The human vitamin D receptor (hVDR) is a member of the nuclear receptor superfamily of transcriptional regulators. Here we show that tryptophan 286 of the hVDR is critical for ligand binding and transactivation of 1,25-dihydroxyvitamin D3 [1,25(OH)2D3] target genes. Two mutants of the hVDR were produced, W286A and W286F, in which the tryptophan was replaced with an alanine or a phenylalanine, respectively. The W286A mutant did not bind 1,25(OH)2D3, interact with steroid receptor coactivator 1 (SRC-1) in vitro, or activate transcription. Moreover, the W286A receptor did not heterodimerize in a ligand-dependent manner with the human retinoid X receptor α (hRXRα). Although the W286F receptor heterodimerized with hRXRα, interacted with SRC-1, and bound 1,25(OH)2D3, its capacity to transactivate was attenuated severely. Thus, tryptophan 286 of hVDR plays an important role in specific 1,25(OH)2D3 ligand interaction and subsequently in hVDR/RXR interaction, SRC-1 binding, and ligand-dependent transactivation of 1,25(OH)2D3 target genes. These results identify the first amino acid that is absolutely required for ligand binding in the VDR and further define the structure-function relationship of 1,25(OH)2D3 interaction with its receptor.
THE BIOLOGICALLY active form of vitamin D3, 1,25-dihydroxyvitamin D3 [1,25(OH)2D3], directly modulates the transcription of several target genes by binding to the vitamin D receptor (VDR), a member of the nuclear receptor family of transcriptional regulators. Binding of ligand activates nuclear receptors by inducing a change in the position of the C-terminal activation function 2 (AF-2) domain.(1) The AF-2 domain mediates transactivation by creating an interface for ligand-dependent interaction with coactivators, a bridge to the transcriptional machinery.(2)
The crystal structures of the ligand binding domains (LBDs) of the retinoid X receptor α (RXRα),(3) the retinoic acid receptor γ (RARγ),(4) and the thyroid hormone receptor α (TRα(5) have been determined. All were found to share a similar secondary structure of 12 α-helices and a very low content of β-pleated sheet(6) in which helix 12 is the highly conserved AF-2 domain. No structural data has yet been published for the VDR; however, other nuclear receptors may serve as models on which to base information about the VDR.
The amino acid sequence of human RARγ revealed a tryptophan residue at position 227 in the LBD. Tryptophan fluorescence studies showed that fluorescence was quenched in a dose-dependent manner on addition of all-trans retinoic acid indicating that tryptophan 227 of RARγ is in close contact with bound ligand.(6) Saturation of the rat VDR with 1,25(OH)2D3 quenches the fluorescence of tryptophan 282, suggesting that the ligand and tryptophan also are in close proximity in the rat VDR.(7) Similar results also have been observed for the human RXRα LBD.(8)
The exact amino acids in the human VDR (hVDR) that contact 1,25(OH)2D3 remain unknown. Extensive work with deletion mutants of the LBD has contributed to the elucidation of regions important for ligand binding. Deletion mapping of the hVDR LBD has revealed that amino acids 232–382 are critical for interaction with ligand.(9) Here, we show that tryptophan 286 of the hVDR is critical for ligand binding, coactivator interaction, and transactivation.
MATERIALS AND METHODS
Preparation of the mutant VDRs
The hVDR expression vector hVDRpSG5 was constructed by inserting an EcoRI fragment containing the entire coding region of the hVDR into the EcoRI site of pSG5.(10) The hVDRpSG5 was used as a template to introduce point mutations into hVDR through a double polymerase chain reaction.(11) The tryptophan at position 286 of hVDR was replaced by either an alanine or a phenylalanine to produce W286A and W286F, respectively.
Cell culture and transfections
COS-7 cells were grown in Dulbecco's modified Eagle's Medium (DMEM) (Gibco-BRL, Burlington, Ontario, Canada) supplemented with 10% fetal bovine serum (FBS). Transfections were performed by incubating 5 μg of each plasmid DNA with 15 μg lipofectAMINE (Gibco-BRL) for 20 h in serum-free OPTI-MEM (Gibco-BRL) and then replacing the medium with fresh DMEM containing 10% FBS.
Cellular extracts of COS-7 cells
Transiently transfected cells were harvested for gel retardation assays 48 h after transfection by washing cells twice with phosphate-buffered saline (PBS) followed by scraping the cells in 1 ml of PBS. Cells were centrifuged at 2000 rpm for 10 minutes at 4°C and the pellets were resuspended in 1 ml of buffer A (25 mM Tris, pH 7.5, 0.3 mM dithiothreitol [DTT], 0.1 M KCl, and 20% glycerol). Cells were lysed by three freeze-thaw cycles and then centrifuged at 11,500 rpm for 15 minutes at 4°C. Supernatants were stored at −80°C.
Nuclear extracts of COS-7 cells
Nuclear extracts were prepared from transiently transfected COS-7 cells as previously described.(12)
Western blot analysis
Western blotting was used to determine the cellular distribution of hVDR. Nuclear and cellular fractions from transiently transfected COS-7 cells (15 μg) were boiled for 5 minutes, resolved on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels and transferred to polyvinylidene difluoride (PVDF) membranes (BIO-RAD, Hercules, CA, USA). Protein content was determined with a protein assay kit (BIO-RAD). Blots were probed with a polyclonal anti-VDR antibody (diluted to 1:200; Affinity Bioreagents, Neshanic Station, NJ, USA). After incubation with a horseradish peroxidase-conjugated secondary antibody (BIO-RAD), the bands were visualized by enhanced chemiluminescence (ECL; Amersham).
Gel mobility shift analysis
Two micrograms of transiently transfected COS-7 cellular extracts were incubated for 20 minutes on ice in the presence or absence of 10−7 M 1,25(OH)2D3, and 1 μg polydeoxyinosine-deoxycytidilic acid (poly dI.dC) in a binding buffer (25 mM Tris-HCl, pH 8.0, 5% glycerol, and 0.5 mM DTT). Five femtomoles of the [3P]-labeled mouse osteopontin VDR element (mOP VDRE; 5′-GTACAAGGTTCACGAGGTTCACGTCTTA-3′) were added and incubated for 20 minutes at room temperature. The anti-RXR antibody 4×1D12 (a kind gift of Dr. P. Chambon, College de France, Illkirch, France) was then added where appropriate and further incubated for 20 minutes at room temperature. The samples were electrophoresed on 5% nondenaturing polyacrylamide gels, dried, and exposed to KODAK XAR-5 film (Eastman Kodak Company, Rochester, NY, USA).
Expression of glutathione-S-transferase fusion proteins in Escherichia coli
Glutathione-S-transferase (GST)-hRXRα, GST-hVDR, GST-hVDR W286A, and GST-hVDR W286F expression vectors were constructed by inserting a fragment encoding the LBDs of hRXRα, wild-type hVDR, hVDR W286A, or hVDR W286F into the EcoRI site of pGEX-2TK (Pharmacia, Baie d'Urfé, Quebec, Canada). The GST fusion proteins were expressed in E. coli. Bacteria transformed with the GST fusion protein expression plasmids were grown to an optical density at a wavelength of 595 nm (A595) of 0.5 at 37°C and expression was subsequently induced with 0.5 mM isopropyl β-D-thiogalactopyranoside. GST alone also was expressed in E. coli for control purposes. Cells were centrifuged and resuspended in 12 ml lysis buffer (50 mM Tris-HCl, pH 7.5, 100 mM KCl, 10 μM ZnCl2, 0.1 mM DTT, 10% glycerol, 100 mg/ml lysozyme, and protease inhibitors). The lysate was sonicated and pelleted by ultracentrifugation for 1 h at 30,000 rpm. Aliquots of the supernatants were stored at −80°C.
In vitro transcription and translation
Plasmids encoding wild-type VDR or the mutant hVDRs were transcribed with T7 polymerase and translated in vitro with a TNT T7-coupled reticulocyte lysate system (Promega, Madison, WI, USA) in the presence of [35S]methionine according to the manufacturer's instructions. Plasmids encoding steroid receptor coactivator 1 (SRC-1) were transcribed with T3 polymerase (Pharmacia) and translated in vitro as described previously. Translation efficiency was monitored by running 0.5 μl of the reaction on an sodium dodecyl sulfate (SDS)-polyacrylamide gel.
GST pull-down assays
GST fusion proteins were purified from 500 μl of bacterial lysate by incubation with 100 μl of glutathione-Sepharose 4B beads (Pharmacia) overnight at 4°C. The beads were washed three times with GST buffer (50 mM Tris-HCl, pH 7.5, 100 mM KCl, 10 mM MgCl2, 0.3 mM DTT, 5% glycerol, 0.1% NP-40, and protease inhibitors) and the protein content was determined by the Bradford method. Beads bound to GST fusion proteins or GST alone were incubated with the in vitro translated proteins in the presence or absence of 10−7 M 1,25(OH)2D3 at room temperature for 1 h. The beads were then washed three times with 400 μl GST buffer, boiled for 10 minutes, and electrophoresed on 10% SDS-polyacrylamide gels. Gels were then treated with Enlightening, an autoradiography enhancer (Mandel Scientific-NEN, St. Laurent, Quebec, Canada), for 20 minutes, dried, and exposed to KODAK XAR-5 film overnight.
Transcriptional activation assay
COS-7 cells were plated at a density of 4 × 104 cells/well into 6-well plates and transfected with 4 μg of the 1,25(OH)2D3-sensitive chloramphenicol acetyltransferase (CAT) reporter plasmid, mOP3,(13) 4 μg of either the wild-type or one of the mutant hVDR constructs, and 0.1 μg of a human growth hormone (hGH) reporter plasmid(14) as an internal control for transfection efficiency. Cells were incubated overnight in serum-free OPTI-MEM and then treated with varying concentrations of 1,25(OH)2D3 for 24 h. The medium was retained for the hGH assay and the cells were trypsinized washed in PBS, resuspended in 0.25 M Tris-HCl, pH 8.0, and lysed by five freeze-thaw cycles. The cell lysate was centrifuged and aliquots of cell extracts were used for the CAT assays. Assays were performed using a CAT ELISA kit (5 Prime-3 Prime, Boulder, CO, USA). The hGH assays were performed using an hGH ELISA kit (Boehringer Mannheim, Laval, Quebec, Canada).
Ligand binding assays
GST-VDR fusion proteins were expressed in bacteria and the fusion proteins were then purified and quantified as described previously. One microgram of purified protein was suspended in KTEDG buffer (10 mM Tris-HCl, pH 7.6, 1 mM EDTA, 300 mM KCl, 1 mM DTT, 0.1 mM phenylmethylsulfonyl fluoride [PMSF], and 1 Complete Mini protease inhibitor cocktail tablet [Boehringer Mannheim]) to a volume of 170 μl. The proteins were incubated with varying concentrations of [H3]-1,25(OH)2D3 (Mandel Scientific-NEN) in 25 μl and 5 μl of vehicle for total binding or 5 μl of nonlabeled 1,25(OH)2D3 for a final concentration of 250 nM for nonspecific binding for 16 h at 4°C. Proteins and bound hormone were pelleted and washed twice rapidly with KTEDG. The pellet was resuspended in 300 μl of KTEDG and 100 μl was assessed by liquid scintillation counting.
Cellular distribution of wild-type and mutant hVDRs expressed in COS-7 cells
Western blot analysis was performed using equal amounts of protein from both nuclear and cytosolic extracts prepared from COS-7 cells transfected with expression plasmids encoding either wild-type VDR, the W286A mutant, or the W286F mutant (Fig. 1). An antibody specific for the hVDR revealed a protein complex in the nuclear fraction of the COS-7 cell extracts from cells transfected with wild-type VDR and from cells transfected with the mutant VDRs (Fig. 1, lanes 1, 3, and 5). No protein complex corresponding to the VDR was detected in the cytosolic fractions extracted from the transfected COS-7 cells (Fig. 1, lanes 2, 4, and 6), indicating that mutation of W286 does not affect nuclear localization of the receptor.
1,25(OH)2D3 ligand binding in vitro
Saturation binding revealed a single class of binding sites (Fig. 2A). Wild-type hVDR had a Kd of 1.536 ± 0.087 nM. Mutation of tryptophan 286 of the hVDR to an alanine completely abolished the capacity of the receptor to bind to 1,25(OH)2D3 while mutation of tryptophan 286 to phenylalanine did not considerably interfere with ligand binding, giving a Kd of 1.216 ± 0.110 nM (Fig. 2B).
Interaction of wild-type and mutant VDRs with hRXRα in vitro
Heterodimerization of wild type and mutant VDR LBDs was examined by in vitro translation (Fig. 3, lanes 1–3) and incubation with GST-hRXRα fusion proteins in the presence or absence of 10−7 M 1,25(OH)2D3. The W286F mutant displayed reduced levels of ligand-dependent interaction with the GST-hRXRα fusion protein (Fig. 3, lane 12) as compared with the wild-type VDR (Fig. 3, lane 6). In contrast, the W286A mutant did not interact with hRXRα in the presence of 1,25(OH)2D3 above basal levels (Fig. 3, lanes 8 and 9). Each of the in vitro translated VDR proteins that were incubated with GST alone as a control did not show any significant interaction (Fig. 3, lanes 4, 7, and 10).
Heterodimerization was analyzed further in gel mobility shift assays with full-length VDR. Under these conditions, DNA binding domain dimerization on a VDRE would be expected to stabilize any weak interaction between hRXRα and hVDR. Cellular extracts were prepared from COS-7 cells transfected with hRXRα and wild-type or mutant VDRs, and formation of specific VDR/hRXRα complexes was monitored on an mOP VDRE. The wild-type VDR (Fig. 4, lanes 1–3) and the W286F mutant (Fig. 4, lanes 7-9) formed complexes with hRXRα on the mOP VDRE, which could be enhanced to similar degrees by addition of 10−7M 1,25(OH)2D3 and supershifted with an anti-RXR antibody (Fig. 4, lanes 3 and 9). However, the extracts from COS-7 cells transfected with hRXRα and the W286A mutant did not respond to 1,25(OH)2D3 as indicated by the absence of enhanced complex formation in the presence of ligand (Fig. 4, lane 5), thus indicating that the W286A mutant is completely defective for heterodimerization with hRXRα.
Interaction of the mutant VDRs with coactivators in vitro
GST fusion proteins of the wild-type and mutant VDR LBDs were incubated with in vitro translated SRC-1 in the presence or absence of 10−7 M 1,25(OH)2D3 (Fig. 5). The wild-type VDR and W286F mutant interacted with SRC-1 in a ligand-dependent manner (Fig. 5, lanes 5 and 9, respectively). In contrast, the W286A mutant was unable to bind SRC-1 (Fig. 5, lane 7).
Ligand-dependent transactivation in COS-7 cells transfected with wild-type or mutant VDRs
Transcriptional activation in response to increasing concentrations of 1,25(OH)2D3 (10−10M-10−7M) was examined by cotransfecting into COS-7 cell plasmids expressing wild-type hVDR or the mutants W286A or W286F and the mOP3 reporter construct, which contains three repeats of the mOP VDRE upstream of a CAT reporter gene.(14) A reporter plasmid encoding hGH was used as an internal control for transfection efficiency.(15) Results were expressed as a percent of vehicle-treated control values. The ligand-dependent transcriptional activation in cells transfected with the W286F receptor was attenuated severely whereas no transactivation was observed in cells transfected with the W286A mutant (Fig. 6).
This study shows that tryptophan 286 of the hVDR is critical for both high affinity binding of 1,25(OH)2D3 to the receptor and for subsequent heterodimerization and gene transactivation. W286 mutants heterodimerized less efficiently with RXRα in GST pull-down assays (Fig. 3). In addition, mutation of W286 abolished transactivation in gene transfer experiments and in the case of the alanine mutation eliminated the ligand-dependent interaction of the receptor with the coactivator SRC-1 in GST pull-down assays ( Figs. 5 and 6). Nuclear fractions of COS-7 cells transfected with wild-type, W286A, or W286F mutants contained similar amounts of receptor (Fig. 1). Thus, mutation of W286 did not affect hVDR expression levels or nuclear localization of the hVDR.
The substitution of W286A was increasingly more disruptive than the more conservative W286F mutation. The W286A mutant was completely defective for ligand binding, heterodimerization, interaction with SRC-1 in vitro, and transactivation. The W286F receptor retained its ability to interact with 1,25(OH)2D3 in ligand binding assays (Fig. 2). However, we did detect slightly attenuated heterodimerization of W286F with RXRα in GST pull-down assays and unaffected heterodimerization with RXRα in gel mobility shift assays ( Figs. 3 and 4). Taken together, mutation of the large, aromatic tryptophan residue to alanine eliminates ligand binding while mutation to the more conservative large, aromatic residue phenylalanine eliminates transactivation. Therefore, tryptophan 286 is critical for both ligand binding and transactivation.
The crystal structure of the hormone binding domain of RXRα(3) showed that ligand interacts with a hydrophobic pocket formed by helices 5 and 7, the C terminus of helix 10 and the N terminus of helix 11, and β-strands.(16) W286 is positioned in an area corresponding to the β-strand region of the VDR LBD. Indeed, it recently has been found that W286 directly contacts bound 1,25(OH)2D3 via packing interactions between the aromatic ring of tryptophan and the A-ring of 1,25(OH)2D3 (Dino Moras, personal communication, 1999). This is consistent with our observation that replacement of tryptophan 286 with a nonaromatic amino acid such as alanine completely abolishes ligand binding.
Both the conservative phenylalanine and the more dramatic alanine substitution completely eliminated transactivation by the VDR, even at high ligand concentrations. Interaction of the W286A mutant with SRC-1 was eliminated, consistent with effects of this mutation on ligand binding and thus ligand-dependent movement of the AF-2 domain from an inactive to an active conformation. However, significant interaction of the W286F receptor with SRC-1 was observed (Fig. 5). This, along with the observation of ligand-dependent heterodimerization of the W286F mutant, suggests that it can form activated DNA-bound heterodimers at high concentrations in vitro. Despite the ligand-inducible interaction of W286F with SRC-1, it is possible that this mutant is unable to bind other important coactivators for hVDR. This may explain the lack of ligand-dependent transactivation observed with this receptor mutant and implies that W286 may be essential for interaction of the VDR with components of the transcription machinery in addition to its role in the stabilization of ligand binding.
The vitamin D receptor mediates the bone mineral homeostatic actions of 1,25(OH)2D3 and therefore natural mutations in the VDR often result in hereditary hypocalcemic vitamin D-resistant rickets (HVDRR).(17) Several mutations have been characterized that are located within the LBD of the VDR and result in the HVDRR phenotype including R271L(18) and H305Q.(19) In the VDR, R274L is located in helix 5, H305Q is in the loop between helix 6 and helix 7, and W286 is positioned in the β-turn between helix 5 and helix 6, all at important interfaces for ligand binding. Mutagenesis studies with the VDR indicate that the LBD is well aligned with those in the hRXRα and hRARγ.(20) Tryptophan is critical for ligand binding and transactivation in the hRARγ,(6) hRXRα,(8) and hVDR, and conservation of the unique tryptophan in the LBD of other members of the nuclear receptor superfamily provides insight concerning the putative structure of the VDR with respect to the hRXRα and hRARγ. Furthermore, mutation of cysteine 288 (C288G) was shown to reduce the affinity of the receptor for 1,25(OH)2D3.(16) Table 1 summarizes the effect of natural and synthetic mutations characterized within the LBD of the VDR thus far on ligand binding and transactivation. Tryptophan 286 is the only amino acid shown absolutely to be required for 1,25(OH)2D3 interaction with hVDR and transactivation.
Table Table 1.. Effect of Natural and Synthetic Mutations in the hVDR on Ligand Binding, Transactivation, and Heterodimerization with hRXRα
In summary, our experiments have shown that tryptophan 286 is essential for 1,25(OH)2D3 interaction with hVDR. Mutation of tryptophan 286 to alanine completely abolishes the ability of the receptor to interact with 1,25(OH)2D3. This, in turn, affects the ligand-activated functions of the receptor including heterodimerization, interaction with coregulatory proteins, and transactivation.
We are grateful to M. Gratton, I. Bolivar, and S.S. Solomon for technical assistance and to Dino Moras (Illkirch, France) for communicating crystallographic data before publication. R.K. is a recipient of Medical Research Council of Canada grant MT-10839. J.H.W. is a recipient of Medical Research Council of Canada grant MT-11704 and is a chercheur-boursier of the Fonds de Recherche en Santé du Québec.