help button home button Endocrine Society Molecular Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Molecular Endocrinology, doi:10.1210/me.2003-0310
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
18/4/874    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow NURSA Molecule Pages Link
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Ismail, A.
Right arrow Articles by Peleg, S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Ismail, A.
Right arrow Articles by Peleg, S.
Molecular Endocrinology 18 (4): 874-887
Copyright © 2004 by The Endocrine Society

Effect of Cellular Environment on the Selective Activation of the Vitamin D Receptor by 1{alpha},25-Dihydroxyvitamin D3 and Its Analog 1{alpha}-Fluoro-16-Ene-20-Epi-23-Ene-26,27-Bishomo-25-Hydroxyvitamin D3 (Ro-26-9228)

Ayesha Ismail, Cuong V. Nguyen, Ago Ahene, James C. Fleet, Milan R. Uskokovic and Sara Peleg

Department of Endocrine Neoplasia and Hormonal Disorders (A.I., C.V.N., S.P.), The University of Texas M. D. Anderson Cancer Center, Houston, Texas 77030; Department of Muscoloskeletal Research (A.A., M.R.U.), Roche Bioscience, Palo Alto, California 94305; and Department of Foods and Nutrition (J.C.F.), Purdue University, West Lafayette, Indiana 47907-1264

Address all correspondence and requests for reprints to: Sara Peleg, Ph.D., Department of Endocrine Neoplasia and Hormonal Disorders, Unit 435, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas 77030. E-mail: speleg{at}mail.mdanderson.org.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIAL AND METHODS
 REFERENCES
 
The vitamin D analog, 1{alpha}-fluoro-16-ene-20-epi-23-ene-26,27-bishomo-25-hydroxyvitamin D3 (Ro-26-9228) is tissue selective, with a gene regulation preference for bone over duodenum in vivo. In the human osteoblast-like cells, hFOB, the vitamin D receptor (VDR)-mediated transcriptional potencies of Ro-26-9228 and 1,25-dihydroxyvitamin D3 (1,25D3) were similar, but in the intestinal cells, Caco-2, transcriptional potency of Ro-26-9228 was 10–50 times lower. We hypothesized that transcriptional activation of the VDR by Ro-26-9228 in the two cell types is regulated differently, and compared VDR extracted from hFOB or Caco-2 cells for their abilities to interact with a p160 coactivator [glucocorticoid receptor-interacting protein (GRIP)] and with retinoid X receptor (RXR) by pull-down assays. 1,25D3 had similar potencies to induce interactions of VDR from the two cell types with these partners of transcription. In contrast, Ro-26-9228 induced interaction of osteoblastic VDR with RXR and GRIP but did not induce these interactions with VDR from Caco-2 cells. Further studies revealed that in hFOB cells the unoccupied VDR was cytoplasmic and proteasome sensitive, and that ligand treatment caused a rapid accumulation of the VDR in the chromatin. Both cytoplasmic and chromatin-associated ligand-bound VDR from hFOB cells had the abilities to interact with GRIP. In contrast, in Caco-2 cells, unoccupied VDR was localized in both the cytoplasm (70%) and the chromatin (30%). In Caco-2 cells, the cytoplasmic VDR was proteasome resistant, and neither 1,25D3 nor Ro-26-9228 induced its binding to GRIP. Only a small fraction of the chromatin-associated VDR was proteasome sensitive, and this fraction was distinguishable by a faster electrophoretic mobility. 1,25D3 induced an accumulation of the proteasome-sensitive VDR in the chromatin of Caco-2 cells and binding to GRIP. Ro-26-9228 failed to induce accumulation of the proteasome-sensitive VDR in the chromatin or binding to GRIP, but a coincubation of Caco-2 cells with the analog and a proteasome inhibitor restored these abilities. These results suggest that Ro-26-9228 has poor ability to promote the accumulation of a proteasome-sensitive, transcriptionally active VDR isoform in Caco-2 cells, whereas it does not have this limitation in hFOB cells.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIAL AND METHODS
 REFERENCES
 
SOME SYNTHETIC LIGANDS for nuclear receptors act in a tissue- or gene-selective fashion. This is true of the selective estrogen receptor modulators (SERMs), tamoxifen, and raloxifene, which antagonize the estrogen receptor-mediated actions of the natural steroid hormone in the breast and in the uterus but promote moderate agonist actions in the bone (1, 2). The structural bases for agonist and antagonist actions have been elucidated from x-ray crystallography studies that revealed distinct conformational differences in agonist-bound and antagonist-bound receptors for several steroid hormones including estrogen receptor (3, 4, 5). However, these studies did not provide a structural explanation for the ability of SERMs to act as agonists in certain cellular environments. The contribution of the cellular environment to the selective action of these compounds has been explained by their cell-specific ability to use a domain of the receptor called activation function 1 to induce transcription through the receptor, even when the activation function 2 domain of the receptor is blocked by binding to an antagonist (6, 7). Another explanation is that the type of response element may contribute to selectivity (8, 9). For example, a SERM may exhibit a preference for induction of estrogen receptor-mediated transcription through an activator protein (AP)-1 like response element, whereas the natural hormone induces transcription through both the classic palindrome hormone responsive element and the nonclassic AP-1-like element (10, 11). Finally, more recent studies demonstrated that the ratio of corepressors and coactivators in target cells might determine whether an analog acts as an antagonist or as an agonist (2, 12).

Analogs of vitamin D have been developed to investigate the structure/function relationship with the vitamin D receptor (VDR) and with other vitamin D binding proteins (13, 14). Many of these analogs have been tested for their potential to be used as drugs for treatment of clinical conditions such as osteoporosis (15, 16, 17), secondary hyperthyroidism (18, 19), psoriasis (20, 21, 22), and cancer (23, 24, 25). A few of these ligands have also been identified as modulators of VDR action, including superagonists, antagonists, or even selective agonists (26, 27, 28, 29, 30). So far, a compound that acts as an agonist in one tissue and an antagonist in another such as the estrogen receptor modulator, tamoxifen, has not been identified. However, at least three low-calcemic analogs have shown selectivity. One is EB1089 (22,24-diene-26,27-bishomo-1,25-dihydroxyvitamin D3), which is selective for types of vitamin D response elements in vitro and in culture (31, 32). The other analog is OCT (22-oxa-1,25-dihydroxyvitamin D3), which recruits different coactivators to the VDR-analog complex than to the VDR-1,25D3 complex in vitro and appears to induce lower intestinal calcium absorption than 1,25D3 (33, 34, 35). The third analog, Ro-26-9228 (1{alpha}-fluoro-16-ene-20-epi-23-ene-26,27-bishomo-25-hydroxyvitamin D3), was recently shown by our laboratories to have tissue preference in vivo because it induced gene expression in bone but not in intestine (36). This preference could be reproduced in culture, in osteoblast and intestinal cell lines. In the study reported here, we further investigated the molecular and cellular mechanism for the selectivity of Ro-26-9228. Here we show that its apparent preference for osteoblasts over intestinal cells was based on differences in the ability of the VDRs from these two cell types to recruit coactivators of transcription to the receptor-analog complex. This difference is probably based on cell-specific mechanism for transformation of an inactive VDR to a transcriptionally active isoform, associated with chromatin.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIAL AND METHODS
 REFERENCES
 
Differential Induction of Transcriptional Activities by Ro-26-9228 in Intestinal and Osteoblast Cells
Our previous studies demonstrated that the potency (defined as the effective dose required to reach 50% of maximal gene expression, or ED50) of Ro-26-9228 to induce transcription in intestinal-like Caco-2 cells is nearly 50 times lower than its potency to induce transcription in the osteosarcoma cell line MG63, whereas 1,25D3 induced transgene expression in these two cell lines with similar potencies (36). We repeated these experiments using instead of MG63 the T-antigen transformed human fetal osteoblast cell line, hFOB, because hFOB cells more closely resemble normal osteoblasts, and can undergo a complete osteoblast differentiation program, concluding with formation of calcified matrix (37).

First we examined endogenous vitamin D receptor-dependent gene regulation (osteocalcin in hFOB cells and calbindin D9K and 24-hydroxylase in Caco-2 cells) in response to graded concentration of 1,25D3 and the analog. Our results showed (Fig. 1Go) that for inducing osteocalcin peptide secretion by hFOB cells 1,25D3 had an ED50 of 1.5 nM, similar to the potency of Ro-26-9228 to induce this gene (ED50 of 0.8 nM). In contrast, in Caco-2 cells, 1,25D3 induced significant amounts of calbindin D9K already at a concentration of 1 nM (Fig. 2Go, A and C), but Ro-26-9228 induced less than half the maximal amount of calbindin D9K mRNA at 100 nM, and this mRNA was not detectable at analog concentrations of 10 and 1 nM. The ED50 for induction of this mRNA by 1,25D3 was 8 nM and the ED50 for induction of this mRNA by Ro-26-9228 was 265 nM, thus the analog’s potency to induce calbindin D9K was 33 times lower than the potency of 1,25D3. When expression of 24-hydroxylase mRNA was examined, we found that the level of this mRNA in Caco-2 cells 8 h after treatment was significantly induced by 10 nM of 1,25D3, whereas Ro-26-9228 did not induce this gene at this concentration (data not shown). When a more detailed dose response was performed 24 h after ligand treatment, we found that the ED50 for induction of this mRNA by 1,25D3 was 2 nM and the ED50 for induction of 24-hydroxylase by Ro-26-9228 was 21 nM (Fig. 2Go, A and B). These results suggested that the analog is less potent than 1,25D3 in inducing gene regulation in Caco-2 cells, but the difference in potency may depend on the target gene as well. To further substantiate that the difference was not only due to the gene examined, we transfected the two cell types with the minimal osteocalcin vitamin D-responsive element (ocVDRE)-thymidine kinase-GH reporter gene and measured reporter gene expression. These assays showed that 1,25D3 and Ro-26-9228 had similar potencies in inducing reporter gene expression in hFOB cells (ED50s of 3 nM and 5.5 nM, respectively, Fig. 1BGo). In contrast, 1,25D3 (ED50 = 2 nM) was significantly more potent than Ro-26-9228 (ED50 = 88 nM) in inducing reporter gene expression in Caco-2 cells (Fig. 2DGo). These results show that in hFOB cells, 1,25D3 and the analog had similar transcriptional potencies, whereas in Caco-2 cells, the analog’s potency to induce gene expression was 10- to 50-fold lower than that of the natural hormone depending on the gene regulatory event examined.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 1. VDR-Mediated Gene Regulation in the Human Osteoblast Cell Line, hFOB

A, Expression of the osteoblast-specific gene osteocalcin was examined in confluent cultures of hFOB cells incubated for 72 h at 37 C and then the indicated ligands were added for 48 h. The amounts of osteocalcin in the culture medium were measured in triplicates by RIA. B, Reporter gene expression was examined in hFOB cells transfected with a reporter gene containing the minimal human ocVDRE attached to the thymidine kinase-GH gene and incubated at 33 C. GH production was assessed in triplicate in medium samples by RIA 48 h after ligand treatment. ED50s (means ± SEM) for gene expression were calculated from three separate experiments.

 


View larger version (36K):
[in this window]
[in a new window]
  		Fig. 2. VDR-Mediated Gene Regulation in the Human Colon Carcinoma Cell Line, Caco-2

A, 24-Hydroxylase and calbindin D9K mRNA levels were assessed by semiquantitative RT-PCR 24 h after ligand treatment. The gene products were separated by agarose gel electrophoresis and detected by ethidium bromide staining. B and C, 24-Hydroxylase and Calbindin D9K mRNA levels were quantified by densitometric scanning. The mRNA levels in panels B and C were normalized against the glyceraldehyde-3-phosphate dehydrogenase mRNA levels and expressed as arbitrary densitometry units obtained by scanning ethidium bromide-stained agarose gels. D, Reporter gene expression was examined in Caco-2 cells transfected with a transgene containing the minimal human ocVDRE attached to the thymidine kinase-GH gene. GH production was assessed in triplicate in medium samples by RIA 48 h after ligand treatment. The ED50s (means ± SEM) for gene expression shown in panels B–D were calculated from three to five separate experiments.

 
We considered the possibility that the weak transcriptional activity of the analog in Caco-2 cells may be due to a rapid clearance rate. To test this possibility, we examined the analog’s stability in Caco-2 cells over a 24-h period. The cells were incubated with 10 nM of Ro-26-9228 and then, the cells and the medium were extracted and the amount of analog was quantified by mass spectrometry. The results showed that there was no apparent catabolism of the compound over that period of time as 85%, 78%, and 80% of the analog were recovered after 1, 6, and 24 h, respectively. In contrast, studies by others have shown that in Caco-2 cells, the amount of newly synthesized 1,25D3 reached a maximum by 3 h, but was rapidly catabolized to nondetectable amounts by 24 h (38). These results suggest that it is not likely that rapid catabolism is the reason for poor transcriptional activity of Ro-26-9228 in Caco-2 cells.

Ligand-Mediated Properties of in Vitro-Synthesized and Cellular VDRs
The sequence of events leading to VDR-mediated transcription includes ligand binding and a conformational change in the VDR. This conformational change is necessary for the VDR to acquire the ability to bind RXR and DNA and eventually recruit coactivators of transcription to the VDR-ligand complex. The apparent selective transcriptional activity of Ro-26-9228 could be due to a number of reasons. One is that the VDR-analog complex may form a poor agonist conformation or even an antagonist conformation, but the composition of transcription factors in hFOB cells may compensate for that (as has been described for SERMs) (2). Alternatively, VDR-analog complexes may be postranslationally modified so that they are transcriptionally active in one cell type and inactive in another.

We and other investigators in the field have shown that the potency of in vitro-synthesized VDR (ivtVDR) to form a protease-resistant conformation in vitro correlates with its ability to recruit dimerization partners and coactivators of transcription to the complex and that these in vitro characteristics also correlate, in most cases, with the transcriptional activities in live cells (39). When antagonists of 1,25D3 are used, the ivtVDR-analog complex assumes a conformation that is not permissive for interaction with coactivators of transcription (29, 40). We rationalized that testing these activities of the ivtVDR, while measuring the activities of VDR-ligand complexes from both cell types, would allow us to determine whether the ivtVDR-analog complex forms an agonist or an antagonist conformation in vitro and whether hFOB and Caco-2 cellular environments directly affect the functional conformation of the VDR [defined as a conformation that enables interaction with RXR and with the p160 coactivator GRIP (glucocorticoid receptor-interacting protein)].

The protease sensitivity assay revealed that the analog stabilized ivtVDR in a conformation that was similar to that stabilized by 1,25D3 by inducing two trypsin-resistant fragments (a prominent 34-kDa and a weak 28-kDa fragments). However, when ED50 for stabilization of the 34-kDa fragment was calculated it was 100 nM for analog-VDR complexes, and 0.5 nM for 1,25D3-VDR complexes (Fig. 3Go). These results suggested that the analog might be a poor agonist rather than an antagonist. To confirm that, we performed pull-down assays with the ivtVDR and glutathione-S-transferase (GST)-GRIP or GST-RXR. The ivtVDR-analog complexes did bind these partners of transcription, but their potency to do so was a more than a 100 times lower than that of 1,25D3 (Fig. 4Go). We then examined how the cellular environment in Caco-2 and in hFOB cells affects these activities. The two cell types were incubated with graded concentrations of 1,25D3 and Ro-26-9228 for 1 h, whole cell extracts were incubated with GST-GRIP or GST-RXR and the bound VDRs were detected by Western blotting. These experiments revealed that 1,25D3 induced a substantial interaction of the VDR from both cell types with both GRIP and RXR. In contrast, Ro-26-9228 did not induce these interactions in VDR from Caco-2 cells even at a concentration of 1 µM, but it induced substantial interaction with GRIP and RXR in VDR from hFOB cells (Fig. 4Go).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 3. Ligand Potencies to Stabilize hVDR in a Protease-Resistant Conformation

In vitro-translated 35S-labeled hVDR was incubated with or without the indicated concentrations of 1,25D3 or Ro-26-9228 before digestion with trypsin. The trypsin-resistant fragments (34 and 28 kDa) were separated by SDS-PAGE and detected by autoradiography.

 


View larger version (37K):
[in this window]
[in a new window]
  		Fig. 4. Differential Interaction of VDR with Partners of Transcription in 1,25D3-Treated and Ro-26-9228-Treated Cells

A, Whole cell extracts from confluent cultures of Caco-2 or hFOB cells were prepared 1 h after incubation with vehicle (un) or the indicated concentration (M) of ligand, in serum-free medium. Western blots were performed with VDR antibodies to assess the amount of VDR in the extracts. B, Quantitative pull-down experiments were performed with extracts of Caco-2 and hFOB cells prepared 1 h after treatment of the cells with the indicated ligand concentration in serum-free medium. Two hundred to 300 µg of each extract were incubated with glutathione-Sepharose beads and GST-GRIP or GST-RXR in a final volume of 300 µl, for 1 h at 4 C. Then the beads were washed and the VDR bound to the beads was extracted by boiling in Lammeli buffer. Samples were separated by SDS-PAGE, and VDR was detected by Western blotting using VDR antibodies. 35S-VDR was prepared by in vitro translation using a human VDR cDNA template. The translated VDR was incubated with ligand for 1 h at 30 C and then subjected to pull-down assay under the same conditions as cellular VDR. 35S-VDR bound to GST-GRIP or GST-RXR was detected by SDS-PAGE and autoradiography of the dried gels. C, The GRIP-bound VDR from the Western blots or the autoradiograph were quantified by densitometric scanning. The results are expressed as percentages of maximal signal on each blot.

 
These results suggested that the initial steps in the acquisition of transcriptionally active conformation are similar for the analog-bound ivtVDR and for the analog-bound VDR from hFOB cells. However, in Caco-2 cells the VDR/analog complexes are completely repressed.

We next considered the possibility that the inability of the analog to induce interaction of VDR from Caco-2 cells with partners of transcription could be due to poor uptake of this ligand or to inability of the cellular VDR in Caco-2 cells to bind the analog. To determine whether VDR in Caco-2 cells forms complexes with the analog at the concentrations used for the coactivator interaction assays (10 nM to 1 µM), we incubated the cells with the analog under those conditions and quantified the occupied and unoccupied receptor complexes. Table 1Go shows that 96% of the VDR binding sites in Caco-2 cells were occupied by the analog at 1 µM, 75% at 100 nM, and 56% at 10 nM, whereas 95–97% of the VDR binding sites were occupied in Caco-2 cells treated with those concentrations of 1,25D3. The ED50 for saturation of the receptor by the analog was therefore about 10 nM, which is very similar to the IC50 value of 9.6 nM, obtained by competition assays with recombinant human VDR (hVDR) in vitro (36). These results suggest that the inability of the VDR-analog complexes to form interaction with coactivators was not due to poor uptake and poor binding to cellular VDR in Caco-2 cells.


View this table:
[in this window]
[in a new window]
  		Table 1. [3H]1,25D3 Binding to Caco-2 Cell Extracts Treated in Vivo with Nonradioactive Ligands1

 
Differential Regulation of VDR’s Subcellular Distribution in Caco-2 and hFOB Cells
The ability of VDR to adopt a transcriptionally active conformation and interact with RXR also determines its ability to localize in the nucleus, where transcription occurs (41). To examine this aspect of ligand-mediated VDR activation and to determine whether it correlates with the results of the in vitro assays, we compared the abilities of the VDR-1,25D3 and VDR-analog complexes to localize in the nuclei of Caco-2 and hFOB cells.

To this end, we implemented fractionation method that examined VDR in three subcellular fractions: a soluble fraction (cytoplasmic), chromatin-associated fraction and nuclear matrix (42). The results of these experiments demonstrated that in hFOB cells the VDR was localized in the cytoplasmic and the nuclear matrix fractions in the absence of ligand, and no VDR was detected in chromatin (Fig. 5Go). Ligand treatment (either 1,25D3 or analog) rapidly induced accumulation of the VDR in the chromatin fraction, decreased VDR in the cytoplasmic fraction, and increased VDR in the nuclear matrix. The accumulation of VDR in the chromatin appeared to be similar in 1,25D3- and analog-treated cells. These results suggest that in hFOB cells, the unoccupied VDR prefers nuclear export to nuclear import, whereas the ligand-bound VDR favors nuclear import. These results also confirmed that the VDR in the hFOB cells was able to form the conformation required for nuclear localization, and both the hormone and the analog effectively induced this activity (Fig. 5Go). Also, the total amount of VDR was increased by ligand treatment; at 1 h, a 5-fold increase was evident in whole cell lysates (Fig. 5DGo) and the subcellular localization suggested this increase was mostly in the chromatin and to a less extent in the nuclear matrix (Fig. 5AGo). After 24 h, the amount increased by 10-fold (Fig. 5DGo), and this increase was observed in all three fractions from both 1,25D3- and analog-treated cells (Fig. 5AGo).



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 5. Subcellular Distribution of VDR in hFOB Cells

A, VDR was detected in the cytoplasmic, chromatin, and nuclear matrix of hFOB cells by Western blotting. B, To detect chromatin-associated and nuclear membrane-associated proteins, each fraction was examined for presence of histone H4 (chromatin) and Lamin A (nuclear envelope) by Western blotting. C, The VDR in subcellular fractions was quantified by densitometric scanning of the Western blots in panel A and is expressed as percentage of total densitometry units in cytoplasm, chromatin and nuclear matrix. D, Detection of VDR by Western blot of whole cell lysates prepared from vehicle-, 1,25D3-, or Ro-26-9228-treated cells. Treatments were for 1 h or 24 h with 100 nM of the indicated ligand. C, Cytosol; Ch, chromatin; NM, nuclear matrix; D3, 1,25D3; Ro, Ro-26-9228.

 
When these experiments were repeated with Caco-2 cells, we found several differences: the VDR localized in both the cytoplasmic fraction (70–80%) and in the chromatin (20–30%), even in the absence of ligand, but not in the nuclear matrix (Fig. 6Go). Treatment with 1,25D3 did not change the abundance of VDR in the cytoplasmic fraction but caused the accumulation of a faster-migrating VDR in the chromatin. In analog-treated cells, there was no change in the abundance of the cytoplasmic VDR and the faster-migrating chromatin-associated VDR was barely detectable. We concluded that the mechanisms for localization of VDR in chromatin are different in Caco-2 and hFOB cells for two reasons. First, in Caco-2 cells the VDR was in the cytoplasm and the chromatin whether or not the VDR was ligand-bound, whereas in hFOB cells only ligand-bound VDR was in the chromatin. Second, if ligand binding did affect chromatin localization in Caco-2 cells, it was only of the less abundant, faster migrating form of the VDR that was either induced, or stabilized by 1,25D3 binding, but not by analog binding.



View larger version (21K):
[in this window]
[in a new window]
  		Fig. 6. Subcellular Distribution of VDR in Caco-2 Cells

A, VDR was detected in the cytoplasmic, chromatin, and nuclear matrix of Caco-2 cells by Western blotting 1 or 24 h after treatment with vehicle or 100 nM of the indicated ligand. B, To detect chromatin-associated and nuclear membrane-associated proteins, each fraction was examined for presence of histone H4 (chromatin) and Lamin A (nuclear envelope) by Western blotting. C, Cytosol; Ch, chromatin; NM, nuclear matrix; D3, 1,25D3; Ro, Ro-26-9228.

 
These experiments suggest that nuclear export and import of VDR are in equilibrium in Caco-2 cells regardless of ligand binding state, whereas in hFOB cells, unoccupied VDR is exported and ligand-occupied VDR is imported. That even 1,25D3 failed to induce substantial VDR accumulation in the chromatin suggests that in Caco-2 cells most of the ligand-bound VDR is not in a conformation that enables it to bind coactivators. That the faster-migrating VDR in Caco-2 cells is localized exclusively in chromatin suggests that it is the form that binds RXR and coactivators of transcription.

Differential Subcellular Localization of Active and Inactive Forms of VDR
To further substantiate the cause-and-effect relationship between ligand-mediated VDR accumulation in the chromatin and VDR’s ability to interact with coactivators, we examined the abilities of cytoplasmic and chromatin-bound VDR to interact with GST-GRIP in vitro.

In the human osteoblast cell line hFOB, we found that the cytoplasmic and nuclear VDRs from ligand-treated cells were able to interact with GST-GRIP, with similar efficiency in 1,25D3-treated and analog-treated cells (Fig. 7Go). When these experiments were repeated with Caco-2 cells we found several differences. First, cytoplasmic VDR from untreated, 1,25D3-treated and analog-treated cells did not bind GST-GRIP (Fig. 8Go). In contrast, the VDR extracted from the chromatin of 1,25D3-treated Caco-2 cells did, but neither VDR from the chromatin of untreated cells nor VDR from the chromatin of analog-treated cells could. Next, we wished to determine whether the VDR that bound GST-GRIP was the slow-migrating or the fast-migrating isoform, and whether the fast-migrating VDR from Caco-2 cells had the same electrophoretic mobility as the VDR from hFOB cells. To that end, we extracted cytoplasmic VDR and nuclear VDR from 1,25D3-treated Caco-2 and hFOB cells and ran these samples side by side with VDR-1,25D3 complexes from Caco-2 cells that bound to GST-GRIP (Fig. 8Go, bottom panel). Our results suggest that the fast-migrating VDR, not the slow-migrating one had the ability to bind GRIP, in vitro. Furthermore, these results showed that the electrophoretic mobility of VDR from hFOB cells is the same as that of the fast-migrating VDR in Caco-2 cells.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 7. Subcellular Localization of Activated VDR in hFOB Cells

The upper row shows Western blotting of the immunoreactive VDR in each subcellular fraction (Input). The second row shows VDR that bound to GST-GRIP by pull-down assay (PD). C, Cytosol; Ch, chromatin; 1,25D3, 1,25-dihydroxyvitamin D3; Ro, Ro-26-9228.

 


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 8. Subcellular Localization of Activated VDR in Caco-2 Cells

The upper row shows Western blotting of the immunoreactive VDR in each subcellular fraction (Input). The second row shows VDR that bound to GST-GRIP by pull-down assay (PD). The bottom panel shows VDR by Western blotting of a gel loaded with subcellular fractions from hFOB or Caco-2 cells, 24 h after treatment with 1,25D3, and a GST-GRIP-bound VDR from 1,25D3-treated Caco-2 cells. C, Cytosol; Ch, chromatin; 1,25D3, 1,25-dihydroxyvitamin D3; Ro, Ro-26-9228.

 
Taken together, these experiments suggested that increased retention of ligand-bound VDR in the chromatin directly reflects VDR’s ability to form a conformation that is transcriptionally active, by enabling the VDR to bind transcription coactivators. However, the question still remained why the 1,25D3-treated cytoplasmic VDR in Caco-2 cells could not interact with coactivators, whereas the hormone- and analog-treated cytoplasmic VDR in hFOB cells could. One simple explanation is that in Caco-2 cells cytoplasmic VDR cannot bind ligand and the chromatin-associated VDR can, or that the fractionation procedure used disrupts ligand-binding activity. To examine these possibilities, we performed subcellular fractionation of Caco-2 cells and assayed [3H]-1,25D3 binding activities in the cytoplasmic and nuclear fractions (Table 2Go). This binding activity was compared with that of whole cell homogenates, prepared by a standard method (26). We found that both cytoplasmic and chromatin-associated VDR could bind [3H]-1,25D3, although the fractionation procedure may have caused some loss of ligand binding activity. Furthermore, VDR that bound 1,25D3 in intact cells remained occupied during the subcellular fractionation procedure (Table 2Go). Therefore, we concluded that the ability of cytoplasmic VDR in Caco-2 cells to bind ligand was not sufficient to induce interaction with GRIP in vitro.


View this table:
[in this window]
[in a new window]
  		Table 2. [3H]1,25D3 Binding in Subcellular Fractions of Caco-2 Cells1

 
Another possible explanation was that in Caco-2 cells the ligand-bound cytoplasmic and chromatin-associated VDRs were biochemically different, whereas in hFOB cells they were the same. Our finding of one cytoplasmic form and two chromatin forms of VDR in Caco-2 cells but not in hFOB cells supports this explanation. Because the chromatin-associated fast-migrating VDR from Caco-2 cells is probably the one that bound GST-GRIP (Fig. 8Go), we hypothesized that in Caco-2 cells ligand treatment transforms the VDR from a slow-migrating, inactive form to a fast-migrating active one. Alternatively, ligand treatment may stabilize the fast-migrating receptor that is otherwise more susceptible to proteolytic degradation and therefore cannot be detected in the absence of ligand. A third possibility is that the slow-migrating, inactive but protease-resistant form transforms to a fastmigrating, degradable form in a ligand-independent fashion, but the binding to ligand stabilizes the fast-migrating form leading to its accumulation in the chromatin fraction.

Several studies have shown that unbound VDR probably undergoes ubiquitination and proteasome-mediated degradation, and ligand binding or treatment with proteasome inhibitors prevents this degradation (43, 44). Therefore, we hypothesized that a degradable form of VDR may be substrate to the proteasome. To distinguish proteasome-sensitive and nonsensitive VDR forms, we treated hFOB and Caco-2 cells with the proteasome inhibitor MG132, and performed Western blot analysis of whole cell lysates and subcellular fractions. These experiments showed that in hFOB cells, VDR increased substantially in vehicle-treated controls incubated with the proteasome inhibitor and in cells treated only with ligand (Fig. 9Go). These results demonstrated that in hFOB cells almost all of the VDR is susceptible to degradation by the proteasome and that ligand binding probably inhibits this degradation (Fig. 9Go).



View larger version (37K):
[in this window]
[in a new window]
  		Fig. 9. Effect of the Proteasome Inhibitor MG132 on Subcellular Distribution of VDR, Its Ability to Interact with GST-GRIP, and Its Transcriptional Activity

A, Caco-2 and hFOB cells were incubated for 24 h without or with 25 µM MG132, in the presence or absence of the indicated ligand at a concentration of 100 nM. VDR was quantified by Western blotting in whole-cell lysates (WC) of hFOB and in subcellular fractions of Caco-2 cells. Chromatin fractions of Caco-2 cells were also subjected to a pull-down assay with GST-GRIP (PD). B, Confluent cultures of Caco-2 cells were transfected with the 24-hydroxylase promoter-luciferase reporter. Three days later, the transfected cells were treated, in triplicates, without or with 2.5 µM MG132, and the indicated ligand. Cell extracts were collected 24 h later and tested for luciferase activity. Enzyme activity was normalized against protein content in the samples and expressed as means ± SEM. *, P < 0.05 for the difference in reporter gene expression between 1,25D3-treated and DMSO-treated samples. **, P < 0.01 for the difference in reporter gene expression between MG132-treated and MG-132 + ligand-treated cells. Vehicle, DMSO; Ch, chromatin; C, cytosol; MG, MG132, 1,25D3, 1,25-dihydroxyvitamin D3; Ro, Ro-26-9228.

 
To determine presence of proteasome-sensitive VDR in Caco-2 cells and its subcellular localization, we prepared cytoplasmic and chromatin fractions from Caco-2 cells treated either with ligand or with a combination of MG132 and ligand. We found that cytoplasmic VDR from Caco-2 cells was not sensitive to MG132 treatment or to ligand binding. On the other hand, both MG132 and ligand increased the abundance of the fast-migrating receptor form in the chromatin and decreased the abundance of the slow-migrating form (Fig. 9Go). This shift in abundance toward the fast-migrating isoform of VDR was more evident in analog-treated than in 1,25D3-treated chromatin samples, suggesting that ligand binding transforms one receptor type to the other. When Caco-2 cell extracts from MG132-treated cells were examined for interaction with GST-GRIP, we found that there was no significant increase in VDR binding to the coactivator in 1,25D3-treated samples but treatment with MG132 significantly increased the GRIP binding activity of the analog-treated VDR. These results suggest that in Caco-2 cells the amount of the proteasome-sensitive form of VDR is rate limiting for analog action but not for 1,25D3 action.

To determine whether the ability of MG132 to increase the abundance of the VDR form that binds GRIP in vitro was reflected in transcriptional activity in intact cells, we examined 1,25D3- or analog-induced expression of a luciferase reporter gene driven by the 24-hydroxylase promoter in Caco-2 cells treated with or without the proteasome inhibitor MG132. Our findings (Fig. 9BGo) showed that 1,25D3 up-regulated reporter gene expression (70%) and the combination of 1,25D3 and MG132 caused a further increase in the 1,25D3-induced reporter gene expression (130%). As for analog-treated cells, there was no detectable stimulation of reporter gene expression in the absence of the proteasome inhibitor, but in its presence, transcription was restored to levels similar to those induced by the combination of 1,25D3 and MG132 (140–150% increase in reporter gene expression).

Taken together, these results suggest that in both cell types the VDR form that can interact with coactivators of transcription in vitro is proteasome-sensitive. The apparent ligand-selective activation of VDR in Caco-2 cells is probably due to the analog’s reduced ability to induce accumulation of a fast-migrating receptor form. Our experiments, however, did not reveal whether this poor ability of the analog to stabilize the active VDR form in Caco-2 cells is due to inefficient transformation from the stable, slow-migrating VDR to the degradable fast-migrating one, or to poor ability of the analog to protect the fast-migrating VDR against proteasome-mediated degradation.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIAL AND METHODS
 REFERENCES
 
The results of the present study suggest that the cell-specific action of Ro-26-9228 might be due to direct modification of the VDR.

Several of our observations support the idea that the VDRs in Caco-2 cells and in hFOB cells are different (These observations are summarized in Fig. 10Go). One is that most of the VDR in hFOB cells was proteasome-sensitive, whereas in Caco-2 cells most of it was not. Second, the proteasome-resistant form of the VDR in Caco-2 cells apparently shuttled between the cytoplasm and the nucleus, whereas the proteasome- sensitive VDR in both cell types had a greater preference for the chromatin and that preference was induced by ligand binding. Third, the proteasome-resistant VDR in Caco-2 cells had ligand binding activity but did not bind coactivators, regardless of whether it bound 1,25D3 or analog, suggesting that it is associated with a factor that blocks access to the coactivator binding site or that the site is modified in a way that makes it inactive. Fourth, a small amount of proteasome-sensitive VDR did accumulate in Caco-2 cells after 1,25D3 treatment but only in the chromatin. That the proteasome-sensitive and the proteasome-resistant VDRs were different in size further supports the idea of a posttranslational modification in Caco-2 cells that renders most of the VDR transcriptionally inactive. Our experiments provided preliminary evidence for a transformation in Caco-2 cells that converts the inactive form of VDR to an active one, especially in chromatin of analog-treated cells or in chromatin of cells treated with both analog and MG132 (Figs. 8Go and 9Go) because ligand binding decreased the abundance of the slow-migrating VDR and increased the abundance of the fast-migrating VDR in these cells. However, the role of ligand in this transformation is inconclusive. If that transformation event does exist and is ligand modulated, then the selective action of the analog is due to poor ability to transform the VDR from the inactive form to the active form. However, we cannot exclude the possibility that the transformation is ligand independent, and the analog is less able than 1,25D3 to protect the proteasome-sensitive form of the VDR against degradation (Fig. 10Go). This possibility may be valid if the critical components of the ubiquitination and proteasome-mediated protein degradation machinery are more abundant in Caco-2 cells than in hFOB cells.



View larger version (22K):
[in this window]
[in a new window]
  		Fig. 10. A Model for the Cell-Selective Activation of the VDR

In hFOB cells most or all of the VDR is substrate to ubiquitination (ub) and is proteasome-sensitive [VDR (PS)]. Upon binding ligand, the VDR (PS) is stabilized [VDR (Pstab)], and acquires the potential to accumulate in the chromatin, and to bind coactivators of transcription (triangles, p160). In Caco-2 cells, the VDR exists in two forms. One shuttles between the cytoplasm and the nucleus, and can bind ligand but is proteasome resistant [VDR (PR)] and unable to bind coactivators of transcription. The other form is substrate to ubiquitination (ub), is proteasome sensitive, is localized exclusively in the chromatin, and can bind coactivators of transcription (p160). We propose that in Caco-2 cells the ability of individual ligands to transform the proteasome-resistant to proteasome-sensitive VDR and/or to protect the proteasome-sensitive VDR from degradation may determine the amount of transcriptionally active VDR in the cells. 1,25D3 is effective at this process, but Ro-26-9228 (Ro) is not.

 
What are the molecular events associated with VDR transformation from the inactive to the active, proteasome-sensitive form? That the inactive VDR had a slower electrophoretic mobility than the active, ligand-stabilized VDR would suggest a change in phosphorylation state, such as dephosphorylation of the inactive VDR that converts it into a proteasome-sensitive active form with faster electrophoretic mobility. A recent publication reporting that the cytoplasmic VDR in Caco-2 cells is in a ternary complex with a phosphatase and a kinase, that this complex is cell specific, and that ligand-induced dephosphorylation of the kinase changes the composition of the complex by releasing the kinase provides an attractive explanation for ligand-modulated transition of VDR between phosphorylation states (45). First, the VDR itself may be a substrate for either the kinase or the phosphatase or both. Second, the release of the kinase from the complex on ligand binding may give rise to a less phosphorylated but transcriptionally active form of the VDR, and the analog may simply be less effective in doing so. If the VDR in hFOB cells does not undergo this transformation process, then VDR activation is more direct, and involves only ligand-mediated stabilization of the proteasome-sensitive form of the VDR in a transcriptionally active conformation, which is directly proportional to the abilities of each compound to induce these activities in synthetic VDR in a cell-free system.

That there is a direct correlation between the ability of the analog to induce the accumulation of transcriptionally active form of the VDR in the two cell types, and analog’s differential transcription activity in these cells might explain its selective action in culture but does not prove that these different mechanisms for ligand-mediated VDR transformation and activation exist in normal intestinal or duodenal mucosa and normal osteoblasts, and that actually contribute to the tissue selectivity of the analog in vivo. However, these findings do suggest regulatory mechanisms of VDR action that merit pursuing, even for the purpose of applying vitamin D analogs for treatment of malignant cells.

Do cell-specific modifications of the VDR completely account for the selective transcriptional activity of the analog in the two cell types we examined? The analog’s potency to induce interaction of VDR from hFOB cells with coactivators was still more than a hundred-fold lower than the ability of 1,25D3 to induce this activity, and yet, the transcriptional potency of the two ligands in these cells was the same, suggesting that there may be another mechanism in hFOB cells that increases the transcriptional activity of the poor agonist, Ro-26-9228. Analysis of the components of the transcription complexes recruited by the analog to the VDR in the two cell types may shed light on this possibility.


   MATERIAL AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIAL AND METHODS
REFERENCES
 
Materials
1,25-Dihydroxyvitamin D3 and Ro-26-9228 were generous gifts from Dr. M. Uskokovic (Roche Bioscience, Palo Alto, CA). MG132 (Z-Leu-Leu-Leu-H) was purchased from Peptides International, Inc. (Osaka, Japan). VDR antibodies (MA1-710) were from Affinity Bioreagents, Inc. (Golden, CO). Lamin A and histone H4 antibodies were from Cell Signaling Technology (Beverly, MA). Immunoradiometric assay kits for measurement of osteocalcin and GH were from Nichols Institute Diagnostic (San Clemente, CA). The luciferase reporter assay system was from Promega, Inc. (Madison, WI)

Preparation of RNA from Cultured Cells and Analysis of Gene Expression
Caco-2 cells (2 d post confluence) were harvested 8 or 24 h after treatment with the indicated doses of either 1,25D3 or Ro-26-9228, and total RNA was extracted by the guanidine/phenol/chloroform procedure (46).

To measure 24-hydroxylase and calbindin D9K mRNA, we performed semiquantitative RT-PCR by using the following primers. The forward primer for human 24-hydroxylase was 5'-TCATGCTAAATACCCAGGTGTTGG-3', and the reverse primer was 5'-TTATCGCTGGCAAAACGCATGG-3'. The forward primer for human calbindin D9K was 5'-AAACTCCTCTTTGATTCTTC-3', and the reverse primer was 5' ATCTCCATTCTTGTCCAGTTC-3'. The forward primer for human glyceraldehyde-3-phosphate dehydrogenase was 5'-TGAAGGTCGGAGTCAACGGATTTGG-3', and the reverse primer was 5'-CATGTAGGGCCATGAGGTCCACCAC-3'. Primer sequences were obtained from GenBank.

RT-PCR was performed as follows. Five micrograms of total RNA were heat-denatured with 1 µg of random primer for 10 min at 70 C and then placed on ice, and first-strand buffer (Invitrogen, Carlsbad, CA), dithiothreitol, and 25 mM deoxynucleotide triphosphate were added. This mixture was incubated 10 min at 25 C and 2 min at 42 C, 1 µl of reverse transcriptase (40 U, Superscript II, Invitrogen) was added, and the reaction was continued for 50 min at 42 C. Semiquantitative PCR was performed by using three different amounts of cDNA (2.5%, 5%, and 10% of the reverse transcription reaction mixture) and the appropriate set of primers in a final volume of 50 µl containing the reaction buffer and Taq polymerase (Roche). The amplification was performed for 30 cycles of 1 min at 95 C, 1 min at 55 C, and 2 min at 72 C, with a final step of 6 min at 72 C, in a DNA thermal cycler (PerkinElmer Cetus, Foster City, CA). A fraction of the reaction mixture was analyzed by agarose gel electrophoresis and visualized by ethidium bromide staining under ultraviolet light. The PCR signals were quantified relative to the G3P product by densitometric scanning of the ethidium bromide-stained gels.

Transfections and Transcription Assays
Caco-2 human colon carcinoma cells were obtained from the American Type Culture Collection and maintained in DMEM and 10% fetal bovine serum (FBS). The T antigen-transformed human fetal osteoblast line, hFOB (a generous gift from Dr. T. Spelsberg, Rochester, MN) was maintained at 33 C in DMEM/F12 medium containing 10% FBS and 400 µg/ml G418. Forty-eight hours before transfection, the cells were plated in 35-mm dishes at a density of 3 x 105 cells/dish. The cells were transfected by the DEAE-dextran method with 2 µg/dish of a reporter construct containing either the human ocVDRE linked to the thymidine kinase promoter and the GH reporter gene (47), or the 24-hydroxylase promoter linked to the luciferase reporter gene (48). Medium was collected 2 d after transfection, and GH in the medium was measured by using a RIA as described by the manufacturer (Nichols Institute). Cells transfected with the luciferase reporter gene were treated 3 d after transfection with dimethylsulfoxide (DMSO) (0.05%) or with 2.5 µM MG132 and the indicated ligands, and cell lysates were collected 24 h later. Luciferase assays were conducted as previously described (48), and enzyme activity in each sample was normalized against total protein content in the cell lysate.

Pharmacokinetics
Caco-2 cells were grown to confluence in 35 mm dishes in DMEM containing 10% FBS. Ro-26–9226 (10 nM) was added for 1, 6, and 24 h and then the cells were scraped into the culture medium and extracted with an equal volume of 20:80 (vol: vol) ethyl acetate: hexane after adding 10 ng of the internal standard, Ro-26-9228/002 (2H10-Ro-26-9228). A control sample (2 ml of the culture medium containing 10 nM of Ro-26-9228) was subjected to the same extraction procedure. The samples (in duplicates) were agitated for 10 min and then centrifuged at 5 C for 10 min. The aqueous (bottom) layer was then frozen on dry ice and the upper, organic layer decanted into a clean tube. The organic layer was evaporated and then reconstituted in 70% (vol: vol) methanol in water. The reconstituted solution was subjected to liquid chromatography/mass spectrometry with a Hewlett-Packard 1090 HPLC pump, a cooled autosampler compartment (Hewlett-Packard), and a Finnigan TSQ 7000 mass spectrometer (Finnigan-MAT, San Jose, CA) as described previously (36). Briefly, the mobile phase consisted of a premixed solution consisting of 20:75:5 (vol: vol: vol) ammonium formate (50 mM, pH 4.2): methanol: acetonitrile. The flow rate was 0.25 ml/min. The column used was a C8 Hypersil column from Column Engineering with dimensions of 2 mm x 150 mm x 5 mm. The column temperature was kept at 40 C. The compounds were ionized by atmospheric pressure chemical ionization and monitored at mass to charge ratio (m/z) 425 for Ro-26-9228 and at m/z 435 for its deca-deuterated internal standard, Ro-26-9228/002. Detection was accomplished by main beam select ion monitoring. The retention time for both was about 7 min. The ratio of Ro-26-9228 peak height to Ro-26-9228/002 peak height was fitted to quadratic equation with 1/(concentration)2 weighting to generate a calibration curve. The amounts of Ro-26-9228 in the cell extracts were interpolated from a calibration curve constructed with known concentrations of Ro-26-9228.

Receptor Binding Assays
To assess the binding of 1,25D3 and Ro-26-9228 in intact cells, cultures of Caco-2 cells, 2 d post confluence were washed twice and scraped into cold PBS, centrifuged at 800 x g for 10 min at 4 C, and resuspended and homogenized in KTED [10 mM Tris-HCl (pH 7.4), 1.5 mM EDTA, 0.3 M KCl, and 1 mM dithiothreitol] (26, 46). The homogenates were aliquoted into tubes containing 0.2 pmol of [3H]1,25D3 without or with excess nonradioactive ligand. The mixtures were incubated on ice for 3–4 h, after which free ligand was separated from bound with hydroxyapatite. The bound ligand was released from the hydroxyapatite by ethanol extraction, and the radioactivity was measured by scintillation counting. To assess binding activity in subcellular fractions, cytosol, chromatin, and nuclear matrix were isolated by sequential extractions as described below (42) and dialyzed for 2 h and the binding assay was performed in duplicates, as described above, except that the samples were incubated with [3H]1,25D3 for 16 h.

Protease Sensitivity Assays
The human VDR was synthesized and labeled in vitro with [35S]-methionine (1000 Ci/mmol) using the transcription and translation (TNT)-coupled transcription/translation system (Promega) and a plasmid containing the hVDR cDNA (pGEM4-hVDR). The translated receptor preparations were incubated with 1,25D3 or analog for 10 min at ambient temperature. Next, trypsin was added to a concentration of 20 µg/ml and the mixtures incubated for 10 min. The digestion products were then separated by SDS-PAGE and detected by autoradiography.

Pull-Down Assays
To determine the potency of ligands to induce interaction of VDR with RXR{alpha} or with the GRIP, we used GST-pull-down assays (28). For pull-down reaction with ivtVDR, binding reactions containing 11 µl of PBSDP buffer [1 mM dithiothreitol and 10 mM phenylmethylsulfonyl fluoride (PMSF) in PBS], 3 µl of 35S-labeled VDR and 1 µl of ethanol or ligand (in ethanol) were incubated at 30 C for 1 h. Then, 3–5 µg of purified GST fusion protein and 20 µl of glutathione-Sepharose beads (equilibrated in PBSDP buffer) were added, and the volume brought up to 100 µl with NETND buffer [20 mM Tris-HCl (pH 7.8), 100 mM NaCl, 1 mM EDTA, 0.2% Nonidet P-40, and 1 mM dithiothreitol]. The mixtures were incubated at 4 C for 1 h, and then the beads were washed once with NETND and twice with PBSDP. The bound proteins were eluted from the beads by boiling in Laemmli buffer for 3 min and analyzed by SDS-PAGE and autoradiography.

For pull-down assays with cellular VDR, 2-d postconfluence cultures of Caco-2 or hFOB cells (in 150-mm x 25-mm plates) were incubated with or without the indicated amounts of 1,25D3 or Ro-26-9228. Then the cells were washed once and scraped into cold PBS. Whole-cell extracts were prepared in buffer C [0.4 M NaCl, 1 mM dithiothreitol, 20 mM HEPES (pH 7.5), 25% glycerol, 1.5 mM MgCl2, 0.2 mM EDTA, 5 mM PMSF, and 1x protease inhibitor cocktail (Roche)]. The cell pellet was resuspended in two volumes of buffer C, subjected to 10 strokes with a Teflon pestle, and incubated on ice for 1 h, and then the supernatant was collected by 15 min of centrifugation at 14,000 x g. Each cell extract (200–300 µg of total proteins) was incubated with 3–5 µg of purified GST fusion protein and 20 µl of glutathione-Sepharose beads were added and the volume was brought up to 300 µl with NETND buffer and to a final ligand concentration as that used in the culture dish. The mixtures were incubated at 4 C for 1 h, and then the beads were washed once with NETND and twice with PBSDP. The bound proteins were eluted from the beads by boiling in Laemmeli buffer for 3 min and analyzed by SDS-PAGE and Western blotting for the presence of VDR by using anti-VDR antibodies.

Subcellular Fractionation
Caco-2 and hFOB cells (2 d post confluence) were scraped into ice-cold PBS and centrifuged for 5 min at 4 C at 800 x g. The cell pellets (80 µl) were suspended by adding 300 µl of cytoskeletal buffer [10 mM PIPES (pH 7.0), 100 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 1 mM EGTA, 0.5% Triton X-100, and 1x EDTA-free protease inhibitor cocktail]. The cell suspensions were incubated on ice for 10 min and then centrifuged at 6000 x g for 20 sec. The supernatants (cytosolic fractions) were transferred to another set of tubes. The pellets were resuspended in 300 µl of chromatin extraction buffer [10 mM PIPES (pH 7.0), 50 mM NaCl, 300 mM sucrose, 3 mM MgCl2, 1 mM EGTA, 0.5% Triton X-100, 1x EDTA-free protease inhibitors cocktail, and 400 IU of ribonuclease-free deoxyribonuclease I (Roche)] and the tubes were rocked at room temperature for 30 min. Then, freshly prepared ammonium sulfate was added to a final concentration of 250 mM and the incubation was continued for additional 10 min at room temperature. The tubes were then centrifuged at 6000 x g for 5 min, the supernatants (chromatin fractions) were collected, and the pellets (nuclear matrix) were extracted by boiling for 3 min in 300 µl of Laemmeli buffer (42).

Aliquots of cytosolic, chromatin, and nuclear matrix fractions were separated by 12% SDS-PAGE, and VDR, histone H4 and lamin A were visualized by Western blotting. When indicated, cytosolic and chromatin extracts were dialyzed for 2 h against a buffer consisting of 20 mM Tris-HCl (pH 8.0), 50 mM NaCl, 0.5 mM EDTA, 1 mM dithiothreitol, and 0.5 mM PMSF and then subjected to GST pull-down assays and Western blotting as described above for buffer C-extracted cellular VDR.


   ACKNOWLEDGMENTS
 
We thank Dr. R. M. Evans, Dr. D. Mangelsdorf, and The Salk Institute for the RXR-{alpha} expression plasmid, Dr. H. DeLuca for the 24-hydroxylase-luciferase reporter plasmid, and Dr. Maureen E. Goode for her helpful comments on this manuscript.


   FOOTNOTES
 
This work was supported by NIH Grants DK 50583 (to S.P.) and DK 54111 (to J.C.F.).

Abbreviations: AP-1, Activator protein-1; DMSO, dimethylsulfoxide; FBS, fetal bovine serum; GRIP, glucocorticoid receptor-interacting protein; GST, glutathione-S-transferase; hFOB, human osteoblast cell line; hVDR, human VDR; ivtVDR, in vitro-synthesized VDR; ocVDRE, osteocalcin vitamin D-responsive element; PMSF, phenylmethylsulfonyl fluoride; Ro-26-9228, 1{alpha}-fluoro-16-ene-20-epi-23-ene-26,27-bishomo-25-hydroxyvitamin D3; RXR, retinoid X receptor; SERMS, selective estrogen receptor modulators; VDR, vitamin D receptor.

Received for publication August 15, 2003. Accepted for publication December 31, 2003.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIAL AND METHODS
 REFERENCES
 

  1. Lonard DM, Smith CL 2002 Molecular perspectives on selective estrogen receptor modulators (SERMs): progress in understanding their tissue-specific agonist and antagonist actions. Steroids 67:15–24[CrossRef][Medline]
  2. Shang Y, Brown M 2002 Molecular determinants for the tissue specificity of SERMs. Science 295:2465–2468[Abstract/Free Full Text]
  3. Pike AC, Brzozowski AM, Walton J, Hubbard RE, Thorsell A, Li Y, Gustafsson J, Carlquist M 2001 Structural insights into the mode of action of a pure antiestrogen. Structure 9:145–153[Medline]
  4. Brzozowski AM, Pike AC, Dauter Z, Hubbard RE, Bonn T, Engstrom O, Ohman L, Greene GL, Gustafsson JA, Carlquist M 1997 Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389:753–758[CrossRef][Medline]
  5. Shiau AK, Barstad D, Loria PM, Cheng L, Kushner PJ, Agard DA, Greene GL 1998 The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95:927–937[CrossRef][Medline]
  6. McInerney EM, Katzenellenbogen BS 1996 Different regions in activation function-1 of the human estrogen receptor required for antiestrogen- and estradiol-dependent transcription activation. J Biol Chem 271:24172–24178[Abstract/Free Full Text]
  7. Berry M, Metzger D, Chambon P 1990 Role of the two activating domains of the oestrogen receptor in the cell-type and promoter-context dependent agonistic activity of the anti-oestrogen 4-hydroxytamoxifen. EMBO J 9:2811–2818[Medline]
  8. Yang NN, Venugopalan M, Hardikar S, Glasebrook A 1996 Identification of an estrogen response element activated by metabolites of 17-estradiol and raloxifene. Science 273:1222–1224[Abstract]
  9. Grese TA, Sluka JP, Bryant HU, Cullinan GJ, Glasebrook AL, Jones CD, Matsumoto K, Palkowitz AD, Sato M, Termine JD, Winter MA, Yang NN, Dodge JA 1997 Molecular determinants of tissue selectivity in estrogen receptor modulators. Proc Natl Acad Sci USA 94:14105–14110[Abstract/Free Full Text]
  10. Webb P, Lopez GN, Uht RM, Kushner PJ 1995 Tamoxifen activation of the estrogen receptor/AP-1 pathway: potential origin for the cell-specific estrogen-like effects of antiestrogens. Mol Endocrinol 9:443–456[Abstract]
  11. Webb P, Nguyen P, Valentine C, Lopez GN, Kwok GR, McInerney E, Katzenellenbogen BS, Enmark E, Gustafsson JA, Nilsson S, Kushner PJ 1999 The estrogen receptor enhances AP-1 activity by two distinct mechanisms with different requirements for receptor transactivation functions. Mol Endocrinol 13:1672–1685[Abstract/Free Full Text]
  12. Liu Z, Auboeuf D, Wong J, Chen JD, Tsai SY, Tsai MJ, O’Malley BW 2002 Coactivator/corepressor ratios modulate PR-mediated transcription by the selective receptor modulator RU486. Proc Natl Acad Sci USA 99:7940–7944[Abstract/Free Full Text]
  13. Bouillon R, Okamura WH, Norman AW 1995 Structure-function relationships in the vitamin D endocrine system. Endocr Rev 16:200–257[CrossRef][Medline]
  14. Norman AW, Ishizuka S, Okamura WH 2001 Ligands for the vitamin D endocrine system: different shapes function as agonists and antagonists for genomic and rapid response receptors or as a ligand for the plasma vitamin D binding protein. J Steroid Biochem Mol Biol 76:49–59[CrossRef][Medline]
  15. Tanizawa T, Imura K, Ishii Y, Nishida S, Takano Y, Mashiba T, Endo N, Takahashi HE 1999 Treatment with active vitamin D metabolites and concurrent treatments in the prevention of hip fractures: a retrospective study. Osteoporos Int 9:163–170[CrossRef][Medline]
  16. Francis RM, Boyle IT, Moniz C, Sutcliffe AM, Davis BS, Beastall GH, Cowan RA, Downes N 1996 A comparison of the effects of alfacalcidol treatment and vitamin D2 supplementation on calcium absorption in elderly women with vertebral fractures. Osteoporos Int 6:284–290[Medline]
  17. Kubodera N, Tsuji N, Uchiyama Y, Endo K 2003 A new active vitamin D analog, ED-71, causes increase in bone mass with preferential effects on bone in osteoporotic patients J Cell Biochem 88:286–289
  18. Funahashi H, Tanaka Y, Imai T, Wada M, Tsukamura K, Hayakawa Y, Matsuura N, Kikumori T, Oiwa M, Tominaga Y, Takagi H 1998 Parathyroid hormone suppression by 22-oxacalcitriol in the severe parathyroid hyperplasia. J Endocrinol Invest 21:43–47[Medline]
  19. Tsukamoto Y, Hanaoka M, Matsuo T, Saruta T, Nomura M, Takahashi Y 2000 Effect of 22-oxacalcitriol on the bone histology of hemodialyzed patients with severe secondary hyperthyroidism. Am J Kidney Dis 35:458–464[Medline]
  20. De Jong EM, Mork NJ, Seijger MM, De La Brassine M, Lauharanta J, Jansen CT, Guilhou JJ, Guillot B, Ostrojic A, Souteyrand P, Vaillant L, Barnes L, Rogers S, Klaber MR, Van De Kerkhof PC 2003 The combination of calcipotriol and methotrexate compared with methotrexate and vehicle in psoriasis: results of a multicentre placebo-controlled randomized trial. Br J Dermatol 148:318–325[CrossRef][Medline]
  21. Kaufmann R, Bibby AJ, Bissonnette R, Cambazard F, Chu AC, Decroix J, Douglas WS, Lowson D, Mascaro JM, Murphy GM, Stymne B 2002 A new calcipotriol/betamethasone dipropionate formulation (Daivobet) is an effective once-daily treatment for psoriasis vulgaris. Dermatology 205:389–393[CrossRef][Medline]
  22. Thaci D, Daiber W, Boehncke WH, Kaufmann R 2001 Calcipotriol solution for the treatment of scalp psoriasis: evaluation of efficacy, safety and acceptance in 3,396 patients. Dermatology 203:153–156[CrossRef][Medline]
  23. Bauer JA, Thompson TA, Church DR, Ariazi EA, Wilding G 2003 Growth inhibition and differentiation in human prostate carcinoma cells induced by the vitamin D analog 1{alpha}, 24-dihydroxyvitamin D2. Prostate 55:159–167[CrossRef][Medline]
  24. Evans TR, Colston KW, Lofts FJ, Cunningham D, Anthoney DA, Gogas H, de Bono JS, Hamberg KJ, Skov T, Mansi JL 2002 A phase II trial of the vitamin D analogue Seocalcitol (EB1089) in patients with inoperable pancreatic cancer. Br J Cancer 86:680–685[CrossRef][Medline]
  25. Gulliford T, English J, Colston KW, Menday P, Moller S, Coombes RC 1998 A phase I study of the vitamin D analogue EB1089 in patients with advanced breasts and colorectal cancer. Br J Cancer 78:6–13[Medline]
  26. Liu YY, Collins ED, Norman AW, Peleg S 1997 Differential interaction of 1{alpha},25-dihydroxyvitamin D3 analogues and their 20-epi homologues with the vitamin D receptor. J Biol Chem 272:3336–3345[Abstract/Free Full Text]
  27. Ozono K, Saito M, Miura D, Michigami T, Nakajima S, Ishizuka S 1999 Analysis of the molecular mechanism for the antagonistic action of a novel 1{alpha},25-dihydroxyvitamin D(3) analogue toward vitamin D receptor function. J Biol Chem 274:32376–32381[Abstract/Free Full Text]
  28. Verstuyf A, Verlinden L, van Etten E, Shi L, Wu Y, D’Halleweyn C, Van Haver D, Zhu GD, Chen YJ, Zhou X, Haussler MR, De Clercq P, Vandewalle M, Van Baelen H, Mathieu C, Bouillon R 2000 Biological activity of CD-ring modified 1{alpha},25-dihydroxyvitamin D analogues: C-ring and five-membered D-ring analogues. J Bone Miner Res 15:237–252[CrossRef][Medline]
  29. Toell A, Gonzalez MM, Ruf D, Steinmeyer A, Ishizuka S, Carlberg C 2001 Different molecular mechanisms of vitamin D(3) receptor antagonists. Mol Pharmacol 59:1478–1485[Abstract/Free Full Text]
  30. Shevde NK, Plum LA, Clagett-Dame M, Yamamoto H, Pike JW, DeLuca HF 2002 A potent analog of 1{alpha},25-dihydroxyvitamin D3 selectively induces bone formation. Proc Natl Acad Sci USA 99:13487–13491[Abstract/Free Full Text]
  31. Nayeri S, Danielsson C, Kahlen JP, Schräder M, Mathiasen IS, Binderup L, Carlberg C 1995 The anti-proliferative effect of vitamin D3 analogues is not related to the AP-1 pathway, but related to promoter selectivity. Oncogene 11:1853–1858[Medline]
  32. Danielsson C, Mathiasen IS, James SY, Nayeri S, Bretting C, Hansen CM, Colston KW, Carlberg C 1997 Sensitive induction of apoptosis in breast cancer cells by a novel 1,25-dihydroxyvitamin D3 analogue is related to promoter selectivity. J Cell Biochem 66:552–562[CrossRef][Medline]
  33. Abe J, Nakano T, Nishii Y, Matsumoto T, Ogata E, Ikeda K 1991 A novel vitamin D3 analog, 22-oxa-1,25-dihydroxyvitamin D3, inhibits the growth of human breast cancer in vitro and in vivo without causing hypercalcemia. Endocrinology 129:832–837[Abstract]
  34. Takeyama K, Masuhiro Y, Fuse H, Endoh H, Murayama A, Kitanaka S, Suzawa M, Yanagisawa J, Kato S 1999 Selective interaction of vitamin D receptor with transcriptional coactivators by a vitamin D analog. Mol Cell Biol 19:1049–1055[Abstract/Free Full Text]
  35. Endo K, Ichikawa F, Uchiyama Y, Katsumata K, Ohkawa H, Kumaki K, Ogata E, Ikeda K 1994 Evidence for the uptake of a vitamin D analogue (OCT) by a human carcinoma and its effect of suppressing the transcription of parathyroid hormone-related peptide gene in vivo. J Biol Chem 269:32693–32699[Abstract/Free Full Text]
  36. Peleg S, Uskokovic M, Ahene A, Vickery B, Avnur Z 2002 Cellular and molecular events associated with the bone-protecting activity of the noncalcemic vitamin D analog Ro-26-9228 in osteopenic rats. Endocrinology 143:1625–1636[Abstract/Free Full Text]
  37. Harris SA, Enger RJ, Riggs BL, Spelsberg TC 1995 Development and characterization of a conditionally immortalized human fetal osteoblastic cell line. J Bone Miner Res 10:178–186[Medline]
  38. Cross HS, Peterlik M, Reddy GS, Schuster I 1997 Vitamin D metabolism in human colon adenocarcinoma-derived Caco-2 cells: expression of 25-hydroxyvitamin D3-1{alpha}-hydroxylase activity and regulation of side-chain metabolism. J Steroid Biochem Mol Biol 62:21–28[CrossRef][Medline]
  39. Peleg S, Nguyen C, Woodard BT, Lee JK, Posner GH 1998 Differential use of transcription activation function 2 domain of the vitamin D receptor by 1,25-dihydroxyvitamin D3 and its A ring-modified analogs. Mol Endocrinol 12:525–535[Abstract/Free Full Text]
  40. Bula CM, Bishop JE, Ishizuka S, Norman AW 2000 25-Dehydro-1{alpha}-hydroxyvitamin D3–26,23S-lactone antagonizes the nuclear vitamin D receptor by mediating a unique noncovalent conformational change. Mol Endocrinol 14:1788–1796[Abstract/Free Full Text]
  41. Prufer K, Racz A, Lin GC, Barsony J 2000 Dimerization with retinoid X receptors promotes nuclear localization and subnuclear targeting of vitamin D receptors. J Biol Chem 275:41114–41123[Abstract/Free Full Text]
  42. van Wijnen AJ, Bidwell JP, Fey EG, Penman S, Lian JB, Stein JL, Stein GS 1993 Nuclear matrix association of multiple sequence-specific DNA binding activities related to SP-1, ATF, CCAAT, C/EBP, OCT-1, and AP-1. Biochemistry 32:8397–8402[CrossRef][Medline]
  43. Li XY, Boudjelal M, Xiao JH, Peng ZH, Asuru A, Kang S, Fisher GJ, Voorhees JJ 1999 1,25-Dihydroxyvitamin D3 increases nuclear vitamin D3 receptors by blocking ubiquitin/proteasome-mediated degradation in human skin. Mol Endocrinol 13:1686–1694[Abstract/Free Full Text]
  44. Jaaskelainen T, Ryhanen S, Mahonen A, DeLuca HF, Maenpaa PH 2000 Mechanism of action of superactive vitamin D analogs through regulated receptor degradation. J Cell Biochem 76:548–558[CrossRef][Medline]
  45. Bettoun DJ, Buck 2nd DW, Lu J, Khalifa B, Chin WW, Nagpal S 2002 A vitamin D receptor-Ser/Thr phosphatase-p70 S6 kinase complex and modulation of its enzymatic activities by the ligand. J Biol Chem 277:24847–24850[Abstract/Free Full Text]</