Molecular basis of the selective activity of vitamin D analogues

More than 2,000 synthetic analogues of the biological active form of vitamin D, 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)), are presently known. Basically, all of them interfere with the molecular switch of nuclear 1alpha,25(OH)(2)D(3) signaling, which is the complex of the vitamin D receptor (VDR), the retinoid X receptor (RXR), and a 1alpha,25(OH)(2)D(3) response element (VDRE). Central element of this molecular switch is the ligand-binding domain (LBD) of the VDR, which can be stabilized by a 1alpha,25(OH)(2)D(3) analogue either in its agonistic, antagonistic, or non-agonistic conformation. The positioning of helix 12 of the LBD is of most critical importance for these conformations. In each of the three conformations, the VDR performs different protein-protein interactions, which then result in a characteristic functional profile. Most 1alpha,25(OH)(2)D(3) analogues have been identified as agonists, a few are antagonists (e.g., ZK159222 and TEI-9647), and only Gemini and some of its derivatives act under restricted conditions as non-agonists. The functional profile of some 1alpha,25(OH)(2)D(3) analogues, such as EB1089 and Gemini, can be modulated by protein and DNA interaction partners of the VDR. This provides them with some selectivity for DNA-dependent and -independent signaling pathways and VDRE structures.     

Ligand-mediated conformational changes of the VDR are required for gene transactivation

The central element of the molecular switch of nuclear 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)) signaling is the ligand-binding domain (LBD) of the Vitamin D receptor (VDR), which can be stabilized by 1alpha,25(OH)(2)D(3) or its analogues in to agonistic, antagonistic or inverse agonistic conformations. The positioning of helix 12 of the LBD is of most critical importance for these conformations, because it determines the distance between the charge clamp amino acids K246 and E420 that are essential for VDR-coactivator (CoA) interaction. Most VDR ligands have been identified as agonists and only a few (e.g., ZK168281 and TEI-9647) as pure or partial antagonists. Antagonists induce corepressor (CoR) dissociation from the VDR but prevent completely or partially CoA interaction and thus transactivation. Gemini is a 1alpha,25(OH)(2)D(3) analogue with two identical side chains that despite its significantly increased volume binds to the VDR and acts under most conditions as an agonist. Interestingly, supramolar CoR concentrations shift Gemini from an agonist to an inverse agonist, which actively recruits CoR to the VDR and thus mediates repression of 1alpha,25(OH)(2)D(3) target genes. Gemini is the first described (conditional) inverse agonist to an endocrine nuclear receptor (NR) and may function as a sensor for cell-specific CoA/CoR ratios.     

Molecular modeling, affinity labeling, and site-directed mutagenesis define the key points of interaction between the ligand-binding domain of the vitamin D nuclear receptor and 1 alpha,25-dihydroxyvitamin D3

We have combined molecular modeling and classical structure-function techniques to define the interactions between the ligand-binding domain (LBD) of the vitamin D nuclear receptor (VDR) and its natural ligand, 1alpha,25-dihydroxyvitamin D(3) [1alpha,25-(OH)(2)D(3)]. The affinity analogue 1alpha,25-(OH)(2)D(3)-3-bromoacetate exclusively labeled Cys-288 in the VDR-LBD. Mutation of C288 to glycine abolished this affinity labeling, whereas the VDR-LBD mutants C337G and C369G (other conserved cysteines in the VDR-LBD) were labeled similarly to the wild-type protein. These results revealed that the A-ring 3-OH group docks next to C288 in the binding pocket. We further mutated M284 and W286 (separately creating M284A, M284S, W286A, and W286F) and caused severe loss of ligand binding, indicating the crucial role played by the contiguous segment between M284 and C288. Alignment of the VDR-LBD sequence with the sequences of nuclear receptor LBDs of known 3-D structure positioned M284 and W286 in the presumed beta-hairpin of the molecule, thereby identifying it as the region contacting the A-ring of 1alpha, 25-(OH)(2)D(3). From the multiple sequence alignment, we developed a homologous extension model of the VDR-LBD. The model has a canonical nuclear receptor fold with helices H1-H12 and a single beta hairpin but lacks the long insert (residues 161-221) between H2 and H3. We docked the alpha-conformation of the A-ring into the binding pocket first so as to incorporate the above-noted interacting residues. The model predicts hydrogen bonding contacts between ligand and protein at S237 and D299 as well as at the site of the natural mutation R274L. Mutation of S237 or D299 to alanine largely abolished ligand binding, whereas changing K302, a nonligand-contacting residue, to alanine left binding unaffected. In the "activation" helix 12, the model places V418 closest to the ligand, and, consistent with this prediction, the mutation V418S abolished ligand binding. The studies together have enabled us to identify 1alpha,25-(OH)(2)D(3)-binding motifs in the ligand-binding pocket of VDR.     

Antagonistic action of a 25-carboxylic ester analogue of 1alpha, 25-dihydroxyvitamin D3 is mediated by a lack of ligand-induced vitamin D receptor interaction with coactivators

A 25-carboxylic ester analogue of 1alpha,25-dihydroxyvitamin D(3) (1alpha,25-(OH)(2)D(3)), ZK159222, was described as a novel type of antagonist of 1alpha,25-(OH)(2)D(3) signaling. The ligand sensitivity of ZK159222, in facilitating complex formation between 1alpha,25-(OH)(2)D(3) receptor (VDR) and the retinoid X receptor (RXR) on a 1alpha,25-(OH)(2)D(3) response element (VDRE), was approximately 7-fold lower when compared with 1alpha,25-(OH)(2)D(3). However, ZK159222 was not able to promote a ligand-dependent interaction of the VDR with the coactivator proteins SRC-1, TIF2, and RAC3, neither in solution nor in a complex with RXR on DNA. Functional analysis in HeLa and COS-7 cells demonstrated a 10-100-fold lower ligand sensitivity for ZK159222 than for 1alpha, 25-(OH)(2)D(3) and, most interestingly, a potency that was drastically reduced compared with 1alpha,25-(OH)(2)D(3). A cotreatment of 1alpha,25-(OH)(2)D(3) with a 100-fold higher concentration of ZK159222 resulted in a prominent antagonistic effect both in functional in vivo and in in vitro assays. These data suggest that the antagonistic action of ZK159222 is due to a lack of ligand-induced interaction of the VDR with coactivators with a parallel ligand sensitivity, which is sufficient for competition with the natural hormone for VDR binding.     

Structural determinants for vitamin D receptor response to endocrine and xenobiotic signals

The vitamin D receptor (VDR), initially identified as a nuclear receptor for 1alpha,25-dihydroxyvitamin D3 [1alpha,25(OH)2D3], regulates calcium metabolism, cellular proliferation and differentiation, immune responses, and other physiological processes. Recently, secondary bile acids such as lithocholic acid (LCA) were identified as endogenous VDR agonists. To identify structural determinants required for VDR activation by 1alpha,25(OH)2D3 and LCA, we generated VDR mutants predicted to modulate ligand response based on sequence homology to pregnane X receptor, another bile acid-responsive nuclear receptor. In both vitamin D response element activation and mammalian two-hybrid assays, we found that VDR-S278V is activated by 1alpha,25(OH)2D3 but not by LCA, whereas VDR-S237M can respond to LCA but not to 1alpha,25(OH)2D3. Competitive ligand binding analysis reveals that LCA, but not 1alpha,25(OH)2D3, effectively binds to VDR-S237M and both 1alpha,25(OH)2D3 and LCA bind to VDR-S278V. We propose a docking model for LCA binding to VDR that is supported by mutagenesis data. Comparative analysis of the VDR-LCA and VDR-1alpha,25(OH)2D3 structure-activity relationships should be useful in the development of bile acid-derived synthetic VDR ligands that selectively target VDR function in cancer and immune disorders without inducing adverse hypercalcemic effects.     

Epigenetic repression of transcription by the Vitamin D3 receptor in prostate cancer cells

Normal prostate epithelial cells are acutely sensitive to the antiproliferative action of 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)), whilst prostate cancer cell lines and primary cultures display a range of sensitivities. We hypothesised that key antiproliferative target genes of the Vitamin D receptor (VDR) were repressed by an epigenetic mechanism in 1alpha,25(OH)(2)D(3)-insensitive cells. Supportively, we found elevated nuclear receptor co-repressor and reduced VDR expression correlated with reduced sensitivity to the antiproliferative action of 1alpha,25(OH)(2)D(3). Furthermore, the growth suppressive actions of 1alpha,25(OH)(2)D(3) can be restored by co-treatment with low doses of histone deacetylation inhibitors, such as trichostatin A (TSA) to induce apoptosis. Examination of the regulation of VDR target genes revealed that co-treatment of 1alpha,25(OH)(2)D(3) plus TSA co-operatively upregulated GADD45alpha. Similarly in a primary cancer cell culture, the regulation of appeared GADD45alpha repressed. These data demonstrate that prostate cancer cells utilise a mechanism involving deacetylation to suppress the responsiveness of VDR target genes and thus ablate the antiproliferative action of 1alpha,25(OH)(2)D(3).     

Natural metabolites of 1alpha,25-dihydroxyvitamin D(3) retain biologic activity mediated through the vitamin D receptor

1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)), the active metabolite of vitamin D, mediates many of its effects through the intranuclear vitamin D receptor (VDR, NR1I1), that belongs to the large superfamily of nuclear receptors. Vitamin D receptor can directly regulate gene expression by binding to vitamin D response elements (VDREs) located in promoter or enhancer regions of various genes. Although numerous synthetic analogs of 1alpha,25(OH)(2)D(3) have been analysed for VDR binding and transactivation of VDRE-driven gene expression, the biologic activity of many naturally occurring metabolites has not yet been analyzed in detail. We therefore studied the transactivation properties of 1alpha,24R, 25-trihydroxyvitamin D(3) (1alpha,24R,25(OH)(3)D(3)), 1alpha, 25-dihydroxy-3-epi-vitamin D(3) (1alpha,25(OH)(2)-3-epi-D(3)), 1alpha,23S,25-trihydroxyvitamin D(3) (1alpha,23S,25(OH)(3)D(3)), and 1alpha-hydroxy-23-carboxy-24,25,26,27-tetranorvitamin D(3) (1alpha(OH)-24,25,26,27-tetranor-23-COOH-D(3); calcitroic acid) using the human G-361 melanoma cell line. Cells were cotransfected with a VDR expression plasmid and luciferase reporter gene constructs driven by two copies of the VDRE of either the mouse osteopontin promoter or the 1alpha,25(OH)(2)D(3) 24-hydroxylase (CYP24) promoter. Treatment with 1alpha,25(OH)(2)D(3) or the metabolites 1alpha,24R,25(OH)(3)D(3), 1alpha,25(OH)(2)-3-epi-D(3), and 1alpha,23S,25(OH)(3)D(3) resulted in transactivation of both constructs in a time- and dose-dependent manner, and a postitive regulatory effect was observed even for calcitroic acid in the presence of overexpressed VDR. The metabolites that were active in the reporter gene assay also induced expression of CYP24 mRNA in the human keratinocyte cell line HaCaT, although with less potency than the parent hormone. A ligand-binding assay based on nuclear extracts from COS-1 cells overexpressing human VDR demonstrated that the metabolites, although active in the reporter gene assay, were much less effective in displacing [(3)H]-labeled 1alpha,25(OH)(2)D(3) from VDR than the parent hormone. Thus, we report that several natural metabolites of 1alpha,25(OH)(2)D(3) retain significant biologic activity mediated through VDR despite their apparent low affinity for VDR.     

Ligand recognition by the vitamin D receptor.

Three-dimensional structure of the ligand binding domain (LBD) of the vitamin D receptor (VDR) docked with the natural ligand 1 alpha,25-dihydroxyvitamin D(3) [1,25-(OH)(2)D(3)] has been mostly solved by the X-ray crystallographic analysis of the deletion mutant (VDR-LBD Delta 165-215). The important focus, from now on, is how the VDR recognizes and interacts with potent synthetic ligands. We now report the docking models of the VDR with three functionally and structurally interesting ligands, 22-oxa-1,25-(OH)(2)D(3) (OCT), 20-epi-1,25-(OH)(2)D(3) and 20-epi-22-oxa-24,26,27-trihomo-1,25-(OH)(2)D(3). In parallel with the computational docking studies, we prepared twelve one-point mutants of amino acid residues lining the ligand binding pocket of the VDR and examined their transactivation potency induced by 1,25-(OH)(2)D(3) and these synthetic ligands. The results indicate that L233, R274, W286, H397 and Y401 are essential for holding the all ligands tested, S278 and Q400 are not important at all, and the importance of S237, V234, S275, C288 and H305 is variable depending on the side-chain structure of the ligands. Based on these studies, we suggested key structural factors to bestow the selective action on OCT and the augmented activities on 20-epi-ligands. Furthermore, the docking models coincided well with our proposed active space-region theory of vitamin D based on the conformational analyses of ligands.     

Critical role of helix 12 of the vitamin D(3) receptor for the partial agonism of carboxylic ester antagonists.

The carboxy-terminal alpha-helix of a nuclear receptor ligand-binding domain (LBD), helix 12, contains a critical, ligand-modulated interface for the interaction with coactivator proteins. In this study, using the example of the vitamin D receptor (VDR) and the partial antagonist ZK159222, the role of helix 12 (residues 417-427) for both antagonistic and agonistic receptor actions was investigated. Amino acid residue G423 was demonstrated to be critical for partial agonism of ZK159222, but not for the activity of the natural VDR agonist, 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)). The amount of partial agonism of ZK159222 increased when helix 12 was truncated by the last four amino acid residues (Delta424-27) and augmented even more, when in addition helix 12 of VDR's dimerization partner, retinoid X receptor (RXR), was truncated. In contrast, the low agonism of a structural derivative of ZK159222, ZK168281, was not affected comparably, whereas other close structural relatives of ZK159222 even demonstrated the same agonistic activity as that of 1alpha,25(OH)(2)D(3). The amount of agonism of ZK159222 and ZK168281 at different variations of helix 12 correlated well with VDR's ability to complex with coactivator proteins and inversely correlated with the strength of the compound's antagonistic action on 1alpha,25(OH)(2)D(3) signalling. Molecular dynamics simulations of the LBD complexed with the two antagonists could explain their different action by demonstrating a more drastic displacement of helix 12 through ZK168281 than through ZK159222. Moreover, the modelling could indicate a kink of helix 12 at amino acid residue G423, which provides the last four amino acid residues of helix 12 with a modulatory role for the partial agonism of some VDR antagonists, such as ZK159222. In conclusion, partial agonism of a VDR antagonist is lower the more it disturbs helix 12 in taking the optimal position for coactivator interaction.     

Sensitivity to growth suppression by 1alpha,25-dihydroxyvitamin D(3) among MCF-7 clones correlates with Vitamin D receptor protein induction

The antiproliferative effect of 1alpha,25-dihydroxyvitamin D(3) (1alpha,25(OH)(2)D(3)) has been studied for a decade in diverse model systems, but the signalling pathways linking 1alpha,25(OH)(2)D(3) to cell cycle arrest remains unclear. In our attempt to establish a model system which would allow further identification of important players in the process of the 1alpha,25(OH)(2)D(3) imposed cell cycle arrest, we have isolated derivatives of the human breast cancer cell line MCF-7 and chosen two nearly 1alpha,25(OH)(2)D(3) resistant and two hypersensitive sub-clones. Investigation of cell cycle proteins regulated by 1alpha,25(OH)(2)D(3) in these clones indicates that activation of one component/pathway is responsible for the linkage between 1alpha,25(OH)(2)D(3) and growth arrest. Protein levels of the Vitamin D receptor (VDR) were elevated in sensitive cells upon 1alpha,25(OH)(2)D(3) treatment, whereas resistant clones were unable to induce VDR upon 1alpha,25(OH)(2)D(3) treatment. Our data show that VDR protein levels and the ability of a cell to induce VDR upon 1alpha,25(OH)(2)D(3) treatment correlate with the antiproliferative effects of 1alpha,25(OH)(2)D(3), and suggest that the level of VDR in cancer cells might serve as a prognostic marker for treatment of cancer with 1alpha,25(OH)(2)D(3) analogues.