Hereditary 1,25-Dihydroxyvitamin D Resistant Rickets due to a ...

0013-7227/04/$15.00/0 Printed in U.S.A.

Endocrinology 145(11):5106 –5114 Copyright © 2004 by The Endocrine Society doi: 10.1210/en.2004-0080

Hereditary 1,25-Dihydroxyvitamin D Resistant Rickets due to a Mutation Causing Multiple Defects in Vitamin D Receptor Function PETER J. MALLOY, RONG XU, LIHONG PENG, SARA PELEG, ABDULLAH AL-ASHWAL, DAVID FELDMAN

AND

Department of Medicine, Stanford University School of Medicine (P.J.M., R.X., L.P., D.F.), Stanford, California 94305; Department of Endocrine Neoplasia and Hormonal Disorders, University of Texas, M. D. Anderson Cancer Center (S.P.), Houston, Texas 77030; and Department of Pediatrics, King Faisal Specialist Hospital and Research Center (A.A.-A.), Riyadh, Saudi Arabia Hereditary vitamin D-resistant rickets (HVDRR) is an autosomal recessive disease caused by mutations in the vitamin D receptor (VDR). We studied a young Saudi Arabian girl who exhibited the typical clinical features of HVDRR, but without alopecia. Analysis of her VDR gene revealed a homozygous T to C mutation in exon 7 that changed isoleucine to threonine at amino acid 268 (I268T). From crystallographic studies of the VDR ligand-binding domain, I268 directly interacts with 1,25dihydroxyvitamin D3 [1,25(OH)2D3] and is involved in the hydrophobic stabilization of helix H12. We recreated the I268T mutation and analyzed its effects on VDR function. In ligand binding assays, the I268T mutant VDR exhibited an approximately 5- to 10-fold lower affinity for [3H]1,25(OH)2D3 compared with the wild-type (WT) VDR. The I268T mutant required approximately a 65-fold higher concentration of 1,25(OH)2D3 to be equipotent in gene transactivation. Both

T

HE BIOLOGICAL ACTIONS of 1,25-dihydroxyvitamin D3 [1,25(OH)2D3], including regulation of calcium homeostasis, cellular differentiation, and immune function, are mediated by the vitamin D receptor (VDR), a member of the steroid/nuclear receptor superfamily of ligand-activated transcription factors (1–3). Binding of 1,25(OH)2D3 to the VDR ligand-binding domain (LBD) triggers a series of molecular events leading to the activation of vitamin D-responsive genes. A two-zinc finger structure in the DNA-binding domain (DBD) near the N terminus of the VDR allows the VDR to bind to vitamin D response elements (VDRE) in the promoter regions of target genes. The regulation of gene transcription by the VDR requires binding to VDREs as a heterodimer with the retinoid X receptor (RXR). Initiation of gene transcription also involves the recruitment of coactivator proteins, including the p160 proteins steroid receptor coactivator 1 (SRC1), glucocorticoid receptor interacting protein 1 (GRIP1), and amplified in breast cancer 1 (AIB1) and

Abbreviations: DBD, DNA-binding domain; DRIP, VDR interacting protein; GST, glutathione-S-transferase; HVDRR, hereditary vitamin Dresistant rickets; LBD, ligand-binding domain; 1,25(OH)2D3, 1,25-dihydroxyvitamin D3; 25(OH)D, 25-hydroxyvitamin D; RXR, retinoid X receptor; SRC-1 steroid receptor coactivator 1; VDR, vitamin D receptor; VDRE, vitamin D response element; WT, wild type. Endocrinology is published monthly by The Endocrine Society (http:// www.endo-society.org), the foremost professional society serving the endocrine community.

retinoid X receptor heterodimerization and coactivator binding were reduced in the I268T mutant. Analogs of 1,25(OH)2D3 have been proposed as potential therapeutics for patients with HVDRR. Interestingly, in protease sensitivity assays, treatment with the potent vitamin D analog, 20epi-1,25(OH)2D3, stabilized I268T mutant proteolytic fragments better than 1,25(OH)2D3. Moreover, 20-epi-1,25(OH)2D3 restored transactivation of the I268T mutant to levels exhibited by WT VDR treated with 1,25(OH)2D3. In conclusion, we describe a novel mutation, I268T, in the VDR ligand-binding domain that alters ligand binding, retinoid X receptor heterodimerization, and coactivator binding. These combined defects in VDR function cause resistance to 1,25(OH)2D3 action and result in the syndrome of HVDRR. (Endocrinology 145: 5106 –5114, 2004)

the VDR interacting protein (DRIP)/mediator complex. These coactivator proteins remodel chromatin and act as bridging factors, linking the nuclear receptors to the preinitiation complexes and RNA polymerase II. They bind to the VDR in a ligand-dependent manner and enhance transactivation. A crystallographic study of the VDR LBD bound to 1,25(OH)2D3 has shown that the LBD is composed of 13 ␣-helices and three ␤-sheets (4). Ligand binding causes helix H12 in the activation function-2 domain to be repositioned forming a hydrophobic groove with helices H3 and H4. The coactivators bind to the hydrophobic groove on the nuclear receptor surfaces through a conserved LXXLL motif in the nuclear receptor-interacting domains. Hereditary vitamin D-resistant rickets (HVDRR; on line Mendelian Inheritance in Man 277440), also known as vitamin D-dependent rickets type II, is a rare recessive genetic disorder caused by mutations in the VDR. Patients with HVDRR exhibit early-onset rickets, hypocalcemia, and secondary hyperparathyroidism and generally have a history of consanguinity in the family. HVDRR patients also have significantly elevated serum levels of 1,25(OH)2D3, distinguishing this condition from 1␣-hydroxylase deficiency, also known as vitamin D-dependent rickets type I. The latter patients present with reduced levels of 1,25(OH)2D3. In some cases, HVDRR patients have total body alopecia. A number of genetic abnormalities in the VDR causing HVDRR have

5106

Malloy et al. • Vitamin D Resistance

been described (2). Mutations in the VDR DBD interfere with VDR-DNA interactions, but not with ligand binding, and result in complete loss of transactivation (5–9). Mutations in the VDR LBD affect 1,25(OH)2D3 binding, heterodimerization with RXR and coactivator interaction and result in partial or total hormone unresponsiveness (10 –14). Some patients with LBD mutations do not have alopecia. There has also been one report of a patient with HVDRR and alopecia whose fibroblasts were resistant to 1,25(OH)2D3 without a mutation in the VDR (15). The 1,25(OH)2D3 resistance in this patient has been postulated to be due to the overabundance of a hormone response element-binding protein that occupies VDREs and prevents VDR transactivation (16). In this report we describe a novel mutation in helix H5 in the VDR LBD that affects ligand binding, RXR heterodimerization, and coactivator binding and results in the syndrome of HVDRR without alopecia. Materials and Methods [3H]1,25(OH)2D3 binding and Western blotting The wild-type (WT) and mutant VDRs were expressed in COS-7 cells. Cell extracts were prepared in M-PER extraction buffer (Pierce Chemical Co., Rockford, IL) containing 300 mm KCl, 5 mm dithiothreitol, and a complete protease inhibitor tablet (Roche, Indianapolis, IN). Cells were incubated at ambient temperature for 10 min on a rotating mixer. Cell debris was removed by centrifugation at 13,000 rpm for 15 min at 4 C. The crude cell extracts were incubated with [3H]1,25(OH)2D3 (Amersham Biosciences, Arlington Heights, IL) with or without a 250-fold excess of radioinert 1,25(OH)2D3 as previously described (17). Hydroxylapatite was used to separate bound and free hormone. For Western blotting, cell extracts were denatured in lithium dodecyl sulfate sample buffer for 10 min at 70 C and electrophoresed on 10% NuPAGE gels in 4-morpholinepropanesulfonic acid-sodium dodecyl sulfate running buffer (Invitrogen Life Technologies, Inc., Carlsbad, CA). Proteins were transferred to nitrocellulose and incubated with a mouse anti-VDR monoclonal antibody (DSL Laboratories, Webster, TX) (18) as previously described (19). Protein concentrations were determined by the Bradford method (20).

Gene amplification and DNA sequencing Exons 2–9 of the VDR gene that encode the VDR protein were amplified by PCR and cloned into pBluescript II KS (Stratagene, La Jolla, CA). The PCR products were cloned and sequenced using [␣-35S]deoxyATP and Sequenase version 2.0 DNA sequencing kit (Amersham Biosciences).

RFLP analysis Exons 7 and 8 of the VDR gene were amplified from the patient’s DNA using oligonucleotide primers 7a and 8b as previously described (17). PCR products were digested with the restriction endonuclease DdeI at 37 C according to the manufacturer’s directions (New England Biolabs, Beverly, MA) and were analyzed on 2% agarose gels in Tris-acetic acid-EDTA buffer. Gels were stained with ethidium bromide, visualized by UV light, and photographed.

Site-directed mutagenesis and plasmid construction Site-directed mutagenesis of the WT VDR cDNA was performed using the pALTER system (Promega Corp., Madison, WI) as previously described (12). The mutant oligonucleotide used was 5⬘-TCA AGT GCC ACT GAG GTC ATC. Clones were sequenced to confirm the presence of the point mutation. The mutant VDR cDNA was subcloned into the mammalian expression vector pSG5 (Stratagene).

Transactivation assays COS-7 cells were grown to 60 – 80% confluence in 12-well tissue culture plates. Cells were transfected with 0.125 ␮g WT or I268T mutant

Endocrinology, November 2004, 145(11):5106 –5114 5107

VDR expression plasmid and 0.25 ␮g rat 24-hydroxylase promoter VDRE-luciferase plasmid using Polyfect (Qiagen, Valencia, CA). A Renilla-luciferase plasmid, pRLnull (0.01 ␮g), that served as an internal control for transfection efficiency was included in each assay. A human RXR␣ expression vector was cotransfected in some experiments. After a 16-h transfection, the cells were incubated in DMEM containing 1% fetal bovine serum with or without 1,25(OH)2D3. In some assays cells were treated with 20-epi-1,25(OH)2D3 (MC1288, Leo Pharmaceuticals, Ballerup, Denmark). Twenty-four hours after transfection, the cells were washed and prepared for dual luciferase assays according to the manufacturer’s instructions (Promega Corp.). Luciferase activities were determined using a luminometer (Turner Design, Sunnyvale, CA).

Glutathione-S-transferase (GST) pull-down GST fusion proteins (RXR␣, SRC-1, and DRIP205) were expressed in Escherichia coli BL21(DE3) after induction with 0.1 mm isopropyl-␤-dthiogalactoside for 3 h at 37 C. Proteins were extracted by incubating the cells in B-PER extraction reagent (Pierce Chemical Co.) containing 100 mm KCl, 10 mm MgCl2, 1 mm dithiothreitol, 10% glycerol, and a complete protease inhibitor tablet (one tablet per 50 ml) (Roche Molecular Biochemicals) for 10 min at ambient temperature with gentle shaking. Cell debris was removed by centrifugation at 12,500 ⫻ g for 20 min at 4 C. WT and I268T mutant VDR were labeled with [35S]methionine (Amersham Biosciences) by in vitro transcription/translation using the TNT-coupled system (Promega Corp.). For binding assays, E. coli extracts containing GST fusion proteins were mixed with glutathione agarose at 4 C for 16 h and then washed. 35S-Labeled VDRs and 1,25(OH)2D3 were added to the beads in GST binding buffer [50 mm Tris buffer (pH 7.5) containing 100 mm KCl, 10 mm MgCl2, 0.3 mm dithiothreitol, 0.1% Nonidet P-40, and 10% glycerol] and incubated at 4 C for 16 h. The agarose beads were then washed three times with binding buffer. Bound proteins were eluted in 2⫻ lithium dodecyl sulfate sample buffer, heated at 70 C for 5 min, and electrophoresed on 10% sodium dodecyl sulfate-NuPAGE gels (Invitrogen Life Technologies, Inc.). Gels were fixed in 50% methanol/10% acetic acid for 10 min, incubated in Amplify (Amersham Biosciences) for 15 min, dried, and exposed to Hyperfilm (Amersham Biosciences) at ⫺80 C. Nonspecific binding was determined using extracts containing GST alone.

Yeast two-hybrid assay Plasmids pAS2-WT VDR, pAS2-I268T mutant, and pGAD-RXR␣ were transformed into Saccharomyces cerevisiae strains CG-1945 or Y187 cells using FROZEN-EZ yeast transformation kit (ZYMO Research, Orange, CA). Diploid colonies were obtained by mating CG1945 with Y187 cells on medium lacking leucine, tryptophan, and histidine and containing 5 mm 3-aminotriazole. Clones were grown overnight at 30 C in selective medium, then diluted 1:20 in fresh medium with or without 1,25(OH)2D3 and grown for an additional 16 –24 h. ␤-Galactosidase activity was determined using the yeast ␤-galactosidase assay kit (Pierce Chemical Co.). One unit of ␤-galactosidase is defined as the amount of enzyme that hydrolyzes 1 ␮mol O-nitrophenyl ␤-d-galactopyranoside /min at 30 C. ␤-Galactosidase activity was normalized to the cell culture density determined by measuring the OD600 at the time of harvest.

DNA binding DNA binding was assessed by gel mobility shift assays. The human osteopontin VDRE was end-labeled using [␥-32P]ATP and polynucleotide kinase. WT and I268T VDRs were expressed in COS-7 cells. Cell extracts were prepared by incubating the cell pellet in M-PER extraction buffer containing 300 mm KCl, 5 mm dithiothreitol, and a protease inhibitor tablet for 10 min. Cell extracts were incubated with vehicle (ethanol) or 10 nm 1,25(OH)2D3 in buffer containing 0.25 ␮g/ml polydeoxyinosinic-deoxycytidylic acid for 20 min. The 32P-labeled osteopontin-VDRE probe was added for an additional 20 min. All incubations were performed at ambient temperature. The final concentration of salt in the binding assay was 150 mm KCl. For supershift assays, a monoclonal antibody against the C terminus of the VDR (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was added before addition of the probe. The samples were then electrophoresed on 5% polyacrylamide gels (acrylamide/bis-acrylamide, 29:1) in 0.5⫻ Tris-borate buffer at 180

5108

Endocrinology, November 2004, 145(11):5106 –5114

V for 90 min at ambient temperature. The gel was then dried and subjected to autoradiography at ⫺80 C.

Protease sensitivity assays WT and mutant VDRs were synthesized and labeled in vitro with [35S]methionine (1000 Ci/mmol) using the TNT-coupled transcription/ translation system (Promega Corp.). The translated receptor preparations were incubated with 1,25(OH)2D3 or 20-epi-1,25(OH)2D3 for 10 min at ambient temperature. Next, trypsin was added to a concentration of 20 ␮g/ml, and the mixtures were incubated for 10 min. The digestion products were then separated by SDS-PAGE and detected by autoradiography.

Results Case history

Informed consent for this study was obtained from the patient’s parents using Stanford University institutional review board-approved protocols. The parents declined to provide their own DNA samples and would not consent to a skin biopsy of the affected child. The patient, a young Saudi girl, first presented to her local hospital at age 11 months with evidence of severe rickets, inability to walk, multiple fractures, and recurrent pulmonary infections. Her family has a history of consanguinity. The parents did not have bone disease; however, two older children died at ages 5 and 6 yr of a condition similar to that of the younger girl. None of the affected children had alopecia. The patient’s blood tests at that time showed a calcium of 1.89 mmol/liter (normal, 2.1–2.6), phosphate of 1.4 mmol/liter (normal, 1–1.5), and alkaline phosphatase of 5080 U/liter (normal, 60 –300). Her 1,25(OH)2D level was 168 pmol/liter (normal, 39 –102), and PTH was 368 pg/ml (normal, 10 – 65). X-rays showed severe rickets, chest deformities, and evidence of restrictive lung disease. She was treated with oral calcitriol, calcium, and phosphate and discharged. At 20 months of age she was severely ill and presented to the local hospital with severe osteopenia, inability to walk, and fractures. She had chest deformities, restrictive lung disease requiring oxygen, and failure to thrive. On admission at that time her calcium was 1.5 mmol/liter, phosphate was 0.9 mmol/liter, alkaline phosphatase was 1570 U/liter, 25hydroxyvitamin D [25(OH)D] was less than 9 nmol/liter (normal, 22–116), indicating severe vitamin D deficiency, 1,25-(OH)2D was 150 pmol/liter (normal, 39 –102), and PTH was 840 pg/ml. The patient received a total of 1.2 million IU vitamin D by injection over 12 wk that brought her 25(OH)D levels almost to normal range. However, she did not show clinical improvement with normalization of 25(OH)D and oral calcium. She received up to 10 ␮g calcitriol twice daily with no response. When she did not improve on oral calcium and calcitriol, she was started on high dose calcium iv infusions of 1000 mg/m2 nightly plus total parenteral nutrition. Over the next 4 months she showed remarkable improvement in her rickets and symptoms. Her calcium rose to 2.47, phosphate to 1.42, alkaline phosphatase fell to 286, PTH to 84 pg/ml, 25(OH)D rose to 200 nmol/liter, and 1,25(OH)2D to 490 pmol/liter. The x-rays reverted toward normal. After repeated infections from her indwelling catheter, the iv therapy was discontinued, and she was discharged from

Malloy et al. • Vitamin D Resistance

the hospital on oral calcium (4 g elemental calcium daily and phosphate). However, after several months off iv infusions, her rickets began to recur. Her PTH rose to 620, alkaline phosphatase to 1136, and x-rays again showed osteopenia. She has been readmitted for repeat iv calcium infusion therapy every 6 – 8 wk as signs of rickets recur when the infusion is discontinued, heralded by rising alkaline phosphatase and PTH levels. DNA sequence analysis and genotype

In our initial studies we examined the patient’s DNA for mutations in the VDR gene and identified a novel T to C missense mutation in exon 7. The mutation occurs in helix H5 in the VDR LBD and changes isoleucine to threonine at amino acid 268 (I268T; Fig. 1). The mutation generated a unique DdeI restriction site in exon 7. We amplified exons 7– 8 by PCR and used DdeI to determine whether the patient was homozygous for the mutation. As shown in Fig. 2, DdeI digestion of exons 7– 8 from a normal control showed two bands of 306 and 180 bp (the smaller size fragments are not visible on the figure). DdeI digestion of exons 7– 8 from the patient showed a smaller size fragment of 271 bp and the 180-bp fragment consistent with the presence of a new DdeI site. Because the normal WT fragment is not present, the data demonstrate that the patient is homozygous for the mutation. DNA samples from the parents or siblings were not available for analysis. We recreated the I268T mutation in the WT VDR cDNA by site-directed mutagenesis and expressed the mutant VDR cDNA in COS-7 cells. We then examined the effects of the

FIG. 1. Sequence analysis of the patient’s VDR uncovers a missense mutation in exon 7. Exon 7 was amplified from the patient’s DNA by PCR and cloned into pBluescript II KS⫹. A, Sequence of the sense strand with the T to C mutation indicated. B, The mutation changed isoleucine to threonine at amino acid 268 (I268T). C, Schematic representation of the VDR LBD and location of the I268T mutation. ␣ Helices are shown as rectangles, with the helix number inside.

Malloy et al. • Vitamin D Resistance

FIG. 2. The patient is homozygous for the missense mutation in exon 7. A, Exons 7 and 8 were amplified from genomic DNA using primers VDR7U and VDR8L to generate a 567-bp fragment. The T to C base change added a new DdeI restriction site that is indicated by an asterisk. B, DdeI digestion products were electrophoresed on 2% agarose 1000 gels and visualized with ethidium bromide stain. Lanes 1–3 symbols: 100, 100-bp marker; N, normal control; Pt, patient. The patient shows only the 271-bp fragment and no 306-bp fragment seen in the WT, indicating that she is homozygous for the I268T mutation.

I268T mutation on VDR function to determine whether the mutation is the molecular cause of the 1,25(OH)2D3 resistance in this patient.

Endocrinology, November 2004, 145(11):5106 –5114 5109

FIG. 3. Reduced transactivation activity of the I268T mutant VDR by 1,25(OH)2D3. COS-7 cells were transiently transfected with I268T mutant, WT VDR expression plasmids, or pSG5 empty vector, a 24hydroxylase promoter-luciferase reporter plasmid, and a Renillaluciferase reporter plasmid that serves as an internal control. COS-7 cells were treated with 1,25(OH)2D3 as indicated. Error bars represent ⫾SD.

mined the effects of the mutation on ligand binding using [3H]1,25(OH)2D3 binding assays and coactivator binding using GST pull-down assays. Figure 5 shows the result of a representative Scatchard analysis of the ligand binding data. The I268T mutant VDR bound [3H]1,25(OH)2D3 with a dissociation constant (Kd) of 4 ⫻ 10⫺9 m, an approximately 5-fold lower affinity than the WT VDR Kd of 8 ⫻ 10⫺10 m assayed at the same time. Coactivator binding

Transactivation

Because the mutation was expected to cause a loss or reduction in VDR function, we first determined the effects of the I268T mutation on VDR transactivation. As shown in Fig. 3, in transactivation assays the I268T mutant VDR required approximately 65-fold more 1,25(OH)2D3 to induce gene transcription from a 24-hydroxylase promoter reporter compared with the WT VDR (average EC50, ⬃4.5 nm for I268T mutant vs. ⬃0.07 nm for WT VDR). Western blot confirmed equal expression of the VDR proteins (data not shown). These results demonstrate that the I268T mutation causes 1,25(OH)2D3 resistance that can be overcome with very high concentrations of ligand. Ligand binding

From the crystallographic studies of the holo-VDR, I268 has been postulated to interact with 1,25(OH)2D3 and be involved in the repositioning of helix H12 for coactivator interaction (4). A model depicting the location of the amino acids contacting 1,25(OH)2D3 is shown in Fig. 4. I268 forms van der Walls interactions with C22 in the side-chain of 1,25(OH)2D3. Also, I268 forms hydrophobic interactions with residues in helix H12. Based on these data, we next deter-

We also determined whether coactivator binding to the VDR was compromised by the I268T mutation. As shown in Fig. 6, in GST pull-down assays the WT VDR bound SRC-1 and DRIP205 in a ligand-dependent manner. In contrast, the I268T mutant VDR exhibited reduced binding to both SRC-1 and DRIP205. At 1–10 nm 1,25(OH)2D3, the I268T mutant VDR was unable to bind to SRC-1 or DRIP205, in contrast to the WT VDR that bound these coactivators at these 1,25(OH)2D3 concentrations. However, as the 1,25(OH)2D3 concentration was raised to 100 nm, the mutant receptor was able to bind the coactivators. The I268T mutant VDR required approximately 100-fold higher concentrations of 1,25(OH)2D3 for binding to occur compared with the WT VDR. These results coupled with the transactivation data demonstrate that the I268T mutation causes a defect in coactivator binding that reduces the receptor’s ability to activate gene transcription in addition to its effects on ligand binding. RXR heterodimerization

The I268T mutation is also located near the E1 domain (amino acids 244 –263), which has been shown to be involved in VDR-RXR heterodimerization (21, 22). To determine

5110

Endocrinology, November 2004, 145(11):5106 –5114

FIG. 4. Model of 1,25(OH)2D3 bound to the VDR LBD. The amino acid residues contacting the ligand are shown. The dashed lines indicate hydrogen bonding. Arrow points to I268. From D. Moras, personal communication.

FIG. 5. The I268T mutant VDR has a decreased binding affinity for binding [3H]1,25(OH)2D3. High salt extracts from COS-7 cells expressing the I268T mutant VDR were incubated with increasing concentrations of [3H]1,25(OH)2D3 with and without a 250-fold excess of radioinert 1,25(OH)2D3 for 24 h at 0 C. Bound and free hormones were then separated by hydroxyapatite. Scatchard plots are shown. The dissociation constants (Kd) were 0.08 nM for the WT VDR and 0.4 nM for the I268T VDR.

Malloy et al. • Vitamin D Resistance

FIG. 6. The I268T mutant VDR requires higher concentrations of 1,25(OH)2D3 to bind the coactivators SRC-1 and DRIP205. In vitro 35 S-labeled I268T mutant and WT VDRs were incubated with GSTSRC-1 or GST-DRIP205 bound to glutathione agarose beads and increasing concentrations of 1,25(OH)2D3. After washing, the bound proteins were eluted and subjected to SDS-PAGE and autoradiography. A, WT VDR and I268T mutant VDR bound to GST-SRC-1; B, WT VDR and I268T mutant VDR bound to GST-DRIP205.

whether the I268T mutation affects VDR-RXR␣ heterodimerization, we used GST pull-down and yeast two-hybrid assays. As shown in the GST pull-down assay, WT VDR binds RXR in the absence of ligand and shows an increase in RXR binding concomitant with increasing concentrations of 1,25(OH)2D3 (Fig. 7). The I268T mutant VDR also binds RXR in the absence of 1,25(OH)2D3; however, RXR binding did not increase with incremental ligand until the ligand concentration was raised to 10 nm 1,25(OH)2D3. We also examined VDR-RXR interaction using the yeast two-hybrid assay (23). In these assays the I268T mutant VDR consistently bound RXR slightly better than the WT VDR in the absence of 1,25(OH)2D3 (Fig. 7). However, as the concentration of 1,25(OH)2D3 was increased, the I268T mutant did not show an increase in RXR binding until the concentration was greater than 10 nm 1,25(OH)2D3. In addition, the I268T mutant VDR was not able to achieve the maximal activity that the WT VDR exhibited at the concentrations tested. As a control we also included an F251C mutant VDR that we have previously shown to have reduced affinity for 1,25(OH)2D3 and a significant defect in RXR binding even at high concentrations of 1,25(OH)2D3 (13). We also determined whether the transactivation function of the I268T mutant VDR could be restored by the addition of excess RXR. As shown in Fig. 7, coexpression of RXR slightly augments the transactivation activity of the mutant VDR; however, it did not shift the dose-response curve to WT levels.

Malloy et al. • Vitamin D Resistance

Endocrinology, November 2004, 145(11):5106 –5114 5111

FIG. 7. The I268T mutant VDR binds RXR␣. A, Analysis of VDR-RXR binding using GST pull-down assays. In vitro 35Slabeled I268T mutant and WT VDRs were incubated with GST-RXR␣ bound to glutathione agarose beads and increasing concentrations of 1,25(OH)2D3. After washing, the bound proteins were eluted and subjected to SDS-PAGE and autoradiography. B, Analysis of VDR-RXR binding using the yeast two-hybrid system. Diploid strains were obtained by mating S. cerevisiae CG-1945 containing pAS2VDRs and S. cerevisiae Y187 containing pGAD-RXR␣. The diploid cells were grown overnight at 30 C with or without 1,25(OH)2D3. VDR-RXR interactions were determined by analyzing ␤-galactosidase activity. An F251C mutant VDR defective in RXR binding was used as a control. Error bars represent ⫾SD. C, Coexpression of RXR␣ increases the transactivation activity of the I268T mutant VDR. VDR and RXR␣ expression plasmids were cotransfected into COS-7 cells and incubated with increasing concentrations of 1,25(OH)2D3. Error bars represent ⫾SD.

DNA binding

Because the I268T mutant VDR was defective in heterodimerization, we examined whether the I268T mutation also affected DNA binding. As shown in Fig. 8, in gel mobility shift assays, no band shifts were observed in the absence of VDR (pSG5 vector alone), in the absence of ligand, or in the presence of 0.1 nm 1,25(OH)2D3 in either WT VDR or I268T mutant VDR. However, both WT and I268T mutant exhibited VDR complex formation as the concentration of ligand increased from 1 to 100 nm 1,25(OH)2D3. Addition of an anti-VDR antibody caused the complexes to supershift, demonstrating the presence of the VDR in the complex. These results indicate that the I268T mutation does not alter the DNA binding function of the mutant receptor. Sensitivity to proteolytic digestion

We have previously shown that the ability of a ligand to stabilize WT or mutant VDR against proteolytic digestion by trypsin may correlate better with the VDR’s ability to interact with coactivators in vitro and better with its transcriptional activity in cells than with its affinity for VDR (24). Furthermore, we have shown that potent vitamin D analogs may restore stabilization against proteolytic digestion and transcriptional activity of VDRs from HVDRR patients with mutations in the LBD that affect ligand binding (24). The results described above demonstrate that the decrease in affinity of

FIG. 8. The I268T mutant VDR binds to the osteopontin VDRE. COS-7 cells extracts containing I268T mutant and WT VDRs were incubated with 32P-labeled osteopontin VDRE with and without 1,25(OH)2D3. The bound complexes were then resolved on nondenaturing gels. In some samples, a monoclonal antibody against the VDR was added to supershift the complex.

the I286T VDR (5-fold) for 1,25(OH)2D3 is less remarkable than its decrease in ability to be transactivated or to interact with p160 coactivator in a hormone-dependent manner (a

5112

Endocrinology, November 2004, 145(11):5106 –5114

100-fold decrease). Therefore, we investigated whether the I268T mutation affects the ability to the VDR to be stabilized in a transcriptionally active conformation by 1,25(OH)2D3 and whether this function can be rescued by the potent analog, 20-epi-1,25(OH)2D3. As shown in Fig. 9, a major protease-resistant band from the WT VDR was stabilized by 10 nm 1,25(OH)2D3 and 10 nm 20-epi-1,25(OH)2D3. In contrast, protease-resistant bands from the I268T mutant VDR were protected when 10 nm 20-epi-1,25(OH)2D3 was added, but not with 1,25(OH)2D3. When the mutant VDR was tested over a range of different ligand concentrations, the natural hormone protected a smaller molecular weight fragment only when 1000 nm 1,25(OH)2D3 was added. Lower concentrations of 1,25(OH)2D3 were ineffective in protecting the I268T mutant VDR from proteolysis. In contrast, addition of 20-epi-1,25(OH)2D3 protected two major fragments, the smaller fragment observed with 1,25(OH)2D3 and a higher molecular weight fragment observed in the WT VDR. These fragments were protected over a wide range of 20-epi1,25(OH)2D3 concentrations (Fig. 9). These data demonstrate that the vitamin D analog, 20-epi-1,25(OH)2D3, stabilized the I268T mutant receptor at least 100 times more effectively than 1,25(OH)2D3. Transcriptional activity of the I268T mutant in response to 20-epi-1,25(OH)2D3

To determine whether the better ability of the 20-epi1,25(OH)2D3 analog to stabilize conformation of I268T VDR against proteolytic digestion was reflected in rescue of transcriptional activity of this mutant, we performed transfection experiments. As shown in Fig. 10, when COS-7 cells were transfected with the WT VDR and the 24-hydroxylase promoter luciferase reporter, 20-epi-1,25(OH)2D3 shifted the dose-response curve to the left and was approximately 5-fold more potent than the natural hormone, 1,25(OH)2D3 [EC50, 0.1 nm for 1,25(OH)2D3 and 0.02 nm for 20-epi-1,25(OH)2D3, average of two experiments] as we have previously shown using this reporter (24). In contrast, 20-epi-1,25(OH)2D3 was approximately 20-fold more potent than 1,25(OH)2D3 [EC50, 3.5 nm for 1,25(OH)2D3 and 0.18 nm for 20-epi-1,25(OH)2D3; average of two experiments] in transactivation of the mutant VDR, which correlated with its better ability to stabilize the mutant VDR conformation in vitro. The 20-epi-analog restored transactivation of the I268T mutant to nearly the same level the

FIG. 9. Ligand potencies to stabilize WT VDR or I268T VDR in a protease-resistant conformation. Protease sensitivity assays were performed with in vitro-translated 35S-labeled WT or I268T VDRs, which were incubated with or without the indicated amounts of 1,25(OH)2D3 or 20-epi-1,25(OH)2D3 before digestion with trypsin. The protease-resistant fragments were separated by SDS-PAGE and detected by autoradiography.

Malloy et al. • Vitamin D Resistance

FIG. 10. Differential transactivation of I268T mutant VDR by 1,25(OH)2D3 and 20-epi-1,25(OH)2D3. Transcriptional activities were assessed in COS-7 cells cotransfected with the 24-hydroxylase luciferase reporter gene and WT or mutant VDR expression vectors. Left panel, WT VDR treated with 1,25(OH)2D3 (E) or 20-epi-1,25(OH)2D3 (□). Middle panel, I268T mutant VDR treated with 1,25(OH)2D3 (F) or 20-epi-1,25(OH)2D3 (f). Right panel, Comparison of WT VDR treated with 1,25(OH)2D3 (E) with I268T mutant VDR treated with 20-epi-1,25(OH)2D3 (f). Each transfection experiment was performed in triplicate, and each titration was performed twice. A representative experiment is shown. Error bars represent ⫾SD.

WT VDR treated with 1,25(OH)2D3 [EC50, 0.18 nm for 20epi-1,25(OH)2D3 for I268T and 0.1 nm for 1,25(OH)2D3 for WT VDR]. These results demonstrate that the analog exhibits increased potency in the mutant and is able to achieve transactivation potency similar to that of 1,25(OH)2D3 in the WT VDR (Fig. 10). Discussion

The young girl described here exhibited the classical clinical pattern of HVDRR. She initially presented with earlyonset rickets, secondary hyperparathyroidism, hypocalcemia, and elevated serum 1,25(OH)2D3 levels, but no alopecia. In analyzing the VDR from this patient, we identified a novel mutation, I268T, as the molecular cause of her vitamin D resistance. Our studies showed that the I268T mutation affects ligand binding, RXR heterodimerization, and coactivator binding. Although we were unable to obtain DNA from the parents, we presume each parent to be heterozygous for the mutation, because this is a recessive disease. The family has a history of consanguinity. From crystallographic studies of the VDR and other members of the steroid receptor superfamily, the LBDs of these receptors are composed of 11 or 12 ␣-helices forming a hydrophobic core that is occupied by the ligand (4). In the VDR, the ligand binding pocket is formed by helixes H1, H3, H4, H5, H7, H8, H9, H10, and H11 (4). The I268T mutation described here occurs in the LBD and is located in helix H5. Figure 4 shows the amino acids in the VDR LBD that interact with 1,25(OH)2D3. I268 forms van der Walls bonds with C22 on the side-chain of 1,25(OH)2D3 (Moras, D., unpublished observation). The I268T mutant VDR described here has an approximately 5- to 10-fold lower affinity for 1,25(OH)2D3 consistent with I268’s interaction with the ligand. The amino acids V234, I268, H397, and Y401, which directly interact with 1,25(OH)2D3, are also involved in the hydrophobic stabilization of helix H12 (4). The fact that these residues participate in both ligand binding and helix H12 stabilization suggests that the ligand may control helix H12 repositioning through these amino acids.

Malloy et al. • Vitamin D Resistance

The repositioning of helix H12 is a critical event that occurs as a consequence of ligand binding and is essential for transactivation. Repositioning of helix H12 involves both hydrophobic contacts and polar interactions and is essential for creating the correct surface interface for coactivator binding. The amino acids residues involved in the hydrophobic contacts include residues T415, L417, V418, L419, V421, and F422 from helix H12; residues D232, V234, S235, I238, and Q239 from helix H3; residues A267 and I268 from helix H5 and residues H397 and Y401 from helix H11 (4). Replacement of the hydrophobic isoleucine residue I268 with a nucleophilic threonine residue I268T would be expected to disrupt the hydrophobic contact between helix H5 and helix H12 and would probably prevent formation of the hydrophobic cleft needed for coactivator binding. Our data clearly demonstrate that the I268T mutation inhibits coactivator binding at low concentrations of 1,25(OH)2D3. In fact, the amount of 1,25(OH)2D3 needed to restore coactivator binding is much higher than the 5- to 10-fold change in the binding affinity for 1,25(OH)2D3, suggesting that the I268T mutation also directly alters helix H12 stabilization. The I268T mutation affects VDR-RXR heterodimerization. Deletion studies of the VDR have shown that the E1 region as well as helix H10 are involved in heterodimerization with RXR␣ and are necessary for transcriptional activity (21, 22, 25). The E1 region encompasses amino acid residues 244 –263 overlapping the C-terminal portion of helix H3, loop 3– 4, and the N-terminal portion of helix H4. Although I268 is not part of the E1 region, the I268T mutation clearly alters VDR-RXR heterodimerization. The data from the GST pull-down experiment suggest that the inability of the I268T mutant to bind more strongly to RXR at low 1,25(OH)2D3 concentrations may be related to its low affinity for 1,25(OH)2D3. In contrast, the data from the yeast two-hybrid assay and partial rescue by RXR in transactivation assays suggest that even at high concentrations of 1,25(OH)2D3, binding of the I268T mutant VDR to RXR is defective. However, it cannot be ruled out that additional factors may exist that contributed to these findings. An interesting finding in this case of HVDRR is the fact that the patient does not have alopecia. Alopecia is found in many, but not all, cases of HVDRR reported to date (2). Molecular analysis of HVDRR cases provides new insights into the association of VDR defects with alopecia. Alopecia is found in all HVDRR patients who have premature termination codons that truncate the VDR and in all patients who have DBD mutations (2). Alopecia is also found in HVDRR patients who have mutations in the VDR LBD that prevent VDR from binding to RXR (11, 13, 26). In cases in which alopecia was not associated with HVDRR, the molecular cause of the disease was due to mutations within the LBD. Three patients without alopecia had mutations in contact points for the ligand. R274, the contact point for the 1hydroxyl group, was mutated to R274L (10); H305, the contact point for the 25-hydroxyl group, was mutated to H305Q (12); and W286, which contacts the conjugated triene connecting the A and C rings, was mutated to W286R (27). A fourth patient without alopecia had an I314S mutation. I314 does not directly contact the ligand, but the mutation changed the ligand binding affinity of the receptor and in-

Endocrinology, November 2004, 145(11):5106 –5114 5113

terfered modestly with RXR heterodimerization (11). A fifth patient had an E420K mutation in helix H12 in the activation function-2 domain. The mutation had no effect on ligand binding, but prevented coactivators from binding to the VDR (14). We also recently reported an HVDRR patient without alopecia who had an insertion/substitution mutation in helix H1 that abolished ligand binding and coactivator binding (28). Taken together, these findings suggest that DNA binding and RXR binding are critical functions of the VDR needed to prevent alopecia and that ligand binding and coactivator binding are not. Because the I268T mutant can bind RXR in the absence of ligand and maintains the ability to bind to DNA, we hypothesize that these properties are sufficient to enable the mutant VDR to perform its role in regulating hair growth and results in the absence of alopecia in the patient. A recent report by Hsieh et al. (29) demonstrated that the hairless gene product (Hr) interacts with the VDR and represses VDR transactivation. We speculate that the VDR and Hr participate in the repression of some gene(s) that, when inappropriately expressed during the hair cycle due to mutations in the VDR or Hr, results in alopecia. Future studies will examine Hr interactions and gene repression with mutant VDRs from HVDRR patients with and without alopecia. The fact that the patient did not respond to vitamin D therapy with calcitriol is also of interest. Despite the presence of the mutation, we showed that the mutant VDR can activate gene transcription at high concentrations of 1,25(OH)2D3. Very high doses of calcitriol therapy might have been sufficient to overcome the ligand binding affinity defect. However, disruption of the coactivator-binding site caused by the mutation apparently was unable to be corrected with the level of calcitriol treatment the patient received. The patient was eventually treated with iv calcium and showed remarkable improvement, as has been reported in other cases of HVDRR (30 –32). The fact that the vitamin D analog 20-epi1,25(OH)2D3 stabilizes the mutant receptor and increases the transactivation capacity suggests that 20-epi-1,25(OH)2D3 or other analogs like it may be useful therapeutically to treat HVDRR patients with this specific type of mutation. However, the mechanism for the increased potency of the 20-epi analog on the I268T mutant vs. the WT VDR is not easily explained. The crystal structures of the VDR LBD bound to the 20-epi compounds MC1288, the 20-epi analog used in this study, and KH1060, another 20-epi analog, have been described (33). The aliphatic side-chains of both 20-epi analogs line the opposite side of the ligand binding pocket compared with 1,25(OH)2D3. Although it is not clear how the mutant would be affected by differences in the positions of the ligands, it is apparent that the 20-epi analog allows the I268T mutant to adopt a more favorable conformation that may affect the positioning of helix H12, thereby allowing better recruitment of coactivators and increasing transactivation. The crystal structure of the I268T mutant may provide the answer. In conclusion, we have identified a novel mutation in the VDR LBD, I268T, that affects ligand binding, RXR heterodimerization, and coactivator interaction and represents the molecular basis of HVDRR in a child with this disease.

5114

Endocrinology, November 2004, 145(11):5106 –5114

Malloy et al. • Vitamin D Resistance

Acknowledgments Received January 23, 2004. Accepted August 6, 2004. Address all correspondence and requests for reprints to: Dr. Peter J. Malloy, Division of Endocrinology, Gerontology and Metabolism, Stanford University School of Medicine, Stanford University Medical Center, Room S025, Stanford, California 94305-5103. E-mail: malloy@cmgm. stanford.edu. This work was supported by National Institutes of Health Grants DK-42482 (to D.F.) and DK-50583 (to S.P.).

16. 17. 18. 19.

References 1. Haussler MR, Whitfield GK, Haussler CA, Hsieh J, Thompson PD, Selznick SH, Dominguez CE, Jurutka PW 1998 The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res 13: 325–349 2. Malloy PJ, Pike JW, Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets. Endocr Rev 20:156 –188 3. Rachez C, Freedman LP 2000 Mechanisms of gene regulation by vitamin D3 receptor: a network of coactivator interactions. Gene 246:9 –21 4. Rochel N, Wurtz JM, Mitschler A, Klaholz B, Moras D 2000 The crystal structure of the nuclear receptor for vitamin D bound to its natural ligand. Mol Cell 5:173–179 5. Hughes MR, Malloy PJ, Kieback DG, Kesterson RA, Pike JW, Feldman D, O’Malley BW 1988 Point mutations in the human vitamin D receptor gene associated with hypocalcemic rickets. Science 242:1702–1705 6. Saijo T, Ito M, Takeda E, Huq AH, Naito E, Yokota I, Sone T, Pike JW, Kuroda Y 1991 A unique mutation in the vitamin D receptor gene in three Japanese patients with vitamin D-dependent rickets type II: utility of single-strand conformation polymorphism analysis for heterozygous carrier detection. Am J Hum Genet 49:668 – 673 7. Sone T, Marx SJ, Liberman UA, Pike JW 1990 A unique point mutation in the human vitamin D receptor chromosomal gene confers hereditary resistance to 1,25-dihydroxyvitamin D3. Mol Endocrinol 4:623– 631 8. Malloy PJ, Weisman Y, Feldman D 1994 Hereditary 1␣,25-dihydroxyvitamin D-resistant rickets resulting from a mutation in the vitamin D receptor deoxyribonucleic acid-binding domain. J Clin Endocrinol Metab 78:313–316 9. Lin NU-T, Malloy PJ, Sakati N, Al-Ashwal A, Feldman D 1996 A novel mutation in the deoxyribonucleic acid-binding domain of the vitamin D receptor gene causes hereditary 1,25-dihydroxyvitamin D resistant rickets. J Clin Endocrinol Metab 81:2564 –2569 10. Kristjansson K, Rut AR, Hewison M, O’Riordan JL, Hughes MR 1993 Two mutations in the hormone binding domain of the vitamin D receptor cause tissue resistance to 1,25 dihydroxyvitamin D3. J Clin Invest 92:12–16 11. Whitfield GK, Selznick SH, Haussler CA, Hsieh JC, Galligan MA, Jurutka PW, Thompson PD, Lee SM, Zerwekh JE, Haussler MR 1996 Vitamin D receptors from patients with resistance to 1,25-dihydroxyvitamin D3: point mutations confer reduced transactivation in response to ligand and impaired interaction with the retinoid X receptor heterodimeric partner. Mol Endocrinol 10:1617–1631 12. Malloy PJ, Eccleshall TR, Gross C, Van Maldergem L, Bouillon R, Feldman D 1997 Hereditary vitamin D resistant rickets caused by a novel mutation in the vitamin D receptor that results in decreased affinity for hormone and cellular hyporesponsiveness. J Clin Invest 99:297–304 13. Malloy PJ, Zhu W, Zhao XY, Pehling GB, Feldman D 2001 A novel inborn error in the ligand-binding domain of the vitamin D receptor causes hereditary vitamin D-resistant rickets. Mol Genet Metab 73:138 –148 14. Malloy PJ, Xu R, Peng L, Clark PA, Feldman D 2002 A novel mutation in helix 12 of the vitamin D receptor impairs coactivator interaction and causes hereditary 1,25-dihydroxyvitamin D-resistant rickets without alopecia. Mol Endocrinol 16:2538 –2546 15. Hewison M, Rut AR, Kristjansson K, Walker RE, Dillon MJ, Hughes MR,

20. 21.

22. 23.

24. 25.

26.

27.

28.

29.

30.

31. 32. 33.

O’Riordan JL 1993 Tissue resistance to 1,25-dihydroxyvitamin D without a mutation of the vitamin D receptor gene. Clin Endocrinol (Oxf) 39:663– 670 Chen H, Hewison M, Hu B, Adams JS 2003 Heterogeneous nuclear ribonucleoprotein (hnRNP) binding to hormone response elements: a cause of vitamin D resistance. Proc Natl Acad Sci USA 100:6109 – 6114 Malloy PJ, Hochberg Z, Tiosano D, Pike JW, Hughes MR, Feldman D 1990 The molecular basis of hereditary 1,25-dihydroxyvitamin D3 resistant rickets in seven related families. J Clin Invest 86:2071–2079 Swami S, Sarabia SF, Diamandis A, Mistry J, Khosravi J, Feldman D 2001 A new enzyme-linked immunosorbant assay for the measurement of human vitamin D receptor. Bone 28:319 –326 Malloy PJ, Xu R, Peng L, Peleg S, Ashwal A, Feldman D 2003 A mutation in the ligand binding domain of the VDR that alters 1,25(OH)2D3 binding, RXR heterodimerization and coactivator binding thereby causing hereditary vitamin D resistant rickets. J Bone Miner Res 18:S177 Bradford MM 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 72:248 –254 Whitfield GK, Hsieh JC, Nakajima S, MacDonald PN, Thompson PD, Jurutka PW, Haussler CA, Haussler MR 1995 A highly conserved region in the hormone-binding domain of the human vitamin D receptor contains residues vital for heterodimerization with retinoid X receptor and for transcriptional activation. Mol Endocrinol 9:1166 –1179 Jin CH, Kerner SA, Hong MH, Pike JW 1996 Transcriptional activation and dimerization functions in the human vitamin D receptor. Mol Endocrinol 10:945–957 Zhao XY, Eccleshall TR, Krishnan AV, Gross C, Feldman D 1997 Analysis of vitamin D analog-induced heterodimerization of vitamin D receptor with retinoid X receptor using the yeast two-hybrid system. Mol Endocrinol 11: 366 –378 Gardezi SA, Nguyen C, Malloy PJ, Posner GH, Feldman D, Peleg S 2001 A rationale for treatment of hereditary vitamin D-resistant rickets with analogs of 1␣,25-dihydroxyvitamin D3. J Biol Chem 276:29148 –29156 Nakajima S, Hsieh JC, MacDonald PN, Galligan MA, Haussler CA, Whitfield GK, Haussler MR 1994 The C-terminal region of the vitamin D receptor is essential to form a complex with a receptor auxiliary factor required for high affinity binding to the vitamin D-responsive element. Mol Endocrinol 8:159 – 172 Cockerill FJ, Hawa NS, Yousaf N, Hewison M, O’Riordan JL, Farrow SM 1997 Mutations in the vitamin D receptor gene in three kindreds associated with hereditary vitamin D resistant rickets. J Clin Endocrinol Metab 82:3156 – 3160 Nguyen TM, Adiceam P, Kottler ML, Guillozo H, Rizk-Rabin M, Brouillard F, Lagier P, Palix C, Garnier JM, Garabedian M 2002 Tryptophan missense mutation in the ligand-binding domain of the vitamin D receptor causes severe resistance to 1,25-dihydroxyvitamin D. J Bone Miner Res 17:1728 –1737 Malloy PJ, Xu R, Cattani A, Reyes L, Feldman D 2004 A unique insertion/ substitution in helix H1 of the vitamin D receptor ligand binding domain in a patient with hereditary 1,25-dihydroxyvitamin D-resistant rickets. J Bone Miner Res 19:1018 –1024 Hsieh JC, Sisk JM, Jurutka PW, Haussler CA, Slater SA, Haussler MR, Thompson CC 2003 Physical and functional interaction between the vitamin D receptor and hairless corepressor, two proteins required for hair cycling. J Biol Chem 278:38665–38674 Balsan S, Garabedian M, Larchet M, Gorski AM, Cournot G, Tau C, Bourdeau A, Silve C, Ricour C 1986 Long-term nocturnal calcium infusions can cure rickets and promote normal mineralization in hereditary resistance to 1,25-dihydroxyvitamin D. J Clin Invest 77:1661–1667 Hochberg Z, Tiosano D, Even L 1992 Calcium therapy for calcitriol-resistant rickets. J Pediatr 121:803– 808 al-Aqeel A, Ozand P, Sobki S, Sewairi W, Marx S 1993 The combined use of intravenous and oral calcium for the treatment of vitamin D dependent rickets type II (VDDRII). Clin Endocrinol (Oxf) 39:229 –237 Tocchini-Valentini G, Rochel N, Wurtz JM, Mitschler A, Moras D 2001 Crystal structures of the vitamin D receptor complexed to superagonist 20-epi ligands. Proc Natl Acad Sci USA 98:5491–5496

Endocrinology is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the endocrine community.