., 1985) of the plasmid were prepared by means of a deletion kit (Takara Shuzo., Kyoto, Japan). The mutants were sequenced on alkaline-dena- tured plasmid DNA templates using a Sequenase kit (U. S. Biochemical Corp.).

Construction of Reporter Plasmids-The 5"flanking region of the 24-hydroxylase gene was deleted using the appropriate restriction enzymes. The promoter region was digested at the 3'-end (+188) with SmaI and at the 5'-end with XbaI, S a d , HincII, PstI, and BstEII to create the fragments -2.2 kilobased+l88, -950/+188, -646/+188, -291/ +188, and -167/+188, respectively. For detailed deletion analyses, the fragment -291/+188 was truncated with RsaI at +9 in the 3' terminus. The fragment -291/+9 was further digested with BstEII, BstUI, and DdeI at the 5'-end to create the fragments -167/+9, -102/+9, and -70/ +9, respectively. These fragments were blunt-ended by the Klenow fragment and ligated to the blunt-ended XbaI site of pCAT-basic vector (Promega). To study the heterologous herpes simplex virus thymidine kinase promoter and the rabbit p-globin promoter, the fragment -2041 -129 excised withHinfI was blunt-ended and ligated to the blunt-ended Sal1 site of pBLCAT8+ and pGCAT vectors (Kat0 et al., 1992). Synthetic fragments with XhoI and Sal1 overhangs corresponding to the regions -1541-134 (oligonucleotide 25 (oligonucleotide of sequence 5"CGGCGC- CCTCACTCACCTCGC-3')) and -1741-151 (oligonucleotide 24 (oligonucleotide of sequence 5'-CGTGTCGGTCACCGAGGCCCCGGC-3')) were ligated into the Sal1 site of pBLCAT8+ and pGCAT vectors. DR3GCAT was constructed by inserting DR3 (5"TAGTAGGTCAkA- GAGGTCAGACA'IT-3'1, which contains two directly repeated AGGTCA motifs spaced by 3 base pairs with HindIII andXbaI overhangs, into the corresponding site of pGCAT vector.

Cell Culture and 'I)ansfection"COS7 cells were maintained in Dul- becco's modified Eagle's medium supplemented with 5% dextran-coated charcoal-treated fetal calf serum. Transfection was achieved by electro- poration using GTE-10 (Shimadzu, Kyoto, Japan) and 10 pg of plasmid5 x 10' cells. The P-galactosidase reporter gene (pSV-P-galactosidase control vector from Promega) was co-transfected. Two days after transfection, 1,25-(OH),D3 was added at various concentrations to the medium. The cells were harvested 24 h later. CAT activity was determined (Gorman et al., 1982) using a Fujix 2000 Bio Image Ana- lyzer (Fuji Film, Tokyo, Japan) after normalization based upon P-galactosidase activity.

'I)anslation of VDR and RXR in Vitro-The plasmid p7Xf-hVDR containing a full-length copy of the human VDR cDNA (a gift from Dr.

Comparison of nucleic acid sequences of VDREs TABLE I

Gene Sequence", Location

Rat 24-hydroxylase AGGTGA gtg AGGGCG (-151--137) Rat osteocalcin GGGTGA atg AGGACA (-460~-446) Human osteocalcin GGGTGA acg GGGGCA (-499--485) Mouse osteopontin Rat calbindin-D9k

GGTTCA cga GGTTCA (-757--743)

Mouse calbindin-D28k GGGGGAt g t g AGGAGA (-198--182) GGGTGT Cgg AAGCCC (-489--475)

a All VDRE sequences are illustrated as direct repeat motifs. Sequences are from top: rat osteocalcin (MacDonald et al., 1991);

human osteocalcin (Ozono et al., 1990); mouse osteopontin (Noda et al., 1990); rat calbindin-D9k (Darwish and DeLuca, 1992); and mouse calbindin-D28K (Gill and Christakos, 1993).

J . W. Pike, Ligand Pharmaceuticals, San Diego, CA) and pSG5RXRp containing the coding region of rat RXRp cDNA (Mano et al., 1993) were linearized by enzyme digestion and utilized to produce capped RNA from the SP6 and T7 promoter, respectively. The RNAs were translated in the reticulocyte lysate system as described by the manufacturer (Promega). The synthesis of VDR was confrmed by SDS-polyacrylamide gel analysis and a ligand binding assay.

Gel Mobility Shi/? Assay-The nuclear extracts from COS7 cells were prepared as described (Sone et al., 1991a). Mobility shift assays were performed as described previously (Ozono et al., 1990). The DNA fragment -204/-129 was prepared from the promoter region by digestion with H i d . Two DNA fragments with XhoI and Sal1 overhangs corre-

CCCGGC-3') and -1541-134 (5'-CGGCGCCCTCACTCACCTCGC-3') sponding to the region -1741-151 (5'-CGTGTCGGTCACCGAGGC-

were synthesized and named oligonucleotides 24 and 25, respectively. A DNA fragment corresponding to the region -145/-98 was also synthesized. The DNA fragments were end-labeled with [Y-~~PIATP (111 TBq/ mmol) using T4 polynucleotide,kinase (about 10' cpdpg). The reticulocyte lysate (2 pl) containing the translated VDR was preincubated with 500 I ~ M 1,25-(OH),D3 for 20 min in buffer containing 25 lll~ Trid HCl, 15 m Hepes, pH 7.9,40 m NaCI, 5 II~M KC], 3 m MgCl,, 4.5 m EDTA, 6% glycerol, 0.08% Tween 20, 1 m dithiothreitol, and 1 pg of poly(d1-dC) in the presence of the nuclear extract (1 pg of protein) or reticulocyte lysate (2 pl) containing translated RXR. One microliter of DNA probe (lo4 cpdreaction) was then incubated with the reaction mixture (total 20 111) for 20 min at room temperature, then the mixture was analyzed by a 5% polyacrylamide gel electrophoresis in 50 m Tris, 380 lll~ glycine buffer, pH 8.5, at 200 V a t 4 "C. The gels were dried and visualized by autoradiography overnight. The oligonucleotides used for the competition assay had the VDRE sequence of human osteocalcin -51W-483 (Ozono et al., 1990); 5'"N'GGTGACTCACCGGGT- GAACGGGGGCA'M-3' with HindIII overhangs (named oligonucleotide 6) and artificial VDRE; 5'-TAGTAGGTCAAAGAGGTCAGACA'M-3' (named DR3) (Umesono et al., 1991).

RESULTS

The 5"Flanking Region of the Rat P45Occ24 Gene-The 5'- upstream region of the rat P45Occ24 gene was sequenced using the genomic clone p30-4.5 (Ohyama et al., 1993) which covered 2.9 kilobase pairs of the 5'-flanking region. The region extend- ing from the nucleotide -1185 to the transcription initiation site (+1) contained a TATA box (at -30) and a CCAAT box (at -58). We previously reported (Ohyama et al . , 1993) that the 5"flanking region up to -516 contains four candidate VDREs -4271-413, -4231-409, -2101-196, and -1751-161, based upon a homology search using the sequences of rat osteocalcin VDRE (GGGTGAAATGAGGACA) (MacDonald et al., 1991) and rat calbindin D-9k VDRE (GGGTGTCGGAAGCCC) (Darwish and DeLuca, 1992). However, these sequences did not have a direct repeat motif. In the present study, other VDRE-like sequences containing direct repeat motifs were found in the regions -2591 -245, -1691-155, -1511-137, and -142/-1281 by a computer search using the sequences of rat osteocalcin (GGGTGA- NNNAGGACA), human osteocalcin (GGGTGANNNGGGWA), mouse osteopontin (GGTTCANNNGGTTCA), and rat calbindin-D9k (GGGTGTNNNAAGCCC) genes (Table I).

Induction of 24-Hydroxylase by Vitamin D in COS Cells- Monkey kidney-derived COS cells contain 24-hydroxylase ac-

Vitamin D-responsive Element in the Vitamin D 24-Hydroxylase Gene 10547

1,25(OH)2D3 250HD3 24,25(OH)2D3

o l o 4 10-7 10-8 l o 6 lo4 IO-^ IO*(M) -

I I 0

Time (h) 10 20

FIG. 1. The induction of 24-hydroxylase activity in COS cells. COS7 cells, 5 x 10fi/9.5-cm2 dish, were cultured with 4 x 1O"j M 25- (OH)D,. At each time point, cells were harvested, and the enzyme activity was measured by the reconstitution system using adrenodoxin and adrenodoxin reductase (Ohyama and Okuda, 1991).

tivity at a high level (Ohyama et al., 1991). As shown in Fig. 1, 25-(OH)D, induced markedly the 24-hydroxylase in the cells despite the low levels of VDR in these cells (about 2000 VDRs/ cell, Baker et al., 1988). The increase in mRNA and protein levels were verified by Northern and Western blotting using p108 cDNA (Ohyama et al., 1991) and a monoclonal antibody (Ohyama and Okuda, 1991) to the rat P45Occ24, respectively (data not shown). These results suggested that the expression of P45Occ24 is controlled by the vitamin D status and that the COS cells contain the nuclear factors that are required for P45Occ24 expression. COS cells are likely to be an appropriate model with which to analyze the functions of the 24-hydroxylase gene.

Effects of Vitamin D Analogs on the Promoter Activity of P450cc24-In the COS cells, 1,25-(OH),D3 at and M enhanced the expression of the CAT construct containing the region -646/+9 in a dose-dependent manner (Fig. 2). 25-(OH)D, and 24,25-(OH),D, at M were also effective in stimulating CAT activity, although they had little effect at 10"' and M. In contrast, a higher concentration (5 x 1O"j M) of vitamin D, had little effect on CAT activity (data not shown). The difference between various vitamin D metabolites in inducing the promoter activity of P45Occ24 may be related to the binding affinity of vitamin D metabolites to VDR (Baker et al., 1988).

Deletion Analysis of the Promoter Region-The COS cells were transfected with several deletion mutants of the 5"flanking region fused to the CAT reporter gene (pCAT-basic, Pro- mega). In the first series of experiments, we used five CAT constructs that contained the promoter region between -2.2 kilobase pairs to -167 base pair at the 5' terminus and at +188 at the 3' terminus. All of the constructs responded to 1,25- (OH),D, at M about 20-30-fold, although the construct -167/+188 showed only 3-fold induction (Fig. 3A). This indi- cates that the VDRE of 24-hydroxylase gene is located near the TATA box, although the VDREs of osteocalcin, osteopontin, and calbindin-D9k are about 500 base pairs upstream from the initiation site (Table I). In the second series of experiments, we prepared four CAT constructs that carried the regions -291/+9, -167/+9, -102/+9 or -70/+9. As shown in Fig. 3B, the constructs -167/+9 and -291/+9 responded to 1,25-(OH),D, to a similar extent (10-fold). However, The constructs -102/+9 and -70/+9 did not respond to 1,25-(OH),D,. The deletion of the region -167/-102 from the construct -950/+9 decreased the 1,25- (OH),D, responsiveness to a very low level (2-fold). These results suggest that the segment between -167 and -102 plays a major role in the vitamin D response of the P45Occ24 gene.

Determination of VDRE Using a Heterologous Promoter- The fragment -204/-129 excised with the HinfI enzyme from the promoter was fused to the region upstream of the heterolo-

Fold 1.0 1.2 4.4 18.4 1.0 1.2 16.5 1.0 1.0 8.8 Induction f f f f ? r f f f f

0.1 1.3 3.7 0.1 0.1 1.4 0.1 0.1 0.1

FIG. 2. The dose dependence of the effect of vitamin D metabolites on the promoter activity. COS7 cells were transfected with a deletion construct containing the -646/+9 fragment, and cultured in the presence of increasing concentrations of 1,25-(OH),D,, 25-(OH)D,, or 24,25-(OH),D,. Values are the means f S.E. from three independent transfections.

A 96 Conversion Fold Induction 1.25(OH)&- _ " + -

- 2.2 kl +la8 29.2 f 7.4 0.8 t 0.4 36.5 - 950 t #y 40.7 f 4.8 1.6 t 0.6 25.4

-646 I 44.9 f 6.3 1.3 f0 .7 34.5 -291 - CAT 41 .8 t7 .1 2 . 6 f l . 3 16.1

- 167 I 10.2f 3.2 3.2t 1.7 3.2

pCAT-basic 1.2tn.5 1.120.5 1 . 1

B %Conversion Fold Induction

+ I.ZS(OHhD3 - - 291 44.2 f 4.6 4.4 C 1.0 10.0

- 167 -? 51.4f5 .6 S . h t 1.8 9.2 - 102 rn 1.6 f 0.2 1.7 f 0.2 0.9

-70 CAT 5 . 4 2 I.? 5 .42 1.0 1.0

- 950 I rn 4.8 2 1.4 2.2 20.6 2.2 - 167 '4 102

FIG. 3. Deletion analysis of the rat P45Occ24 promoter. COS7 cells were transfected with various CAT constructs containing promoter fragments as indicated. Numbers on the left of each deletion represent the 5'-end of the DNA sequence relative to the transcription initiation site. A, low resolution analysis. Each fragment of the promoter was excised by XbaI (-2.2 kilobases), Sac1 (-950), HincII (-646), PstI (-291), and BstEII (-167) at the 5'-end, and SmaI (+188) at the 3'-end from P45Occ24 gene. These fragments were ligated into the pCAT-basic vector. After transfection of 20 pg of the CAT constructs cells were maintained for 2 days, then cultured in the absence (-) or presence (+) of 10" M 1,25-(OH),D, for 24 h. B, detailed deletion analysis. Each fragment was prepared by digestion with PstI (-291). BstEII (-167), BstUI (-1021, and DdeI (-70) at the 5'-end and with RsaI (+9) a t the 3'-end. Adeletion mutant between -167 and -102 was constructed by inserting the fragment -950/-167 into the construct -102/+9. Cells were exposed to 1,25- (OH),D, as described for experiment A. Values are the means 2 S.E. from three independent transfections.

gous promoter @-globin or thymidine kinase), and the fused constructs were used for the CAT assay. These constructs ex- hibited the vitamin D responsiveness in an orientation-independent fashion (about 6-fold) (Table IIA), indicating that this fragment works as an enhancer. As shown in Fig. 4A, the region -204/-129 contained two putative VDREs. Thus, we synthesized two oligonucleotides (oligonucleotides 24 and 25) and fused them to the P-globin promoter. In addition, we prepared a DR3GCAT construct containing DR3, designated as the consensus of VDRE (Umesono et al., 1991). Exposure to 1,25- (OH),D, a t M caused a 2.5-fold increase in CAT activity in the cells transfected with the construct containing oligonucleotide 25 or DR3, whereas the oligonucleotide 24 construct showed only a 1.2-fold induction (Table IIB).

10548 Vitamin D-responsive Element in the Vitamin D 24-Hydroxylase Gene TABLE I1

7hzscriptional activity of the fragment of 5’-flanking region of P45Occ24 gene through the heterologous promoters

Transfection and vitamin D treatment were the same as described in the legend to Fig. 3. -Fold inductions are the means 2 S.E. from three independent transfections. A, fragment -2041-129 was blunt-ended by Klenow fragment and ligated to the blunt-ended Sal1 site of pBLCAT8+ (containing the thymidine kinase promoter) and pGCAT (containing the P-globin promoter). GLOB, globin promoter; TK, thymidine kinase promoter; Control, vector only; F, forward orientation; R, reverse orientation. B, two oligonucleotides containing putative VDRE sequences (illustrated in Fig. 4A) and DR3 were ligated into pGCAT as described under “Experimental Procedures.”

A. DNA fragment -204 to -129

Promoter Orientation -Fold Induction

GLOB Control 1.0 2 0.1 GLOB F GLOB R

6.3 2 0.6

TK 5.8 f 1.0

TK Control 1.1 * 0.1

F 5.0 * 0.3 TK R 6.1 2 1.6

B. Synthetic DNA fragments

Promoter Fragment Orientation -Fold Induction

GLOB Oligonucleotide 25 F GLOB Oligonucleotide 25 R GLOB Oligonucleotide 24 F GLOB Oligonucleotide 24 R GLOB DR3 F

2.5 2 0.2 2.6 2 0.5 1.2 * 0.1 1.2 * 0.1 2.6 2 0.1

Gel Mobility Shift Analysis-The human vitamin D receptor and the rat RXRP were synthesized by in vitro translation of their respective cDNAs. In the gel shift assay using the fragment -204/-129, a retardation band appeared only in the co- presence of 1,25-(OH),D,, VDR, and nuclear extracts of COS cells (or RXRP) (Fig. 4 B , lanes 3 and 5). In contrast, no protein- DNA complex was observed when the synthetic fragment -145/ -98 was used (data not shown). Removal of either VDR, nuclear extracts, or 1,25-(OH),D, abolished the formation of retarded band (Fig. 4B, lanes 1, 2, 4, and 6). The band also disappeared when the protein-DNA complexes were formed in the presence of the anti-VDR monoclonal antibody 9A7 (Ozono et al., 1991) (data not shown). RXRP, a steroid hormone receptor, works as a counterpart of VDR (Yu et al., 1991). In fact, RXRP could substitute for the nuclear extracts of COS cells under our conditions. It has not been determined whether the factor in COS cells extracts is RXR, but it is noteworthy that the complex formed in the presence of nuclear extracts or RXRP showed the same mobility on polyacrylamide gels (Fig. 4B, lanes 3 and 5, and Fig. 5 , lanes 5 and 7).

Furthermore, oligonucleotide 6, DR3, and oligonucleotide 25, which contain human osteocalcin VDRE, the consensus sequence of VDRE, the proximal VDRE-like sequence, respectively, competed with the fragment -204/-129 for the formation of the retarded band (Fig. 4B, lanes 7, 8, and IO). However, oligonucleotide 24 bearing the distal direct repeat had less effect. These results suggest that the sequence containing oligonucleotide 25 has a higher affinity for the VDR.RXR complex than oligonucleotide 24, even though they have similar VDRE- like sequences.

Finally, oligonucleotides 24 and 25 were used as probes in the gel mobility shift assay in the presence or absence of VDR, nuclear extracts of COS cells, RXRP, and 1,25-(OH),D3 (Fig. 5) . Only oligonucleotide 25 produced the retarded band solely in the presence of both nuclear accessory factor(s) and 1,25- (OH),D,. These results confirm that the sequence AGGTGAGT- GAGGGCG is required for efficient binding of the VDR-RXR complex.

A -204

I 5’-AGTCCACCGGTGCGTCTGCCGGGCCAGCAG

GGTGGCCACGCAGACGGCCCGGTCGTC Oligo24

~ ~ T G ~ ~ ~ C ~ ~ ~ C G G ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ GCACAGCCAGTGGCTCCGGGGC+GEE~~GT$ZZE~GCGACTGA-S

I < * I Olig025 -1 67 -1 29

B 1 2 3 4 5 6 7 8 9 1 0

VDR : + - + + + + + + + + cos : - + + + - - + + + + RXR : - ” - + + ” ”

1,25(OH)ZD3 : + + + - + - - + + + Competitor : - - - - - - 6 DR3 24 25

FIG. 4. M)R-M)RE complex formation in gel mobility shift assay. A, the Hinfl-digested fragment -2OU-129 containing putative VDREs was used as a probe. Two oligonucleotides covering each VDRE were synthesized and designated oligonucleotide 24 and oligonucleotide 25, as indicated in the diagram. B, VDR and R X R , two microliters of the reticulocyte lysate containing in vitro translated VDR or RXR, were used. COS, nuclear extracts (1 pg/l pl) of COS cells was prepared according to the method of Sone et al. (1991a). One microliter of the extract was added to the reaction mixture. 1,25-(OH),D,, 1 pl of 1,25- (OH),D, (10 y) was used. Competitor, 6,24, and 25 indicate oligonucleotides 6, 24, and 25, respectively. oligonucleotide 6 contained human osteocalcin WIRE -512/-483. DR3 is an artificial M R E containing two directly repeated AGGTCAmotifs spaced by 3 base pairs. One microliter of each competitor (20 ng) was added to the mixture with the probe. In all experiments, the final volume of the mixture was 20 pl.

DISCUSSION

To identify a hormone response element, the following criteria are necessary: 1) the element in the native promoter mediates the hormone effect, 2) the element fused to heterologous promoteds) mediates the hormone effect in an orientation-independent manner, and 3) the hormone receptor directly binds to the element. These criteria are satisfied by the VDRE -151/ -137 in the rat 24-hydroxylase gene, as indicated by the three observations in the present study: (a) the region -167/-102 in the homologous promoter was required for the 1,25-(OH),D, response (Fig. 3, A and B ), ( b ) oligonucleotide 25 containing the region -El/-137 fused to the P-globin promoter in both forward and reverse directions increased CAT activity in response to 1,25-(OH),D3 (Table IIB), (c) oligonucleotide 25 bound to VDR in the gel mobility shift assay (Fig. 5) .

Besides the VDRE -El/-137, other VDRE-like sequences are present in the promoter region of the rat 24-hydroxylase gene. I t remains unknown whether or not these VDRE-like sequences are also involved in the hormone response. The unidirectional deletion of the region -291/-167 from the native promoter which had a 3‘-end a t +188, significantly decreased

Vitamin D-responsive Element in the Vitamin D 24-Hydroxylase Gene 10549

1 2 3 4 5 6 7 8

Probe : e 24 e 25

VDR : + + + + + + + + cos : + + - - + + - - RXR : - - + + - - + +

1,25(OH)ZD3 : + - + - + - + - FIG. 5. Band shift analysis using synthetic oligonucleotides.

Each of two oligonucleotides (Fig. 4A) covering the putative M R E was synthesized and used as probes. The experimental conditions were as described in the legend to Fig. 4.

the hormone response (Fig. 3A), although the same deletion had little effect upon the construct which had a 3'-end a t +9 (Fig. 3B). This discrepancy may be due to the stability of the transcribed RNA. The deletion mutant of the promoter region -950/+9 which lacked the region -167/-102 retained a 2-fold response to 1,25-(OH),D, (Fig. 3B). These findings suggested that the upstream VDRE-like sequences may also have some role in the hormone response in certain situations, but the effect did not appear as important as the VDRE in the proximal region -167/-102.

The region -204/129 induced CAT activity in response to 1,25-(OH),D, through the heterologous promoters (6-fold). This segment contained two sets of the direct repeat VDRE motif (Fig. 4A). The proximal direct repeat in the antisense strand formed a complex with VDR in the gel mobility shift assay (Fig. 5 ) , whereas the distal element (oligonucleotide 24) did not. However, oligonucleotide 24 competed markedly with the fragment -204/129 for the complex formation (Fig. 4B). The VDRE -154/-134 (proximal direct repeat) fused to the P-globin promoter conferred only 2.6-fold activation, whereas the fragment -204/-129, which contained both direct repeats showed about 6-fold activation, when inserted into the same heterologous CAT vector. These results suggested that the distal element provides the second VDR binding site, although it has less effect on the vitamin D response. I t is known that multiple copies of VDREs fused to heterologous promoters are more effective than a single copy (Kerner et al., 1989). Thus, the distal VDRE-like sequence may facilitate the recognition of VDR and/or RXR.

The sequences of the VDRE in the rat 24-hydroxylase gene and the other VDREs identified previously are summarized in Table I. The alignment of the VDRE reveals that the VDRE in the rat 24-hydroxylase gene, like the sequences in gene of osteocalcin, osteopontin, and calbindin-D9k genes, consists of a direct repeat of two half-sites separated by a 3-base spacer. The VDRE in the rat 24-hydroxylase gene is located near the transcription initiation site, as compared with the other VDREs. The unique feature is that the direction of VDRE in the 24- hydroxylase gene is the opposite of that of the other VDREs. Recent studies have shown that the VDRE in the parathyroid hormone gene has the same direction as that of the rat 24- hydroxylase, although this element has one half-site and mediates the negative effect of 1,25-(OH),D, (Demay et al., 1992).

A comparison of the half-site sequences suggested that the sequence of the upstream half-site is more conserved than that of the downstream (Table I), being GGGTGA and (A/G)GG(G/ A/C)CA, respectively. The molecular events of vitamin D ac- tions are thought to be mediated by the binding of the VDR.RXR heterodimer complex to VDRE (Sone et al., 1991a, 1991b). The difference between the two half-site sequences may indicate preferential binding of each receptor to either. Recent studies have shown that RXRs preferentially bind to the upstream half-site in retinoic acid and thyroid hormone response elements as well as in VDRE (Perlmann et al., 1993; Kurokawa et al., 1993).

The expression of the rat 24-hydroxylase gene is up-regulated by 1,25-(OH),D,, but down-regulated by parathyroid hormone in the kidney (Shinki et al., 1992). Chen et al. (1993) found that 12-0-tetradecanoylphorbol 13-acetate increased the rat 24-hydroxylase gene expression in the presence of 1,25- (OH),D, in renal primary cell cultures. Recently, Uchida et al. (1993) reported that the phorbol ester modulates 24-hydroxylase expression through the promoter region between -169 and +9 in the presence of 1,25-(OH),D,. The effect of the phorbol ester could be mediated by an AP-1-like sequence in the region of -140/-146 overlapping with proximal VDRE in rat 24-hydroxylase. In the human osteocalcin and mouse calbindin-D28k genes, the AP-1-responsive site, and the sodium butyrate-responsive element are juxtaposed VDREs, respectively (Ozono et al., 1990; Gill and Christakos, 1993).

In conclusion, the expression of P45Occ24 is modulated through the VDRE located between nucleotides -151 and -137 in the rat 24-hydroxylase gene promoter and that this segment contains a sequence homologous to the known VDREs. The identification of the VDRE in the rat 24-hydroxylase gene will be useful for understanding the molecular mechanism by which 1,25-(OH),D, regulates target genes.

REFERENCES

Baker, A. R., McDonnel, D. P., Hughes, M., Crisp, T. M., Mangelsdorf, D. J.,

Acad. Sci. U. S. A. 85,3294-3298 Haussler, M. R., Pike, J. W., Shine, J., and OMalley, B. W. (1988) Proc. Natl.

Brommage, R., and DeLuca, H. F. (1985) Endocr Reu. 6,491-511 Chen, M. L., Boltz, M. A,, and Armbrecht, H. J. (1993) Endocrinology 132,1782-

Danvish, H. M., and DeLuca, H. F. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,

DeLuca, H. E (1988) FASEE J. 2,224-236 Demay, M. B., Gerardi, J. M., DeLuca, H. F., and Kronenberg, H. M. (1990) Proc.

Natl. Acad. Sci. U. S. A. 87, 369-373 Dernay, M. B., Kiernan, M. S., DeLuca, H. F., and Kronenberg, H. M. (1992) Proc.

Natl. Acad. Sci. U. S. A. 89, 8097-8101 Gill, R. K., and Christakos, S. (1993) Proc. Natl. Acad. Sci. U. S. A. 90,2984-2988 Gorman, C. M., Moffat, L. F., and Howard, B. H. (1982) Mol. Cell. Biol. 2, 1044-

1051 Haussler. M. R., Mangelsdorf, D. J., Komm, B. S., Terpening, C. M., Yamaoka, K.,

Allegretto, E. A,, Baker, A. R., Shine, J., McDonnell, D. P., Hughes, M., Weigel, N. L., O'Malley, B. W.. and Pike, J. W. (1988) Recent Prog. Horm. Res. 44,

Kato, S., Tora, L.,Yamauchi. J., Masushige. S.. Bellard, M., and Chambon, P. (1992) 263-305

Kerner, S. A., Scott, R. A,, and Pike, J. W. (1989) Proc. Natl. Acad. Sci. U. S. A. 86, Cell 68,731-742

44554459 Kurokawa, R.. Yu, V. C., NBlr, A,, Kyakumoto, S., Han, Z., Silverman, S., Rosen-

Liao, J., Ozono, K., Sone, T., McDonnell, D. P., and Pike, J. W. (1990) Proc. Natl. feld, M. G., and Glass, C. K. (1993) Genes & Deu. 7, 1423-1435

MacDonald, P. N., Haussler, C. A,, Terpening, C. M., Galligan, M. A,, Reeder, M. C., Acad. Sci. U. S. A. 87,9751-9755

Mano, H., Ozawa, T., Takeyama, K., Yoshizawa, Y., Kojima, R., Kato, S., and Whitfield, G. K., and Haussler, M. R. (1991) J. Eiol. Chem. 266, 18808-18813

Markose, E. R., Stein, J. L., Stein, G. S., and Lian, J. B. (1990) Proc. Natl. Acad. Masushige, S. (1993) Biochem. Biophys. Res. Commun. 191,943-949

Minghetti, P. P., and Norman, A. W. (1988) FASEE J. 2,3043-3053 Sci. U. S. A. 87, 1701-1705

Nakamura, T., Nagai, Y., Yarnato, H., Suzuki, K., and Orimo, H. (1992a) Calcif

1788

603-607

Nakamura, T., Suzuki, K., Hirai, T., Kurokawa, T., and Orimo, H. (1992b) Bone 13, Tissue Int. 50, 221-227

Noda, M., Vogel, R. L., Craig, A. M., Prahl, J., DeLuca, H. F., and Denhardt. D. T.

Ohyama, Y., and Okuda, K. (1991) J. Biol. Chem. 266,8690-8695 Ohyama. Y., Hayashi, S.-I., and Okuda. K. (1989) FERS Lett. 255,405-408

229-236

(1990) Proc. Natl. Acad. Sci. U. S. A. 87,9995-9999

10550 Vitamin D-responsive Element in the Vitamin D 24-Hydroxylase Gene Ohyama, Y., Noshiro, M., and Okuda, K. (1991) FEBS Lett. 278, 195-198 Ohyama. Y., Noshiro, M., Eggertsen, G., Gotoh, O., Kato, Y., Bjorkhem, I., and

Ozono, K., Liao, J., Kerner, S. A,, Scott, R. A., and Pike, J. W. (1990) J. Biol. Chem.

Ozono, K., Sone, T., and Pike, J. W. (1991) J. Bone Miner: Res. 6, 1021-1027 Perlmann, T., Rangarajan, P. N., Umesono, K., and Evans, R. M. (1993) Genes &

Pike, J. W. (1991) Annu. Rev. Nutr: 11, 189-216 Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular CZoning: A

Shinki, T., Jin, C. H., Nishimura,A., Nagai, Y., Ohyama,Y., Noshiro, M., Okuda, K., Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY

Okuda, K. (1993) Biochemistry 32,76432

266, 21881-21888

Dew. 7, 1411-1422

Sone, T., Kerner, S., and Pike, J. W. (1991a) J. Biol. Chem. 266,23296-23305 Sone. T.. Ozono. K.. and Pike. J. W. (1991b) Mol. Endocronol. 6. 1578-1586

and Suda, T. (1992) J. Biol. Chem. 267, 13757-13762

~ ~~ ~. I I~ ~. ~~ ~~

Suda, T., Shinki, T., and Takahashi,’N. (19901 Annu. Reu. Nutc’lO, 195-211 Tanaka, Y., and DeLuca, H. F. (1974) Science 183,1198-1200 Uchida, M., Shinki, T., Ohvama. Y., Noshiro, M.. Okuda, K., and Suda, T. (1993) J.

Umesono, K., Murakami, K. K., Thompson, C. C., and Evans, R. M. (1991) Cell 66, Bone Miner: Res. 8, S l i l

1255-1266 Yanisch-Perron, C., Vieira, J., and Messing, J. (1985) Gene (Amst.) 33, 103-109 Yu, V. C., Delsert, C., Andersen, B., Holloway, J. M., Devary, O., Naar, A. M., Kim,

S. Y., Boutin, J-M., Glass, C. K., and Rosenfeld, M. G. (1991) Cell 67,1251-1266

TltE OF BIOUXlcaL CHEMISTRY Vol. 269, No. 14, Issue April ...TltE JOliRNAL OF BIOUXlcaL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 269,

Documents