Cloning and Characterization of the Type I Inositol 1,4,5-Trisphosphate Receptor Gene Promoter

The inositol 1,4,5-trisphosphate (InsP3) receptor is essential for signal Ca2+ release from intracellular stores and for capacitative Ca2+ entry. We have isolated the promoter and proximal DNA segments of the human type I InsP3 receptor gene. Transcription initiation in human G-292 osteosarcoma and HL-60 promyelocytic leukemia cells was shown to occur predominantly from an adenine residue located 39 base pairs downstream of a consensus TATA box element. Upstream DNA including the TATA box promoted directional transcription of a chloramphenicol acetyltransferase reporter gene when transfected into G-292 cells. A negative regulatory element in the distal promoter and a positive element in the proximal region were identified by deletion mapping and transcription assays. The proximal region enhanced transcription in response to 12-O-tetradecanoylphorbol-13-acetate or serum, but conferred transcriptional repression in response to 1,25-dihydroxyvitamin D3 or 17β-estradiol. The repressive effect of 17β-estradiol was mediated by the nuclear estrogen receptor, as estrogen-dependent transcriptional repression was inhibited by the antiestrogen tamoxifen and the estrogen receptor antagonist ICI 182,780. This is the first study of the type I InsP3 receptor gene promoter, and the results suggest a mechanism by which chronic estrogen treatment of osteoblasts affects type I InsP3 receptor gene expression, signal transduction, and secretion.

Inositol 1,4,5-trisphosphate (InsP 3 ) 1 mediates the calciummobilizing effects of a wide range of hormones, cytokines, and neurotransmitters (1). Activation of G-protein-linked and tyrosine kinase-linked receptors stimulates phosphatidylinositolspecific phospholipase C-mediated hydrolyis of phosphatidylinositol 4,5-bisphosphate with the resultant generation of the second messenger molecules, diacylglycerol, and InsP 3 . InsP 3 binds to specific receptors located on endoplasmic reticulum, causing the release of sequestered Ca 2ϩ into the cytosol and the stimulation of Ca 2ϩ -dependent processes including secretion (1)(2)(3).
InsP 3 receptors have been purified as 250 -313-kDa proteins and have been shown to function as tetrameric Ca 2ϩ channel complexes whose conductance is regulated by the binding of InsP 3 (4,5). Three distinct, highly homologous (Ͼ60%) cDNAs of 8.8 -10.7 kb encoding full-length InsP 3 receptors have been cloned from rodent, human, and other sources (6 -13). Partial sequences of two additional InsP 3 receptor types, which are highly homologous to type II, have also been identified (14). Recent studies have shown that individual InsP 3 receptor types are variably expressed in different tissues, and that multiple receptor types can be expressed within a given cell and these may assemble as heterotetramers in a single Ca 2ϩ conductance channel (15,16). Furthermore, InsP 3 -induced calcium release and InsP 3 receptor-mediated calcium entry may be variably controlled by different InsP 3 receptor subtypes (17,18).
The expression and function of InsP 3 receptors are hormonally regulated. Treatment of promyelocytic HL-60 cells with retinoic acid or 1,25-dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ) stimulates InsP 3 receptor gene transcription, increases the expression of functional InsP 3 -regulated intracellular Ca 2ϩ channels, and up-regulates cell secretory capacity (19,20). In pancreatic islets of fasted and then fed rats and also in insulin-secreting RINm5F cells treated chronically with high glucose, type III InsP 3 receptor expression is increased (21). On the other hand, stimulation of neuronal and pancreatoma cells with phosphoinositide-linked hormones causes the rapid down-regulation of the type I InsP 3 receptor protein (22,23). In primary osteoblasts and osteoblastic cell lines, type I InsP 3 receptor messenger RNA expression is decreased by 17␤-estradiol and 1,25(OH) 2 D 3 but increased by phorbol esters and serum (24). These studies suggest that chronic regulators of cell secretory capacity may affect these changes through regulation of InsP 3 receptor levels.
As an initial step to help understand how InsP 3 receptor gene expression may be regulated under different physiological or pathological situations, we have isolated genomic clones encoding the promoter and six of the initial exons of the human type I InsP 3 receptor gene. The major transcription start sites are identified, and it is shown that the type I InsP 3 receptor gene promoter functions in human osteoblastic cell lines to regulate transcription of a reporter gene. Furthermore, it is shown that several chronic regulators of osteoblasts secretory activity affect type I InsP 3 receptor promoter activity. Significantly, 17␤-estradiol via the estrogen receptor down-regulates promoter activity, consistent with its effects on type I InsP 3 * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AF009963.

EXPERIMENTAL PROCEDURES
Cell Culture-Cells from the G-292 human osteosarcoma cell line (ATCC CRL 1423) were cultured in McCoy's 5a medium (Life Technologies, Inc.) supplemented with 10% fetal bovine serum, 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin at 37°C in a humidified atmosphere of 5% CO 2 in air. Experiments with G-292 cells were conducted in phenol red-free Dulbecco's modified Eagle's medium (Life Technologies, Inc.) and 2% charcoal-stripped fetal bovine serum depleted of endogenous steroids as described (24,25). G-292 cells have osteoblastic phenotype in that they express osteocalcin mRNA, bone morphogenic protein (BMP-7) receptor, and secrete interleukin-6 in response to interleukin-1␤ and other osteoblast secretagogues (26). Cells from the human promyelocytic leukemia HL-60 cell line (ATCC CCL 240) were cultured and maintained in RPMI 1640 medium with 20% fetal bovine serum as described previously (19).
Isolation of Human Genomic DNA Clones Containing the Type I InsP 3 Receptor Gene Promoter-A human placental genomic DNA library in bacteriophage FIXII (Stratagene Inc., La Jolla, CA) was screened by plaque hybridization (27) to obtain type I InsP 3 receptor gene clones. Probes for the human type I InsP 3 receptor gene were derived by PCR amplification of human HL-60 cell cDNA using primers based upon the mouse and rat type I InsP 3 receptor cDNA sequences (7,8). Primer pairs were numbered according to the mouse type I InsP 3 receptor cDNA and included: p489s, 5Ј-GAG ACT GCC TCT TTA AGC TAT GTC C-3Ј (489 -513 nt) and p970a, 5Ј-CAG TTG ACG GAG TTG ACC TCA TTG C-3Ј (antisense of 945-969 nt); p85s, 5Ј-GAG AGA AAG CGC ACG CCG AGA GGA G-3Ј (85-109 nt) and p513a, 5Ј-GGA CAT AGC TTA AAG AGG CAG TCT G-3Ј (antisense of 489 -513 nt); and p85s, 5Ј-GAG AGA AAG CGC ACG CCG AGA GGA G-3Ј (85-109 nt) and p255a, 5Ј-GCT CCT CTG TAG TCA GCT CCT TGG T-3Ј (antisense of 241-255 nt). The p85s oligonucleotide was selected as the most 5Ј primer in the cDNA because the reported mouse and rat type I InsP 3 receptor cDNA sequences diverge upstream of this point. PCR products were gel-purified, subcloned into pGEM-11Zf(ϩ) (Promega Corp., Madison, WI), and sequenced to confirm their identities.
Type I InsP 3 receptor cDNA probes were labeled with [␣-32 P]dCTP by the random primer method and column-purified. For library screening, SRB/P2 host bacteria (Stratagene Inc., La Jolla, CA) were infected with the FIXII phage and plated at 5 ϫ 10 4 plaque-forming units onto each of 10 150-mm plates. Resultant plaques were transferred in duplicate to Colony/Plaque Screen nylon membranes (NEN Life Science Products). The membranes were prehybridized and hybridized according to manufacturer's recommendations using 5 ϫ 10 5 cpm/ml 32 P-labeled type I InsP 3 receptor cDNA probe (1 ϫ 10 9 dpm/g). Positive plaques were identified by autoradiography and selected through secondary and tertiary screenings. Purified phage DNAs were digested with various restriction enzymes for mapping, hybridization, subcloning, and sequencing. For analysis of the type I InsP 3 receptor gene promoter region, a 3.2-kb XbaI/HindIII fragment containing 5Ј-flanking gene sequences, the first exon, and part of the first intron was subcloned into pGEM-11Zf(ϩ). An internal 1.5-kb BglII fragment of the XbaI/HindIII fragment was subcloned and sequenced from both strands in its entirety using M13 and internal primers. DNA sequencing reactions were performed by the dideoxy chain termination method using modified T7 DNA polymerase (Sequenase version 2.0, U. S. Biochemical Corp.) and [␣-35 S]ATP (NEN Life Science Products). Sequences were analyzed using MatInspector version 2.0 (28).
Primer Extension Analysis-Primer extension was carried out as described (27) with modification. A 25-mer oligonucleotide, p109a, corresponding to the antisense sequence of bases ϩ109 through ϩ85 of the reported mouse type I InsP 3 receptor cDNA (7) (ϩ37 through ϩ13 of the later reported human sequence; Ref. 11) was end-labeled using [␥-32 P]ATP (7,000 Ci/mmol, ICN Biomedical Inc, Costa Mesa, CA) and T4 polynucleotide kinase and purified using Bio-spin 6 columns (Bio-Rad). The labeled primer (50 fmol, 4 ϫ 10 5 cpm/reaction) was allowed to anneal to 20 g of total cellular RNA derived from either G-292 osteoblasts or HL-60 promyelocytic cells by incubation for 1 h at 62°C in Superscript II reverse transcriptase buffer (Life Technologies, Inc.) containing RNasin. The reactions were cooled to 46 -48°C, and then extended for 25 min by addition of dNTPs and Superscript II reverse transcriptase. The mixtures were subjected to RNase A digestion, phenol/chloroform extraction, and ethanol/sodium acetate precipitation.
Samples were analyzed by electrophoresis on denaturing 6% polyacrylamide gels alongside DNA sequencing reactions of the 0.5-kb human type I InsP 3 receptor SacI genomic subclone using the same p109a type I InsP 3 receptor oligonucleotide as used for the extension.
Type I InsP 3 Receptor Gene Promoter-CAT Reporter Plasmid Constructs-The eukaryotic expression vector pCAT-Basic (Promega Corp, Madison, WI), which contains the bacterial chloramphenicol acetyltransferase (CAT) gene but lacks promoter and enhancer sequences, was used as the parent vector in constructing reporter plasmids for the analysis of putative InsP 3 receptor promoter and transcriptional regulatory sequences. The 1.5-kb BglII restriction fragment, containing bp Ϫ1150 through ϩ350 of the human type I InsP 3 receptor gene, was filled-in and blunt end-ligated into the filled-in, XbaI-cut pCAT-Basic vector. Clones with the promoter in the forward orientation in front of the CAT gene were isolated, digested with PstI and SalI, and subjected to unidirectional digestion using exonuclease III followed by mung bean nuclease. This produced a series of constructs containing progressively larger deletions starting from the SalI site in the vector multiple cloning site and proceeding into 5Ј end of the upstream region of the promoter construct. The resultant DNAs were ligated and used to transform competent DH5␣ Escherichia coli. Recombinant pCAT constructs were purified by Midi Kit procedures (Qiagen), sequenced to identify the extent of the deletion, and used in transient transfection assays. In some experiments, the 0.5-kb SacI restriction fragment including bp Ϫ359 through ϩ153 of the type I InsP 3 receptor gene was cloned in front of the CAT gene of the pCAT-Basic vector and used in transient transfections.
Transfection and CAT Expression Assays-G-292 cells were cultured to ca. 50% confluency in 60-mm 2 tissue culture dishes, washed twice with Opti-MEM (Life Technologies, Inc.), and then transfected with a DNA-LipofectAMINE (Life Technologies, Inc.) mixture according to manufacturer's instructions. Equal molar amounts of the various recombinant pCAT-Basic expression vectors (approximately 3 g) were transfected along with a fixed amount (1.5 g) of the pSV-␤-galactosidase control vector (Promega Corp.). Cells were incubated for 12-14 h with the DNA-LipofectAMINE mixture, washed twice, and then incubated with phenol red-free Dulbecco's modified Eagle's medium containing 2% charcoal-stripped fetal bovine serum. Hormones and other test agent were then added. After 48 -72 h, the cells were harvested, washed, and soluble extracts prepared by repeated freeze-thawing in 0.15 ml of 0.25 M Tris, pH 8.0, and used for determination of ␤-galactosidase and CAT activities. ␤-Galactosidase activity in 20 l of each sample was determined spectrophotometrically as described (29) using o-nitrophenyl-␤-D-galactopyranoside as substrate. After normalization for ␤-galactosidase activities, CAT activities in the soluble extracts (usually ϳ50 -100 l) were determined by incubation at 37°C for 12-18 h with 3 l of [ 14 C]chloramphenicol (50 -60 mCi/mmol, NEN Life Science Products), 5 l of 5 mg/ml acetyl coenzyme A, and 0.25 M Tris-HCl, pH 8.0, in a volume of 150 l. The mixtures were phase extracted and subjected to thin layer chromatography. Results were analyzed by autoradiography using Bio-Rad Phosphor Imaging and Molecular Imager Systems, version 1.4.
Reagents and Materials-Water-soluble 17␤-estradiol, 17␣-estradiol, tamoxifen, retinoic acid, and 12-O-tetradecanoylphorbol-13-acetate (TPA), were purchased from Sigma. Retinoic acid and TPA were dissolved as stock solutions in dimethyl sulfoxide. 1,25(OH) 2 D 3 was generously provided by Dr. J. Napoli (SUNY, Buffalo). The concentration and purity of the 1,25(OH) 2 D 3 were assessed by UV spectroscopy with an absorbance maximum at 265 nm (⑀ 265 ϭ 18,200), minimum at 229 nm, and a maximum/minimum ratio of 1.80. ICI 182,780 was generously provided by Dr. A. E. Wakeling of Zeneca Pharmaceuticals, Macclesfield, United Kingdom. Hormones and reagents were added from 2,000-fold stock concentrations. Equivalent volumes of vehicle had no effect on cell viability or CAT activity. Restriction enzymes and other DNA modification reagents were from Promega Corp. Oligonucleotides were obtained from Bio-Synthesis, Inc. (Lewisville, TX).

RESULTS
Cloning and Sequence Analysis of the 5Ј-Region of the Human Type I InsP 3 Receptor Gene-To study the regulatory elements governing the tissue-specific expression of the type I InsP 3 receptor, a human placental genomic DNA library in a FIXII host was screened for clones containing the 5Ј-untranslated and promoter regions of the type I InsP 3 receptor gene. Using a series of PCR-derived probes corresponding to the 5Ј region of the human type I InsP 3 receptor cDNA, five independ-ent clones were isolated and characterized by restriction enzyme mapping, Southern blotting, DNA subcloning, and DNA sequencing (Fig. 1). Three of these clones ( 1-5-1 , 3-1-2 , and 5-2-1 ) were selected by their hybridization to a human probe corresponding to bp 489 -970 of the type I InsP 3 receptor cDNA (numbering is in reference to the homologous mouse sequence) and were shown to overlap as indicated. Three separate exons were found by restriction enzyme digestion and Southern hybridization of these overlapping clones. A fourth independent clone ( 124 ) encoding two exons was isolated using a human cDNA probe derived from bp 85-513 of the cDNA sequence (mouse numbering). The more 5Ј of these exons, denoted exon 3, is at least the third exon in the gene and begins at base 241 of the human cDNA sequence (11). The ATG initiation methionine codon at nucleotides 257-259 is included within this exon. The clone 5-1-7 was isolated using a human 174-bp probe derived from bp 85-187 (mouse cDNA numbering) of the 5Јuntranslated region of the type I InsP 3 receptor cDNA. Subclones of this 5-1-7 DNA, including a 0.5-kb SacI fragment and a 3.2-kb XbaI/HindIII fragment were isolated by their hybridization to the same 174-bp probe. The 1.5-kb BglII internal fragment of the 3.2-kb XbaI/HindIII subclone was sequenced completely (Fig. 2). This BglII fragment contains ϳ1.2 kb of upstream nontranscribed DNA including a consensus TATA box element, the first exon, and part of the first intron of the human type I InsP 3 receptor gene.
Identification of the Transcription Start Site of the Human Type I InsP 3 Receptor Gene-The transcription start site was determined by primer extension and rapid amplification of cDNA ends (5Ј-RACE). The antisense primer, p109a, corresponding to bases ϩ109 through ϩ85 of the mouse type I InsP 3 receptor cDNA (7) (ϩ37 through ϩ13 of the human type I InsP3R cDNA; Ref. 11) was hybridized to and extended by reverse transcription of RNA from human G-292 osteosarcoma cells and human HL-60 promyelocytic cells. The primer extension (Fig. 3) identified an adenine (A) residue as the predominant transcription start site. This adenine is located 39 residues downstream of a consensus TATA box motif (TATATAA) and is referenced in the sequence as position ϩ1 (Fig. 2). Results from three separate experiments using independently isolated RNAs from G-292 and HL-60 cells indicated the same major transcription start site in both cell types. A minor start site (C), also identified in both cell types, was identified at position Ϫ26 relative to the major start site and 14 bp down-stream of the TATA box. Interestingly, the 5Ј-RACE analysis performed at 37°C consistently yielded products shorter (indicated by asterisks in Fig. 2) than expected from the published human and rat cDNAs or from the observed primer extension products. Inspection of the DNA sequence immediately upstream and encompassing the 5Ј-RACE terminated sites indicates a conserved GGA-rich region, which may impart an unusual secondary structure on the DNA and prevent full-length extension. Successful primer extension through this region required temperatures of 46 -48°C; extensions at 37°C were terminated prematurely. CAT activity averaged 4-fold greater (n ϭ 3) than activity in cells transfected with constructs in the reverse orientation (Fig.  4). In G-292 cells transfected with the forward promoter construct, CAT activity was further increased by treatment with 1 nM phorbol ester (TPA) but decreased in cells treated with either 10 nM 17␤-estradiol or 10 nM 1,25(OH) 2 D 3 . Both 17␤estradiol and 1,25(OH) 2 D 3 decreased CAT expression when added together with TPA when compared with TPA alone.
Transcription promoting activity of the 5Ј-flanking region of the type I InsP 3 receptor gene promoter was also determined by transient transfection of G-292 cells with pCAT-Basic constructs containing the 1.5-kb BglII fragment (Ϫ1152 through ϩ350 bp) or 5Ј-deletions of this fragment. These data (Fig. 5) have been normalized and are reported as percent of activity relative to the Ϫ174 bp construct, which consistently exhibited the greatest CAT activity. The results summarized in Fig. 5 show that the longer promoter constructs (Ϫ1150 bp, Ϫ911 bp) had less CAT promoting activity that shorter constructs (Ϫ691 bp through Ϫ134 bp), and that even shorter constructs (Ϫ110 bp through Ϫ39 bp) had less CAT promoting activity than the Ϫ174 bp promoter construct. Significantly, CAT expression from deletion constructs including just the TATA box, as in the Ϫ39 construct, was greater than expression from the pCAT-Basic parental vector, whereas the slightly longer constructs (Ϫ54, Ϫ134, and Ϫ174 bp) that included a consensus NF-B binding element and a serum response element had greater transcription promoting activity than constructs containing only the TATA box (p Ͻ 0.05, paired Student's t test). 3 Receptor Promoter in G-292 Cells-We have reported that treatment of human G-292 osteosarcoma cells and rat primary calvarial osteoblasts with estrogen decreases steady state expression of type I InsP 3 receptor mRNA (24). That data were consistent with mechanism involving reduced gene transcription. Using the human type I InsP 3 receptor promoter CAT reporter constructs, the regions of the promoter that might confer transcription repression by 17␤-estradiol were investigated. Following transient transfection of G-292 cells with various deletion constructs of the type I InsP 3 receptor gene promoter-pCAT plasmid, cells were incubated with or without 10 nM 17␤-estradiol for 48 h and then assayed for CAT activity. Fig. 6 shows that cells transfected with the longer promoter constructs (Ϫ2.9 kb and Ϫ1152 bp) expressed relatively low CAT activity and 17␤-estradiol had little affect on this expression. However, cells transfected with the shorter promoter constructs (Ϫ486 bp and Ϫ174 bp) expressed greater CAT activity, and this activity was inhibited 20 -40% by 10 nM 17␤-estradiol (p Ͻ 0.05, paired Student's t test).

17␤-Estradiol Represses Expression of the Type I InsP
The Effect of 17␤-Estradiol is Inhibited by Tamoxifen and ICI 182,780 -To determine whether the inhibitory effect of 17␤estradiol on type I InsP 3 receptor promoter activity in G-292 cells might be mediated by the estrogen receptor, CAT expression from the type I InsP 3 receptor promoter construct was determined in cells treated with 17␤-estradiol plus the antiestrogen tamoxifen, the estrogen receptor antagonist ICI 182,780, or 17␣-estradiol, the inactive isomer of 17␤-estradiol. The data in Fig. 7A show that 10 nM 17␤-estradiol reduced CAT expression in cells transfected with the Ϫ357 bp type I InsP 3 receptor gene promoter and that 100 nM tamoxifen did not FIG. 3. Primer extension analysis on the 5 end of the human type I InsP 3 receptor gene. Twenty g of total RNA from human G-292 osteosarcoma cells and human HL-60 promyelocytes were subjected to primer extension using the p109a antisense oligonucleotide corresponding to bases ϩ109 through ϩ85 of the published mouse type I InsP 3 receptor cDNA (ϩ118 through ϩ95 of the genomic DNA). The location of the major extended product is shown by the arrow. The corresponding DNA sequence was determined using the p109a primer and the pGEM11/SacI-0.5-kb subclone identified in Fig. 1. The asterisk indicates the adenine identified as the major start site. The putative TATA box sequence is also shown. mimic this response. However, tamoxifen did block the effect of 17␤-estradiol. In this experiment, a slight inhibitory effect of 17␣-estradiol was seen but this was not reproduced in the other experiments. In a separate series of experiments, the pure estrogen receptor antagonist, ICI 182,780, was used. In these experiments (Fig. 7B), 10 nM 17␤-estradiol reduced CAT expression from the Ϫ357 bp type I InsP 3 receptor promoter construct to 87 Ϯ 2% (mean Ϯ S.E., p Ͻ 0.01) of control, whereas CAT expression in cells treated with 17␤-estradiol plus 20 nM ICI 182,780 was 122 Ϯ 9%. Although cells treated with 17␤-estradiol plus ICI 182,780 had elevated CAT activity, this was not significantly different from cells treated with ICI 182,780 alone (124 Ϯ 8%). In these series of experiments, 17␣-estradiol was again inactive as an inhibitor.

DISCUSSION
The effects of many bone resorptive agents are mediated through their actions on osteoblasts to stimulate the production and secretion of osteoclast-activating factors, including IL-6 (30, 31). Parathyroid hormone, parathyroid hormone-related peptide, bradykinin, prostaglandins, and endothelins induce bone resorption, require osteoblasts for this action, and stimulate IL-6 secretion from osteoblasts (32). These hormones all increase phosphoinositide turnover, InsP 3 formation, and calcium mobilization in osteoblasts, suggesting that the InsP 3 receptor may be important in osteoblast signal transduction and may be an essential component in the regulation of IL-6 secretion. On the other hand, estrogens have antiosteoporotic effects mediated in part by their effects on osteoblasts to reduce IL-6 secretion (33). The mechanism for this action of estrogen in osteoblasts has been attributed to the repression of IL-6 gene transcription, despite the lack of consensus estrogen response elements in the IL-6 promoter. Indeed, additional experimental evidence suggests that 17␤-estradiol affects estrogen receptordependent IL-6 gene repression in osteosarcoma cells by direct inhibitory interaction with the activator transcription factors NF-B and C/EBP␤ (34). The present study indicates that an additional mechanism may contribute to the reduced IL-6 secretory capacity of osteoblasts following estrogen treatment, viz. that chronic estrogen treatment reduces the expression of the type I InsP 3 receptor gene resulting in diminished osteoblast secretory activity.
In the present investigation, the 5Ј region of the human type I InsP 3 receptor gene including its putative promoter was cloned and its regulation in osteoblasts by estrogen and other agents is described. Three overlapping genomic clones and two additional neighboring clones of the human type I InsP 3 receptor gene were isolated by hybridization with cDNA probes. Approximately 50 kb of genomic DNA was isolated and determined to encode eight of the promoter proximal exons within the first 10% (through bp 880) of the 10-kb human type I InsP 3 receptor cDNA. The relative large size of this portion of the gene suggests that the entire human type I InsP 3 receptor gene may be as large as 500 kb. An interesting feature of the promoter proximal region of the gene is that the exon containing the translation start site is preceded by at least two noncoding exons. Examination of the exon-intron boundaries, as illustrated for the first such boundary at bp ϩ246 included in the sequence reported in Fig. 2, showed them all to be classical AGgt . . . agG sequences. Another interesting feature of the first exon is the repetitive GGAGG(A/T) sequence from ϩ45 through ϩ94. This sequence is found with high homology in several other signal transduction genes including those for c-Myb, c-ErbB2, epidermal growth factor receptor P1, IGF II, and snRNA U1 (NCBI). An unusual RNA secondary structure may be imparted by this sequence since the 5Ј-RACE analyses performed at 37°C were arrested within this sequence, and the primer extension required temperatures between 46 -48°C to traverse through this region.
The identification of the major transcription start site by primer extension locates the initiation site 40 bp downstream from a TATATAA sequence, suggesting that the latter serves as a functional TATA box. Indeed, the transcription reporter constructs showed that inclusion of the DNA through and including the TATA box conferred significantly greater transcription promoting activity than the promoterless CAT parental vector. This TATATAA stretch is surrounded by GC-rich sequences, also typical of consensus TATA elements. With this orientation, examination of the upstream sequence reveals several consensus transcription regulatory sequences that potentially could control type I InsP 3 receptor gene expression in a cell-specific and hormonally regulated fashion. Included among these are recognition sequences for AP-1, serum response element (SRE)-binding factors, NF-B, C/EBP, and nuclear receptors (see Fig. 2).
Paired AP-1 sequences are identified at Ϫ334 and Ϫ261 in the upstream region. AP-1 sites can act as positive regulatory elements for phorbol ester-induced transcription through activation of c-Jun/c-Fos signaling (35). In G-292 cells transfected with the Ϫ357 bp type I InsP 3 receptor promoter pCAT construct, TPA stimulated 3 times more CAT expression compared with untreated cells. An SRE-like sequence is found at Ϫ159 and maximal CAT activity was stimulated by 2% charcoalstripped serum with the Ϫ174 construct (Fig. 5). However, the Ϫ134 construct, which lacks the consensus SRE, still exhibited CAT activity comparable to the Ϫ174 construct. Consensus regulatory sequences for NF-B at positions Ϫ274 and Ϫ52, as well as for C/EBP at Ϫ382, are also identified, although a role for these elements in regulating transcription in osteoblasts is not immediately apparent.
Paired steroid receptor half-sites separated by a 4-base spacer are found at Ϫ746. Previous studies have demonstrated that retinoic acid induces InsP 3 receptor mRNA expression and increases the rate of type I InsP 3 receptor gene transcription in HL-60 cells (19). The nuclear receptor half-sites found at Ϫ746 (CGGTCAggacTGAGGT) have a 4-base spacer and are in an inverted repeat orientation (IR-4); however, the IR-4 configuration of the hormone response half-site is not an optimal arrangement for retinoids or other nuclear hormones (36). Interestingly, consensus vitamin D-responsive elements or estrogen-responsive elements were not identified, even though previous studies have shown that levels of type I InsP 3 receptor mRNA are positively regulated by vitamin D in HL-60 cells (20) and negatively regulated by estrogens and vitamin D in primary osteoblasts and osteosarcoma cell lines (24).
The effects of 17␤-estradiol on type I InsP 3 receptor promoter activity were investigated in greater detail. Our previous studies of human G-292 osteosarcoma cells and rat calvarial osteoblasts have indicated that estrogens decrease type I InsP 3 receptor mRNA levels without an effect on mRNA stability, suggesting an inhibitory effect on the rate of gene transcription (24). The data presented here are consistent with this contention. In G-292 cells transfected with InsP 3 receptor promoter constructs, exposure to 10 nM 17␤-estradiol inhibited CAT activity up to 40%. CAT repression by 17␤-estradiol was most prominent with constructs that also had the greatest basal activity in the 2% charcoal-stripped serum, i.e. the Ϫ326 and Ϫ174 constructs. Of additional interest is the fact that maximal promoter repression by 17␤-estradiol averaged 40%, whereas reduction of the type I InsP 3 receptor mRNA is typically about 75% in similarly treated cells (24).
The lack of effect of 17␣-estradiol and the observed antagonism of the 17␤-estradiol response by tamoxifen and ICI 182,780 strongly support the contention that the transcriptional repression observed with 17␤-estradiol is mediated by the estrogen receptor. However, the lack of consensus estrogen response elements in the type I InsP 3 receptor promoter sequence suggests a non-classical action of the estrogen receptor on gene expression (37,38). Stein and Yang (34) observed similar estrogen receptor-dependent inhibition of IL-6 promoter activity in human U2-OS osteoblastic cells and presented evidence that the inhibition is not mediated through consensus estrogen response elements in the promoter but instead involves estrogen receptor binding to and repression of the stimulatory transcription factors NF-B and C/EBP␤. Several observations reported here are consistent with a similar mechanism acting at the InsP 3 receptor promoter. First, consensus response elements for the estrogen receptor are not found in the isolated InsP 3 receptor promoter DNA. Second, transcription repression by 17␤-estradiol was observed with the proximal promoter constructs, the same constructs that show maximal stimulation with serum. Furthermore, consensus binding sites for NF-B, C/EBP␤, and serum response factor are found in the proximal promoter. Thus, the Ϫ350 bp and Ϫ174 bp promoter constructs had greatest activity in the presence of the 2% charcoal-stripped serum, and these were the constructs that showed the most significant repression with 17␤-estradiol. The same parallel was observed in U2-OS osteoblastic cells between the regions of the IL-6 promoter necessary for transcription enhancement by either TPA or IL-1␤ and the ability of 17␤-estradiol to repress transcription. In the case of the IL-6 promoter, experimental observations supported a mechanism by which the agonist-activated estrogen receptor was not required to be bound to the promoter to depress transcription, but rather the receptor sequestered and transrepressed the activator transcription factors, NF-B and C/EBP␤. Studies are directed toward determining whether a similar mechanism is operative at the type I InsP 3 receptor promoter in osteoblasts and whether this estrogen-dependent down-regulation of the type I InsP 3 receptor plays a role in the blunted the secretion of osteoclast activating cytokines from estrogen-treated osteoblasts.