Differential Regulation of Direct Repeat 3 Vitamin D3and Direct Repeat 4 Thyroid Hormone Signaling Pathways by the Human TR4 Orphan Receptor*

In situ hybridization analysis demonstrated that abundant testicular orphan receptor (TR4) transcripts were detected in kidney, intestine, and bone, which are vitamin D3 target organs. Cell transfection studies also demonstrated that the expression of the vitamin D3 target gene, 25-hydroxyvitamin D3 24-hydroxylase, can be repressed by TR4 through high affinity binding (K d = 1.32 nm) to the direct repeat 3 vitamin D3 receptor response element (DR3VDRE). This TR4-mediated repression of DR3VDRE is in contrast to our earlier report that TR4 could induce thyroid hormone target genes containing a direct repeat 4 thyroid hormone response element (DR4T3RE). Electrophoretic mobility shift assay using several TR4 monoclonal antibodies when combined with either TR4-DR3VDRE or TR4-DR4T3RE showed a distinct supershifted pattern, and proteolytic analysis further demonstrated distinct digested peptides with either TR4-DR3VDRE or TR4-DR4T3RE. These results may therefore suggest that TR4 can adapt to different conformations once bound to DR3VDRE or DR4T3RE. The consequence of these different conformations of TR4-DR3VDRE and TR4-DR4T3RE may allow each of them to recruit different coregulators. The differential repression of TR4-mediated DR3VDRE and DR4T3RE transactivation by the receptor interacting protein 140, a TR4 coregulator, further strengthens our hypothesis that the specificity of gene regulation by TR4 can be modulated by protein-DNA and protein-protein interactions.

In situ hybridization analysis demonstrated that abundant testicular orphan receptor (TR4) transcripts were detected in kidney, intestine, and bone, which are vitamin D 3 target organs. Cell transfection studies also demonstrated that the expression of the vitamin D 3 target gene, 25-hydroxyvitamin D 3 24-hydroxylase, can be repressed by TR4 through high affinity binding (K d ‫؍‬ 1.32 nM) to the direct repeat 3 vitamin D 3 receptor response element (DR3VDRE). This TR4-mediated repression of DR3VDRE is in contrast to our earlier report that TR4 could induce thyroid hormone target genes containing a direct repeat 4 thyroid hormone response element (DR4T 3 RE). Electrophoretic mobility shift assay using several TR4 monoclonal antibodies when combined with either TR4-DR3VDRE or TR4-DR4T 3 RE showed a distinct supershifted pattern, and proteolytic analysis further demonstrated distinct digested peptides with either TR4-DR3VDRE or TR4-DR4T 3 RE. These results may therefore suggest that TR4 can adapt to different conformations once bound to DR3VDRE or DR4T 3 RE. The consequence of these different conformations of TR4-DR3VDRE and TR4-DR4T 3 RE may allow each of them to recruit different coregulators. The differential repression of TR4-mediated DR3VDRE and DR4T 3 RE transactivation by the receptor interacting protein 140, a TR4 coregulator, further strengthens our hypothesis that the specificity of gene regulation by TR4 can be modulated by protein-DNA and protein-protein interactions.
Steroid hormones are key physiological mediators of development and homeostasis (1)(2)(3)(4). Understanding the crosstalk between steroid hormone-dependent and -independent signaling pathways is critical for gaining further insight into the integration of cellular regulatory cues that modulate development and tissue-specific gene expression. The biological effects of steroids and related hormones, including derivatives of vitamins A and D 3 , are mediated through their cognate receptors (1)(2)(3)(4). These receptors are members of a large group of ligandactivated proteins that act as transcriptional activators or repressors. However, there is another group of nuclear receptors that shares the same molecular structure as steroid hormone receptors but has no known ligands. The members of this group have therefore been named orphan receptors (5). A common characteristic of many of these orphan receptors is that they exert at least part of their function as regulators of other receptors. This may occur by several different mechanisms: competition for the same response element, heterodimer formation with the regulated receptor, or heterodimer formation with the retinoid X receptor (RXR), 1 thus titrating out available RXR protein. For instance, chicken ovalbumin upstream promoter transcription factors have been demonstrated to suppress ligand-induced gene activation, including that of vitamin A, vitamin D 3 , and thyroid hormone target genes (6). This interfering effect might involve the formation of a transcriptional silencing complex with RXR or competition with steroid hormone receptors for DNA binding sites (7).
The human TR4 orphan receptor (TR4) was first identified by using degenerate polymerase chain reaction cloning. The open reading frames of TR4 cDNA encode 615 amino acids with a calculated molecular mass of 67 kDa (8). On the basis of sequence similarities, TR4 is classified as a member of the steroid hormone receptor superfamily, very close to the TR2 orphan receptor (9,10). A comparison of the amino acid sequence in the p-box of the DNA binding domain groups TR4 in the estrogen receptor/thyroid hormone receptor (T 3 R) subfamily, which recognizes the AGGTCA core consensus motif. From this information we were able to identify several target genes that are up-regulated by TR4, including the fifth intron of the ciliary neurotrophic factor receptor ␣ (11,12) and DR4T 3 RE (13). In contrast, TR4 represses the SV40 major late promoter (14) and retinoic acid receptor/RXR target genes (15). Similar results were also obtained with TR2 (16,17). Molecular mechanisms of the differential regulation by TR4 remain unclear. Determining whether TR4 is capable of interfering with gene regulation by binding to other AGGTCA-like motifs and/or interacting with cofactor(s) might provide us more detailed information about this orphan receptor. A vitamin D 3 target gene is a potential candidate of interest for more study, because its receptor functions through interaction with the vitamin D 3 response element, which contains two AGGTCA repeat motifs with a 3-nucleotide space between repeats (DR3-VDRE). These response element motifs may also be recognized by TR4. Despite the similarity of hormone response elements (HRE) recognized by both VDR and TR4, these two receptors have also been found to interact in vitro with the receptor interacting protein 140 (RIP140) cofactor (18,19). Investigating the regu-lation of TR4 in the vitamin D signaling pathway provides us with a molecular mechanism to explain the way TR4 regulates the gene expression in terms of protein-DNA and protein-protein interactions.
Vitamin D 3 is hydroxylated first in the liver at carbon 25 to yield 25-hydroxyvitamin D 3 and then in the kidney at the ␣-position of carbon 1 to generate 1,25-(OH) 2 D 3 , the active form of vitamin D. This has various biological effects, including the maintenance of calcium homeostasis, regulation of bone remodeling, and modulation of cell growth and differentiation (20 -23). The 25-hydroxyvitamin D 3 24-hydroxylase gene (P450cc24) encodes a key enzyme involved in vitamin D metabolism, which is responsible for the conversion of 25-(OH)D 3 and 1,25-(OH) 2 D 3 to 24,25-dihydroxyvitamin D 3 (24,25-(OH) 2 D 3 ) and 1,24,25-trihydroxyvitamin D 3 , respectively (24). These metabolites are thought to be inactive forms of vitamin D (25). However, 1,25-(OH) 2 D 3 induces 24-hydroxylase activity in its target cells, and thus its presence may play a crucial role in eliminating hormone activity of the vitamin D compound (25). Two vitamin D-responsive elements (VDRE-1 and VDRE-2) responsible for 1,25-(OH) 2 D 3 stimulation of transcription were identified at nucleotides Ϫ151 to Ϫ137 and nucleotides Ϫ259 to Ϫ245 of the 5Ј-flanking region of the rat P450cc24 gene (26 -28). Both VDREs contain two AGGTCA-like repeat motifs with a 3-nucleotide space in sense or antisense orientation and are similar to the VDREs found in the human P450cc24 gene (29). Examination of regulation via the 24-hydroxylase induction mechanism at the molecular level may contribute to the understanding of vitamin D in the endocrine system.
In this study, we analyzed the regulation by TR4 on the vitamin D 3 signaling pathway and compared the expression pattern of TR4 to that in vitamin D 3 target organs. The differential regulation of target genes containing DR3VDRE and DR4T 3 RE by TR4 was further investigated, and the results suggested that conformational changes because of DNA-protein and protein-protein interactions might play major roles in this regulation.

MATERIALS AND METHODS
Plasmid Construction-For the transient transfection or coupled in vitro transcription/translation of the full-length TR4 protein, the pCMX-TR4 and pET14b-TR4 plasmids were constructed as described previously (14). The chimeric receptor pCMX4A4 was constructed as described previously (13). The reporter plasmid P450cc24-CAT, the 5Ј-flanking region (nucleotide Ϫ2200 to ϩ188) of the rat vitamin D 3 24-hydroxylase gene, was kindly provided by Dr. Y. Kato (26). The glutathione S-transferase (GST) fusion protein was constructed by inserting the full-length TR4 to the SmaI site of pGEX-3X (Amersham Pharmacia Biotech). Full-length RIP140 was cloned into the pET-28a(ϩ) vector at the NotI site for in vitro transcription/translation reaction. The coupled in vitro transcription and translation expression plasmids, pET14b-TR4, pCMX-RXR␣, and pSG5-VDR (29), were in vitro transcribed and translated by the TNT system (Promega) as described previously (14).

RESULTS
TR4 Specifically Binds to DR3VDRE, the AGGTCA Motif with 3-nucleotide Spacing-EMSA was used to determine the binding specificity of TR4 to DR3VDRE. In vitro translated TR4 protein was incubated with 32 P-labeled DR3VDRE and analyzed on a 5% polyacrylamide gel. As shown in Fig. 1, free probe was detected at the bottom of the gel (lane 1), and a specific DNA⅐protein complex was identified when 1 l of in vitro translated TR4 protein was added (indicated by an arrow, lane 2). This complex was competed out by adding 100-fold excess of unlabeled DR3VDRE (lane 3). This DNA⅐protein complex was abolished when anti-TR4 monoclonal antibody 4 was added (lane 4). In contrast, when the anti-TR4 monoclonal antibody 2 was added to the reaction, a DNA⅐receptor⅐antibody complex supershifted band was visible (indicated by an arrowhead, lane 5).
DR3VDRE Binds to TR4 with Higher Affinity than to the RXR/VDR Heterodimer-Vitamin D 3 decreases the affinity of the VDR/VDR homodimer for DNA targets, enhances the formation of RXR/VDR heterodimers, and potentiates RXR/VDR affinity for VDREs (33,34). Comparing the different affinities of these receptors and TR4 for DR3VDRE might provide more information about how these receptors regulate their target genes. EMSA was performed to determine receptor response element K d . Fixed amounts of in vitro translated TR4 protein (1 l) or RXR/VDR (1 l of each) were incubated with an increasing amount of 32 P-DR3VDRE (from 0.1 to 12.8 ng) and analyzed by EMSA. As shown in Fig. 2, Scatchard plot analysis of the DNA⅐protein complexes in the EMSA demonstrated that the K d values of TR4 for DR3VDRE and RXR/VDR for DR3VDRE were 1.32 nM and 7.31 nM, respectively. Therefore TR4 has an affinity for DR3VDRE that is 5.5-fold higher than that of the RXR/VDR heterodimer. These data suggest that TR4 was able to compete with RXR/VDR for the binding to DR3VDRE.
TR4 Represses the Vitamin D 3 -induced Rat P450cc24 Gene Expression-The consequence of high affinity binding between TR4 and DR3VDRE was then investigated by reporter assay. The target gene used here was the 5Ј-flanking region (nucleotide Ϫ2200 to ϩ188) of the rat P450cc24, containing VDREs that are responsible for the 1,25-dihydroxyvitamin D 3 enhancement and are located at nucleotides Ϫ167 to Ϫ102 and nucleotides Ϫ204 to Ϫ129 (22). Co-transfection of 2.5 g of pSG5VDR with 3 g of P450cc24-CAT into CHO cells enhanced the transactivation of P450cc24-CAT to 32-fold in the presence of 10 Ϫ7 M 1,25-(OH) 2 D 3 (Fig. 3B, lane 2 versus 3). However, there is no CAT activity with the co-transfection of 2.5 g of pCMX-TR4 or pCMX-4A4 either in the absence or presence of 1,25-(OH) 2 D 3 (Fig. 3B, lanes 4 -7). The chimeric receptor pCMX-4A4, which replaces the DNA binding domain of TR4 with that of an androgen receptor (Fig. 3A), was unable to bind to the AGGTCA DR motif sequence and served here as a negative control. The transcriptional activity of P450cc24-CAT induced by vitamin D 3 was repressed when co-transfected with pCMX-TR4, but not when co-transfected with the chimeric receptor, pCMX-4A4 (Fig. 3B, lane 3 versus 8 and 9). These results suggest that the DNA binding domain of TR4 is essential for the TR4-mediated repression of the vitamin D 3 responsive enhancement of the P450cc24 gene. To examine the expression levels of both wild-type and chimeric receptor TR4 proteins after transfection, polyclonal antibodies against both proteins were produced and examined. This repression effect was further proven by co-transfection of pSG5VDR with different amounts of pCMX-TR4 (from 1 to 5 g) into CHO cells. As shown in Fig. 4, the repression effect mediated by TR4 was gradually increased when an increasing amount of TR4 was co-transfected (lanes 4 -8). This result clearly demonstrated that TR4 could suppress the vitamin D 3 -induced P450cc24 gene promoter activity in a TR4 dose-dependent manner.
Localization of TR4 Transcripts in Vitamin D 3 Target Organs during Mouse Embryogenesis-To determine if TR4 is expressed in vitamin D target tissues, we applied in situ hybridization analysis to mouse embryos. As shown in Fig. 5, high levels of TR4 transcripts were detected in the perichondrium, which contains cells active in bone formation (Fig. 5B). Distinct TR4 distribution was also observed in the developing glomeruli and tubule structures of the kidney as well as the intestinal villi (Fig. 5, C and D). Fig. 5A shows strong TR4 expression detected in certain nonclassical vitamin D target tissues, such as lung and hair follicles (35,36). Active TR4 expression in these tissues with the known expression domains of VDR (37) and 24-hydroxylase (38,39) suggests TR4 may interact with the in vivo vitamin D signaling pathway.
The Conformational Differences between TR4⅐DR3 Complex and TR4⅐DR4 Complex-The above results suggest that TR4 may exert a repression effect on vitamin D 3 responsive target gene expression by binding to DR3VDRE. In our previous studies, we concluded that TR4 activates the expression of the genes that contain DR4T 3 RE in both HRE sequence-and TR4 dose-dependent manners (13). This contrasting and differential regulation by TR4 could be because of different DNA-protein or protein-protein interactions. When compared, we found the K d values of TR4 to DR3VDRE and TR4 to DR4T 3 RE to be very similar (1.32 versus 2.0 nM). This result eliminates the possibility that a different binding capacity between TR4 and DR3 or DR4 results in the distinctive regulation. One microliter of in vitro translated TR4, VDR, and RXR was incubated with serially diluted 32 P-labeled DR3VDRE (concentrations from 0.1 to 12.8 ng) and resolved by EMSA. After autoradiography, the respective bands were excised, placed in scintillation fluid, and quantified directly in a scintillation counter. The ratio of activity between specific DNA protein binding (Bound) and free DNA probe with respect to specific DNA protein binding (bound/free (B/F)) was plotted. The dissociation constant (K d ) and B max values were generated using the EBDA program (Biosoft).
The second possibility is that the mechanism of regulation may be through different protein-protein interactions that are dependent on the distinct conformations of TR4 once bound to either DR3VDRE or DR4T 3 RE. To test this second possibility, EMSA was used to examine a series of monoclonal antibodies raised against TR4, which are able to recognize conformational epitopes. Four monoclonal antibodies (mAbs) were initially characterized by EMSA on the basis of their ability to supershift or abolish the TR4⅐DNA complex. As shown in Fig. 6A, we observed a specific DR3⅐TR4 complex (lane 2) that was distinct from mock-translated protein (lane 1); with the addition of mAbs 1, 2, and 3, supershifted bands were found (lanes 3-5). In contrast, the specific band formed by DR3⅐TR4 was abolished when the mAb 4 was added (lane 6). The same band-shifted pattern was observed when DR3 was replaced with DR4 (Fig.  6B, lane 1-6). Interestingly, different band-shifted patterns were observed with various combinations of different antibodies. As shown in Fig. 6A, an enhanced supershifted band shows migration positions that are different from the supershifted band (lane 8). This suggests that mAbs 1 and 3 recognize different epitopes because their simultaneous addition resulted in an increased mobility shift beyond that of either antibody alone. In contrast, the addition of mAbs 1 and 2 did not lead to an enhanced supershifted band (lanes 3 and 4 versus 7). These results suggest that the mAb 1 recognizes the same epitope of TR4 as the mAb 2 when TR4 is bound to DR3. However, a supershifted band was detected when the mAb 1 was added with the mAb 3 simultaneously to TR4⅐DR4 complex (Fig. 6B,  lane 8). These results indicate that TR4 might fold into different conformations upon binding to diverse HREs and that different antibodies can recognize these conformations.
This second hypothesis was further proven by proteolytic analysis. [ 35 S]methionine-incorporated, in vitro translated TR4 was incubated in the absence or presence of HRE (DR4 or DR3) at 25°C for 1 h. As shown in Fig. 4C, TR4-DR4T 3 RE has a similar trypsin-resisting fragment pattern to that of unbound TR4 control. In contrast, the TR4⅐DR3VDRE complex was more sensitive to trypsin when 2.5 g/ml trypsin was applied. As results showed, the full-length of TR4 was completely degraded and some trypsin-resisting fragments disappeared compared to that with TR4⅐DR4T 3 RE at the same concentration of protease treatment. Similar results were also obtained when we replaced trypsin with ␣-chymotrypsin. We concluded that TR4bound DR4T 3 RE or DR3V3RE had a different sensitivity to protease digestion and different protease-resisting fragments could be obtained with a higher concentration of protease digestion. These results further confirm our hypothesis that, to exert its proper function, TR4 may adopt distinct conformations when bound to DR3VDRE or DR4T 3 RE, leading to different protein environments.
RIP140 Interacts with TR4 and Differentially Modulates the TR4-mediated DR3VDRE-CAT and DR4T 3 RE-CAT Activities-In vitro interaction of TR4 and RIP140 was performed by GST pull-down assay. As shown in Fig. 7A, TR4 could interact with RIP140 but not RXR (lane 3 versus 4). No interaction could be detected when GST-TR4 was replaced with GST (lane 5 and 6). The effects of RIP140 on TR4-mediated DR3VDRE-CAT and DR4T 3 TE-CAT activities were also investigated. As shown in Fig. 7B, RIP140 can further repress the TR4-mediated DR3VDRE-CAT suppression (lane 8 versus 9). In contrast, RIP140 can repress both TR4 and T 3 R-mediated DR4-T 3 RE-CAT activities (Fig. 7C, lane 5 versus 6 and lane 7 versus 10) significantly, but RIP140 has no significant effect on the DR4T 3 RE-CAT induction when TR4 and liganded T 3 R were co-transfected (Fig. 7C, lane 8 versus 9). It is also worth noting that whereas RIP140 can further enhance the TR4-mediated DR3VDRE-CAT suppression (lanes 8 versus 9), RIP140 has no significant effect on DR3VDRE-CAT transactivation in the absence of TR4 (Fig. 7B, lane 7 versus 10). DISCUSSION According to the sequence of the TR4 DNA binding domain and our previous results from EMSA, we conclude that TR4 binds to the AGGTCA core consensus motif arranged in a direct repeat orientation with various numbers of nucleotide spacings. In this paper, we demonstrated that TR4 binds to DR3VDRE with high affinity and thus suppresses vitamin CHO cells were transfected with P450cc24-CAT reporter plasmid (3 g of each) and pSG5VDR (2.5 g of each) and co-transfected with increasing amounts of pCMX-TR4 (from 0 to 5 g). The percentage of CAT conversion was plotted. All CAT assays were normalized relative to ␤-galactosidase activity. Significant differences (*, p Ͻ 0.005) of CAT activity reduced by TR4 at different doses were determined by one-way analysis of variance. Each value represents the mean Ϯ S.D. of three independent experiments. D 3 -induced P450cc24 gene activation. However, in our previous studies, we demonstrated that TR4 activates the genes that contain DR4-T 3 RE or nonclassical T 3 RE by binding to their response elements (13). The mechanisms of different binding responses of TR4 to the same sequence in core motifs with different spacings remain unclear. The responsiveness of genes to steroid hormones involves both the binding of regulatory proteins to specific DNA sequences and the formation of critical protein-protein associations. Throughout the past decade, a number of nuclear receptor coregulators has been characterized and provides us a more detailed molecular model of how nuclear receptors regulate their target genes. Because TR4 binds to DR3VDRE and DR4T 3 RE with similar affinities (1.32 versus 2.0 nM), we propose that whether TR4 works as a transcriptional enhancer or a silencer might be mediated not only by direct DNA binding but also by protein-protein interactions. This indicates that in addition to the receptor and the DNA, other factors may also contribute to the selectivity of receptors in the recognition of their target genes.
To test this hypothesis, EMSAs and proteolytic analyses were performed. In Fig. 4, A and B, distinct EMSA patterns were observed when TR4 was bound 32 P-DR3VDRE and 32 P-DR4T 3 RE in the presence of different combinations of various TR4 monoclonal antibodies. Meanwhile, different peptide patterns were obtained when DR4-or DR3-bound TR4 was digested with trypsin. Taken together, these data suggest that TR4 binding to diverse HREs may result in distinct conformational changes that can then trigger differential regulation.
This finding supports the existence of a unique spacing of the direct repeat, which serves as a binding site for an auxiliary protein that modifies receptor activity (39). Similar approaches have also been used to study a thyroid receptor response element, RSVT 3 RE, which contains an inverted repeat with a 6-nucleotide space. RSVT 3 RE allows strong activation by c-ErbA␣ in the absence of thyroid hormone, and the results of antibody-induced supershift experiments indicate that binding to this element may result in a different conformation as compared with binding to a typical DR4T 3 RE (40). These results suggest that different conformational changes may be involved in determining whether TR4 would function as a positive or negative regulatory factor.
As previous studies indicated that RIP140 could interact with VDR, T 3 R, and TR4 (18, 41), we were interested in determining if RIP140 could differentially affect TR4-mediated gene induction and repression. Using the GST pull-down assay, we demonstrated that TR4 could interact in vitro with RIP140. We then proved that RIP140 could also enhance the trans-repressive effect of TR4 on the vitamin D 3 -signaling pathway. In contrast, RIP140 has only marginal effects on the vitamin D 3 -induced VDRE-CAT activity, although it might interact with VDR (18). On the other hand, RIP140 repressed the transactivation mediated by both TR4 (Fig. 7B, lane 5 versus 6) and liganded T 3 R (Fig. 7B, lane 7   RIP140 on DR4T 3 RE-CAT is that TR4 and liganded T 3 R may both interact with RIP140 and sequester RIP140 blocking the repressive effect of RIP140 on DR4-T 3 RE-CAT (Fig. 8). Because RIP140 can repress both pathways mediated by TR4, it might not be recruited differently by TR4 when it is bound to DR4T 3 RE or DR3VDRE. However, RIP140 did show differential effects in the regulation of DR4T 3 RE-CAT and DR3VDRE-CAT when receptors, TR4 and VDR versus TR4 and T 3 R, are co-transfected. These data support the role of the protein environment surrounding the DNA as a key factor in determining gene regulation. Therefore, the dynamic interaction between the receptors and cofactors in response to their target DNA may determine the specificity of gene regulation.
Although the functional reporter assay presented here was carried out in vitro, the expression patterns of TR4 in the vitamin D target organs support the potential role of TR4 in regulation of the in vivo vitamin D pathway in the endocrine system. Vitamin D and its receptor play essential roles in the regulation of calcium homeostasis and bone formation. Bone, kidney, and intestine are three major targets for such action. In addition, vitamin D is involved in the regulation of cell proliferation, differentiation, and the immune response. At present, the nonclassic targets reported include lung (35) and hair follicles (36). It has been demonstrated that high levels of VDR FIG. 7. RIP140 interacts with TR4 and exerts repressive control over DR3VDRE-CAT and DR4T 3 RE-CAT. 35 S-labeled RIP140 and RXR␣ were incubated with GST-TR4 or GST bound to glutathione-Sepharose beads in a pull-down assay. The input represents 20% of the amount of labeled protein used in the pull-down assay (A). CHO cells were co-transfected with 1 g of pCMX-TR4, 2 g of pSG5VDR (B) or pCD-T 3 R␣1 (C), 1 g PEF-RIP140, and 3.5 g DR3-VDRE-CAT or DR4-T 3 RE-CAT in different combination sets. Twenty-four hours after transfection, the cells were treated with 100 nM vitamin D 3 (B) or T 3 (C). All CAT assays were normalized relative to ␤-galactosidase activity. Each value represents the mean Ϯ S.D. of three independent experiments. transcripts are present in bone, kidney, and all identified nonclassic target organs. Throughout the entire process of bone differentiation, VDR transcripts were detected in osteoblasts and osteocytes of both normal and actively remodeling bone tissue (42). In the kidney, VDR and P450cc24 were co-localized in the distal renal tubules, the collecting ducts, the proximal tubes, and the parietal epithelial cells of the glomerulus (37,43). Our data show TR4 transcripts in perichondrium, kidney, hair follicles, and lung, suggesting TR4 and VDR may be colocalized during the development of these tissues. Among the vitamin D target genes, P450cc24 has been shown to respond to vitamin D treatment in the kidney and intestine. The distribution of this enzyme in the kidney is exclusive to the proximal tubules (38), whereas its location in the intestine was detected in the lower part of the villi and columnar epithelium of the crypt. Our in situ data showed TR4 expression in renal tubules and intestinal villi, indicating TR4 and P450cc24 may be coexpressed in these places, thus providing physiological significance for the interaction between TR4 and VDR on the regulation of the P450cc24 gene. Moreover, immunostaining for VDR revealed that VDR was apparent in rat fetuses as early as embryonic days 9 and 11. In developing limbs at embryonic days 13-15, VDR epitopes were present in the skin, the cytoplasm of condensing mesenchymal cells, and the chondrocytes (44). Whether TR4 could also interact with VDR in regulating other vitamin D target genes, such as osteocalcin and alkaline phosphatase, is an intriguing question to ask.
On the other hand, vitamin D 3 decreases VDR/VDR homodimer formation and enhances the formation of RXR/VDR heterodimers. Thus, it potentiates RXR/VDR action by enhancement of the RXR/VDR binding affinity to VDREs and activation of target gene expression (33,34). We then compared the binding affinity of TR4 and RXR/VDR to DR3VDRE and found that TR4 bound to DR3VDRE with a 5.5-fold higher affinity than the RXR/VDR heterodimer. These binding affinity data suggest that TR4 may be able to compete with RXR/VDR to bind to DR3VDRE and in this way, exert its repressive effect. The ratio of functional TR4⅐DR3VDRE complex to RXR/ VDR⅐DR3VDRE might also be important in determining target gene action. Additionally, the failure of the repression effect observed when PCMX-TR4 was replaced with the chimeric receptor PCMX-4A4 in a reporter gene assay further supports the idea that DNA binding is essential for TR4 to exert its proper function.
The responsiveness of genes to a steroid hormone receptor is principally mediated by functional interactions between DNAbound receptors and components of the transcription initiation machinery. In this paper we demonstrated that TR4 represses the P450cc24 gene activation induced by vitamin D 3 via in vitro and in vivo evidence. Both DNA binding and proper proteinprotein interactions may be the key factors in determining the specific function of TR4. This study may lead to an understanding of the role of DNA binding in altering the conformation of TR4 and allowing different protein interactions resulting in a complex that is capable of mediating differential regulation.