HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Schneider, L. Articles by Torresani, J. Search for Related Content PubMed PubMed Citation Articles by Schneider, L. Articles by Torresani, J.

¹ INSERM U476 and IFR 35, Université de la Méditerranée, Faculté de Medecine, 13385 Marseille, Cedex 5, France
² Pennington Biomedical Research Center, Louisiana State University, Baton Rouge, LA 70808, USA

(Requests for offprints should be addressed to J Torresani; Email: Janine.Torresani{at}medecine.univ-mrs.fr)

	Abstract

Top Abstract Introduction Materials and methods Results Discussion References

 
1,25-(OH)₂-Vitamin D3 (1,25-D3) and the thyroid hormone tri-iodothyronine (T₃) were previously shown to behave as adipogenic agents in murine Ob17 preadipocytes. Moreover, these agents interfere with each other’s action during adipocyte differentiation. T₃ receptor (TR) expression and a downmodulation of T₃ binding sites (TR sites) by 1,25-D3 were also reported. A cross talk at the T₃ and 1,25-D3 receptor (VDR) level was suggested. We report here that Ob17 cells contain VDR receptor sites in markedly modulated number. This includes a sharp decrease during differentiation that was largely counteracted by 1,25-D3 added to preadipocytes in physiological, adipogenic concentrations. In parallel, the VDR mRNA level did not change significantly, neither did a variant produced by alternative splicing in the penultimate exon and defined for the first time in the mouse. The differentiation- and 1,25-D3-related modulations of VDR sites are likely to be, at least for the most part, the result of variations in abundance of the VDR protein, and may thus mainly involve post-translational events. In contrast, the addition of T₃ to the preadipocytes amplified the differentiation-related decrease in VDR sites, even in the presence of 1,25-D3. T₃ significantly decreased the levels of VDR transcripts and thus mainly exerts a pretranslational action. With regard to the reciprocal downmodulation of the TR sites (identified as almost exclusively of the TR type) by physiological concentrations of 1,25-D3, a post-translational action and a sequestration of the TR sites had previously been suggested and are further studied here. Analyses of receptor properties after co-incubations of recombinant VDRs and TRs did not favour direct VDR–TR interaction as a main cause of TR site sequestration. Interestingly, when taken together, the data on downregulation of VDRs and TRs by the alternate ligands define a potential step in the cross talk exerted between 1,25-D3 and T₃ for their adipogenic action. In addition, the present results also show for the first time that 1,25-D3 acts as a strong trigger of a transient expression of TRß1 subtype at an early preadipocyte step, an effect that had previously been assigned to T₃. This last interesting event introduces further incentive for deciphering the T₃/1,25-D3 cross talk in preadipocyte differentiation.

	Introduction

Top Abstract Introduction Materials and methods Results Discussion References

 
The 1,25-dihydroxyvitamin D3 (1,25-(OH)₂D3 or calcitriol; 1,25-D3) is the main active form of vitamin D. It has an important role in calcium and phosphate homeostasis and in the regulation of cell proliferation and differentiation (Walters 1992, Jones et al. 1998). In previous reports, we demonstrated that 1,25-D3, like the thyroid hormone 3,5,3'-tri-iodo-L-thyronine (T₃), exerts an adipogenic action on murine Ob17 preadipocytes. Either T₃ or 1,25-D3 alone were sufficient to trigger the differentiation of preadipocytes cultured under conditions that did not allow adipogenesis (use of culture medium either serum free or supplemented with fetal bovine serum (FBS) depleted of adipogenic factors). Adipogenesis was optimal under physiological concentrations of either agent and was decreased or suppressed by greater concentrations. Interestingly, when added together, both agents interfered in each other’s adipogenic action and behaved as synergistic agents, and then as antagonistic agents when the 1,25-D3 concentration was enhanced (Lenoir et al. 1996, Dace et al. 1997). These results thus strongly suggested that T₃ and 1,25-D3 exert redundant actions in the triggering of adipocyte differentiation and that a cross talk exists between their respective pathways of activity.

Most of the actions of 1,25-D3 and T₃ are exerted through modulations of gene expression and after their binding to specific high-affinity nuclear receptors, the vitamin D receptor (VDR) and the T₃ receptors (TRs), which belong to the steroid/thyroid-retinoid family of nuclear hormone receptors (Mangelsdorf & Evans 1995). To date, three main functional TRs have been identified (TR1, TRß1 and TRß2), which are encoded by two homologous c-erbA and c-erbAß genes that can also give rise to non-receptor variants. TR1 and TRß1 are widely distributed in different tissues, although large differences exist in their relative distribution (Yen 2001). To date, only one VDR gene has been identified, giving rise to one main functional VDR in different tissues (Malloy et al. 1999) although VDR isoforms may also be produced (Ebihara et al. 1996, Byrne et al. 2000). The TRs and VDR display structural and functional similarities. They can form homodimers, but preferentially heterodimerise with the retinoid receptors (RXRs). The heterodimers bind with high affinity to DNA response elements (HRE) in the regulatory regions of target genes and consequently activate or repress their transcription. The HREs that are recognised by VDR (VDRE) or TRs (TRE) are composed of two similar core motifs. Differences in the arrangement and spacing of these motifs define some degree of specific recognition (Mangelsdorf & Evans 1995). In addition, like other nuclear receptors, VDR and TRs can interact with several nuclear co-modulators (co-activators, co-repressors) that play an important part in the control of transcription and are shared by several nuclear receptors (Lemon & Freedman 1999, Xu et al. 1999).

In the Ob17 cells, nuclear T₃ binding sites and TR mRNAs were first identified as being products of the TR gene (Bismuth et al. 1995, Lenoir et al. 1996), although we also recently identified a transient expression of the TRß1 isoform in early preadipocytes (Dace et al. 1999), that is, at the onset of a sequential expression of several adipogenic transacting factors (MacDougald & Lane 1995, Rosen et al. 2000). This early transient expression of TRß1 (TR sites, mRNA) was always markedly lower than that of TR1. It was obliterated under non-adipogenic culture conditions and enhanced by T₃. Contrasting with this peculiar pattern of TRß1 regulation, the total number of TR sites, mainly of the TR1 type, was found to be markedly downmodulated by both T₃ and 1,25-D3. Both acted within their physiological adipogenic concentration range at a post-translational level (Lenoir et al. 1996). The hypothesis of a cross talk at the level of TR and VDRs was raised, and the possible involvement of TR/VDR interactions was suggested.

We previously presented, as preliminary data, the characterisation of VDR transcripts in Ob17 cells (Dace et al. 1997). Nevertheless, interpreting the cellular and molecular levels of the cross talk between the 1,25-D3 and T₃ signalling pathways required a better knowledge of VDR gene expression at the transcript and protein levels, of the ability to bind 1,25-D3 and of the types of modulation involved. This has been achieved in the present study, which demonstrates a downmodulation of the 1,25-D3 binding sites (VDR sites) by T₃. A cross-downmodulation of VDR and TR sites by the alternate ligand is thus obvious. It is further shown that these crossed events involve distinct pre- or post-translational molecular pathways, but not direct VDR–TR interactions. This study also led to the identification of a new VDR mRNA splicing variant that differs from the canonical VDR in its 3'-coding terminus and is close to a VDR1 variant initially described in the rat (Ebihara et al. 1996). Interestingly, it was also shown that 1,25-D3 is as efficient as T₃ in triggering a transient expression of TRß1 that occurs early during adipose differentiation. This adds another molecular level in the cross talk between the T₃ and 1,25-D3 signalling pathways.

	Materials and methods

Top Abstract Introduction Materials and methods Results Discussion References

 
Cell culture

Ob17 preadipocytes, cloned from periepididymal fat pads of adult genetically obese mice (Ob/Ob, C57/BL6J) (Negrel et al. 1978), were seeded at a density of 2000/cm² and grown in Dulbecco’s minimal essential medium (DMEM) supplemented with 10% FBS, 33 µM biotin, 17 µM sodium pantothenate and antibiotics (200 U/ml penicillin, 50 µg/ml streptomycin) (standard medium), as we described previously (Lenoir et al. 1996). At confluence (day 0, occurring 5 days after seeding), the medium was or was not supplemented with 17 nM insulin as an amplifier of the adipose differentiation. The adipogenic agents 1,25-D3 and T₃ (0.2 nM and 1.5 nM respectively; Sigma Chemical Co.) were generally added at confluence or as indicated. FBS stripped of adipogenic agents was obtained by applying the AG1-x8 exchange resin method (Lenoir et al. 1996). Stripped FBS, when used, was substituted with normal FBS 24 h after seeding. Under conditions that allow adipogenesis, clusters of lipid-filled cells progressively appear from day 6–8 after confluence and develop over 2–3 weeks. For some experiments, cells were synchronised using a serum-deprivation method (Dace et al. 1999): the day after plating, cells were washed with PBS and grown for 1 day in DMEM containing 0.05% fatty acid free BSA. The medium was then exchanged for standard medium containing normal or stripped FBS (stripped medium). 1,25-D3, T₃ or both were dissolved in ethanol. Ethanol was added to the assays and control culture media to a total of 0.1%.

Assay of VDR binding sites

VDR site abundance was assayed in Ob17 cells following previously described procedures (Zhao & Feldman 1993), all the steps being performed at 0–4 °C. Cells were homogenised in TKEMo buffer (20 mM Tris–HCl, 300 mM KCl, 1 mM EDTA, 10 mM sodium molybdate, pH 7.9) containing 0.5 mM dithiothreitol (DTT). After centrifugation (150 000 g for 30 min), aliquots of supernatants were incubated for 18 h with 0.06 nM [³H]1,25-D3 (418 mCi/mg, Amersham International plc), either in the absence of radioinert 1,25-D3 or with increasing concentrations from 0.04 to 0.20 nM. Non-specific binding was determined in parallel samples containing a 1600-fold excess of unlabelled 1,25-D3. Bound and free hormones were separated by hydroxyapatite (Bio-Gel HTP, Biorad Laboratories, Hercules, CA, USA) in TKE buffer (10 mM Tris–HCl, 0.1 M KCl, 1 mM EDTA) (Hermey & Popoff 1995). Receptor-bound 1,25-D3 retained on washed hydroxyapatite was solubilised in 100% ethanol at 37 °C and, after ethanol evaporation, solubilised in Permafluor scintillation cocktail (Permafluor I; Packard, Meriden, CT, USA) for estimation of radioactivity. 1,25-D3 receptor site concentration and affinity for 1,25-D3 (K_a) were estimated by Scatchard analysis of the specific binding data. DNA was estimated by the method of Labarca & Paigen (1980). Protein determination was carried out using the Coomassie Plus protein assay kit provided by Pierce Chemical Co. (Rockford, IL, USA) with BSA as standard.

RNA extraction and mRNA analyses

Total RNA was extracted from Ob17 cells using Trizol reagent according to the supplier’s recommendations (Life Technologies) and treated with deoxyribonuclease RQ1 (Promega). RNA purity was judged as an A260:A280 ratio greater than 1.8. First-strand cDNA was synthesised from 1 µg total RNA primed with random hexamers using Moloney mouse leukaemia virus reverse transcriptase (200 U/assay) as described by the supplier (Life Technologies).

For VDR mRNA analysis in northern blots, a mouse VDR cDNA probe was designed as described previously (Dace et al. 1997) and produced by RT-PCR amplification on total RNA using a set of primers selected within the reported mouse VDR sequence (Kamei et al. 1995) (forward primer a from nucleotide (nt) 486 to 511, reverse primer b from nt 868 to 843). Briefly, total RNA (20 µg) was run in 2.2 M formaldehyde–1% agarose gel, transferred onto Hybond N⁺ membrane (Amersham International plc) and hybridised with the [³²P]dCTP-labelled mouse VDR cDNA probe. The radiolabelled bands were analysed in a Fuji Bas 2000 apparatus (Fuji Photo Film Co., Inc, Tokyo, Japan) and normalised to total ribosomal RNA estimated after staining with methylene blue.

The abundance of transcripts encoding VDR, TRß1 and C/EBP (CAAT enhancer binding protein , expression of which is representative of adipose terminal differentiation (MacDougald & Lane 1995)) was also analysed in semiquantitative RT-PCR assays relative to ß-actin transcripts in co-amplification reactions (Lenoir et al. 1996). The ß-actin transcript level was used as internal standard because it does not significantly change per cell during the Ob17 differentiation process (Dani et al. 1990). The primers used for the specific amplification of VDR (primers a–b), C/EBP and ß-actin cDNAs have been described previously (Dace et al. 1999). In addition to the initially described VDR, a VDR variant has been defined in the rat (Ebihara et al. 1996) and, here, is detected in the mouse (see below). The VDR set of primers that we initially selected cannot discriminate between the canonical VDR (VDR0) and the variant (VDR1). Two other sets of primers were therefore designed that would selectively analyse the VDR0 and VDR1 transcripts (see below). The TRß1 primers were as follows: forward primer from nt 40 to 67 (5' CCCAGCATGACTACTAACCTATGACTCC 3') and reverse primer from nt 399 to 374 (5' CTTTGTC CCCACACACTACACAGAGC 3') in the mouse TRß1 sequence (Wood et al. 1991). Because of the disparity between the abundance of the different transcripts, ß-actin primers were added during the PCR reaction at different moments, to limit ß-actin amplification. Under these conditions, the transcripts of interest and ß-actin products accumulated exponentially, following parallel slopes, thus allowing estimation of individual mRNAs with regard to ß-actin mRNA. PCR products were separated on 2.2% agarose gel and stained with GelStar nucleic acid gel stain (FMC BioProducts, Rockland, ME, USA). Band intensities were analysed using the Kodak Scientific Imaging Systems (Kodak Digital Science 1D).

Identification of mouse VDR1 mRNA variant

A VDR mRNA variant described in the rat (rVDR1) results from alternative splicing of the VDR primary transcript at codon 337 and retention of intron 8. An open reading frame and a stop codon in intron 8 were found to be able to direct the production of a VDR1 protein isoform shorter than VDR0 in its C-terminus. Both rVDR0 and rVDR1 mRNAs exhibited a similar size, 4.6 kb (Ebihara et al. 1996). To search for the possible existence of this variant in the mouse (m), and assuming a general organisation of the mVDR gene similar to that of the rat, two primers were selected on each side of intron 9 of the mVDR gene (equivalent to intron 8 in the rat and following the nomenclature proposed for the mVDR gene (Jehan & DeLuca 1997)) and referred to as 1 and 2 (nt 991–1010 and nt 1179–1160 in the mVDR cDNA sequence (Kamei et al. 1995) respectively (see Fig. 3A)). A DNA fragment containing intron 9 was generated by PCR from genomic DNA of Ob17 cells using this set of primers, then purified and sequenced (Abi Prism 310, Applied Biosystem, Foster City, CA, USA). Based on this sequence, a reverse primer, designated 3, was designed in intron 9 (nt 23–42 from the splice site). RT-PCR was carried out on mRNA of Ob17 cells using the direct primer 1 and reverse primers 2 or 3 for VDR0 (1–2) or VDR1 (1–3) analyses respectively. The products were verified by sequencing.

View larger version (33K): [in this window] [in a new window]   Figure 3 Identification of a novel variant of VDR in the mouse Ob17 cells. (A) Representation of the mouse VDR gene region of interest [schematised exon (E)–intron (I) structure from E6 to E10] and of the putative protein structures of the two VDR variants (canonical VDR0 and VDR1) produced by alternative splicing from the primary transcript. The different primers used for RT-PCR and northern blot analyses are shown by arrows; brackets show the positions of epitopes recognised by the antibodies mAb and Ab C-t. aa, amino acids. (B) RT-PCR analysis of VDR0 and VDR1 transcripts in Ob17 cells. A co-amplification using the two sets of primers for VDR0 (primers 1 and 2) and VDR1 (primers 1 and 3) was carried out by RT-PCR with 1 µg total RNA from Ob17 preadipocytes. The number of amplification cycles was from 24 to 29 (lanes 1–6). Transcripts were visualised, after electrophoretic separation in 2.2% agarose gel, by staining with GelStar nucleic acid gel stain. (C) Presentation of the nucleotide and deduced 3'-end amino acid sequences of the mVDR1 cDNA isolated from Ob17 cell cDNA library. The boxed region delimits the intron 9 nucleotide sequence and thus the deduced C-terminal VDR1-specific sequence. The nucleotides are numbered [according to the numbering standard proposed by Kamei et al.(1995); accession number D31969 [GenBank] ] on the left and amino acids on the right. The VDR1 cDNA thus contains a 1113 bp open reading frame, which encodes a 370 amino acid protein. Amino acids are expressed with the case letter denomination. The position of the putative TGA stop codon is marked by an asterisk. (D) Alignment of the C-terminal amino acid sequences of mouse and rat VDR1. Horizontal bars indicate the amino acids common to the two sequences.

Different expression vectors containing the full-length coding sequences of nuclear receptors were used. Human (h)VDR, subcloned into the pSG5 vector, was provided by Dr M Thomasset (INSERM U458, Paris, France). Rat RXR1 and human c-erbAß1 (hTRß1), cloned into the pGEM3 vector, were provided by Dr J A Gustafsson (Karolinska Institute, Huddinge, Sweden) and Dr R Evans (The Salk Institute of Biological Studies, La Jolla, CA, USA) respectively. These receptors were translated in vitro using a TNT–T7-coupled rabbit reticulocyte lysate system in the presence or absence of ³⁵S-methionine (1000 Ci/mmol, Amersham) and following the supplier’s recommendations (Promega). Recombinant hTR1 was also produced in Escherichia coli and partially purified as we described previously (Daadi et al. 1995).

Analyses of VDR protein in western blots

PBS-washed Ob17 cells or nuclei purified from these cells (Lenoir et al. 1996) were solubilised in electrophoresis sample buffer (2% SDS, 10% glycerol, 5% ß-mercaptoethanol, 62.5 mM Tris–HCl, pH 6.8) and denatured (5 min at 95 °C). Equivalent amounts (DNA normalised) of whole cells or nuclear extracts were separated by SDS-PAGE in 10% gels and checked for protein amount equivalence by staining of the membranes with Ponceau Red after gel transfer. For western blotting, the following anti-VDR antibodies (Ab) were applied: rabbit polyclonal Ab directed against a 20 amino acid C-terminal peptide in rat VDR [supplied as purified IgGs and termed Ab C-t (sc-1008 X; Santa Cruz Biotechnology Inc, Santa Cruz, CA, USA) and used at 1/5000 dilution]; rat monoclonal Ab (mAb) directed against a central region in chicken VDR, C-terminal to the DNA binding domain (amino acids 89–105 in mVDR) [supplied as ascitic fluid, MAB 1360 (clone 9A7; Chemicon International Inc, Temecula, CA, USA) and used at 1/100 dilution] (see Fig. 3A for respective epitope localisation). VDR protein was detected using a peroxidase-conjugated AffiniPure F(ab')₂ fragment goat anti-rat IgG (IM0825, Immunotech, Marseille, France) as secondary antibody and enhanced with the enhanced chemiluminescence detection system (ECL Western Blotting Detection Reagents, Amersham). In vitro synthesised hVDR and hTRß1 were used as positive and negative controls of VDR detection.

Electrophoretic mobility shift assay

Electrophoretic mobility shift assays (EMSAs) were performed with nuclear extracts of 1,25-D3-treated Ob17 cells or with the recombinant receptors (Dace et al. 1999). For the DNA-binding reaction, the proteins were incubated at room temperature for 30 min in shift buffer (50 mM NaCl, 20 mM Tris–HCl, 1 mM EDTA, 20% glycerol, 1 mM MgCl₂, 2.5 mM DTT, pH 7.8) containing 40 ng/µl poly(dI-dC) and 20–40 fmol double-stranded VDRE or TRE oligonucleotides labelled with [³²P]dCTP by the Klenow fill-in method. For immunodetection of receptors (supershift or inhibition of binding to a VDRE), the mixture was preincubated for 1 h at room temperature with the appropriate antibody (anti-VDR as described above or anti-TR 150–166 (Daadi et al. 1995)) before the labelled oligonucleotides were added. Equivalent amount of either non-relevant ascitic fluid or preimmune IgGs were applied as controls. The oligonucleotides used were a DR3-type mouse osteopontin VDRE (Liu & Freedman 1994) (5' gatccACAAGGTTCACGAGGTTCACGTCTg 3'), a DR3-type rat osteocalcin VDRE (Demay et al. 1990), inverted palindrome IP7-type VDREs designed by Schräder et al.(1994) and, as described previously (Dace et al. 1999), a DR4-type rat malic enzyme TRE and an IP6-type lysozyme TRE (F2). DNA–protein complexes were resolved in non-denaturing 6% polyacrylamide gels in 0.5% TBE buffer (50 mM Tris–HCl, 42.5 mM H₃BO₃, 10 mM EDTA, pH 8.3). Gels were dried under vacuum and analysed electronically using a Fuji Bas 2000.

T₃ binding to recombinant T₃ receptors and to Ob17 nuclei in vitro

Recombinant VDR and TRs, obtained either by in vitro transcription–translation or by partial purification from bacterial production (TR1), were co-incubated in the buffers selected for analysis of T₃ binding (Daadi et al 1995) or 1,25-D3 binding (TKEMo). Incubates were adjusted to equal concentrations of rabbit reticulocyte lysate with unprogrammed lysate, and supplemented or not with 2.5 nM 1,25-D3 (T₃ binding analysis) or 10 nM T₃ (1,25-D3 binding analysis). After 30 min at 0 °C, aliquots were incubated with increasing concentrations of either radiolabelled ligand, to determine specific binding site concentration and apparent affinity constant (K_a) in Scatchard analyses as described above for 1,25-D3 binding and previously for T₃ binding (Daadi et al. 1995). In a parallel approach, nuclei were purified from Ob17 preadipocytes cultured under standard conditions and incubated (50 µg DNA) with a nearly saturating concentration of [¹²⁵I]T₃ (1 nM), supplemented or not with 1,25-D3 (0.25–250 nM) or vehicle, to determine specific T₃ binding sites as previously described (Lenoir et al. 1996).

Statistics

Experimental data are reported as means ± S.E. The number of assays in each group is indicated in the legends of figures or in the text. Statistically significant differences from the control were analysed using the paired Student’s t-test. Differences in kinetics among treatments were evaluated by analysis of variance (ANOVA) followed by the Student–Newman–Keuls multirange test and considered significant at P < 0.05.

	Results

Top Abstract Introduction Materials and methods Results Discussion References

 
High-affinity receptor sites for 1,25-D3 in murine Ob17 preadipocytes decrease markedly during the differentiation process and are modulated by 1,25-D3 and T₃

High-affinity VDR sites were detected in Ob17 cell extracts (Fig. 1A, C), with a mean affinity constant (K_a) of 4.0 ± 1.3 x 10¹⁰ M^–1 and a maximum binding capacity of 443 ± 35 fmol 1,25-D3 per mg DNA (Fig. 1B) in confluent preadipocytes (day 0) cultured under standard conditions (n=8). Figure 1 (A, B) also indicates that the abundance of the VDR sites decreased markedly after confluence as the preadipocytes initiated their terminal differentiation, and could reach a fourfold lower abundance 5 days after confluence. This low level was maintained over the following 2 weeks, a period of time during which cells progressively differentiate. This significant decline in the number of VDR sites (2.7 ± 0.3-fold, n=8, between day 0 and day 9; Fig. 1B), without any significant change in the affinity for 1,25-D3 (Fig. 1A), was similar when insulin was included in the culture medium as an amplifier of the differentiation process (Fig. 1D).

View larger version (31K): [in this window] [in a new window]   Figure 1 Analysis of 1,25-D3 binding parameters in Ob17 cells during adipose differentiation and under the action of adipogenic agents. Ob17 cells were cultured in standard medium, supplemented or not as indicated, and homogenised in TKEMo buffer as described in Materials and methods. Aliquots (250 µl) of cell extracts were incubated with 0.06 nM [3H]1,25-D3 with or without increasing concentrations of 1,25-D3 (0.04–0.20 nM) before separation of bound (B) and free (F) hormone. (A) Scatchard representation of 1,25-D3 binding data in a single representative Ob17 cell series analysed after different periods of culture in unsupplemented standard medium (day 0=confluence). (B) Estimation of 1,25-D3 binding sites (VDR sites) in eight independent Ob17 cell series analysed on different days of culture in unsupplemented standard medium. Data are expressed as mean ± S.E. in the eight cell series at day 0 and day 9 and in two of these cell series at days 5, 8 and 14. ****P < 0.001, significant difference (paired Student’s t-test). (C) Scatchard analysis of 1,25-D3 binding data in a representative Ob17 cell series estimated at day 9 after culture in the presence of 1,25-D3 (0.2 nM), T3 (1.5 nM), 1,25-D3 (0.2 nM) + T3 (1.5 nM) or vehicle (control) added at day 0. (D) Abundance of VDR sites 9 days after confluence in Ob17 cells cultured in the absence (day 9 control) or presence of different agents added at day 0 or at day 7 as indicated. The data are expressed as (mean ± S.E.) percentage of the respective day 9 control data values in four cell series cultured for 9 days with either 1,25-D3 or T3 or according to the depicted drug application schedule in two additional assays in these cell series. ***P < 0.01 and ****P < 0.001, significant differences compared with day 9 control. The day 0 abundance of VDR sites in these four cell series is also estimated as percentage of the respective day 9 control level and depicted by the horizontal solid and dashed lines (mean ± S.E.).

VDR mRNA level is not significantly modulated during Ob17 cell differentiation and by 1,25-D3 but is decreased by T₃

In northern blot analyses using a PCR-designed murine cDNA probe [encompassing codons 125–252 in mVDR (Kamei et al. 1995) and perfectly matching the mouse VDR sequence] a single band was detected at the expected size (4.6 kb) (Dace et al. 1997). The level of VDR mRNA has now been estimated in several series of cells cultured either under standard conditions or in a non-adipogenic culture medium (stripped FBS) complemented or not with 1,25-D3, T₃, or both, at day 0 (Fig. 2A). In parallel, semiquantitative RT-PCR assays were also performed using the same cDNA probe (Fig. 2B). Interestingly, the relative abundance of VDR mRNA did not significantly change during the adipose differentiation of cells cultured under unsupplemented conditions. This contrasts with the sharp decrease in the number of VDR sites described in Fig. 1. This discrepancy between the extents of variation in the number of VDR mRNA and VDR sites was also evident after addition of 1,25-D3 (0.2 nM) to cells at day 0, as the chronic presence of 1,25-D3 did not significantly increase the VDR mRNA level analysed at day 9. In contrast, addition of T₃ at day 0 always decreased the relative abundance of VDR mRNA analysed at day 9. The actions of T₃ on VDR mRNA level and on VDR site number thus seem to be in good correlation.

View larger version (21K): [in this window] [in a new window]   Figure 2 Comparative estimation of VDR mRNA level by northern blot and RT-PCR in Ob17 cells during adipose differentiation and under the action of 1,25-D3 or T3. Ob17 preadipocytes were cultured in the presence of normal FBS or adipogenic agent-stripped FBS and treated for 5 days or 9 days with 1,25-D3 or T3 added at day 0 and as indicated. Total RNA was extracted and used for (A) northern blot analyses (performed on duplicate cell series) or (B) semiquantitative RT-PCR analyses (assayed on three or four cell series and expressed as mean ± S.E.) as described in Materials and methods. The data are expressed as percentage of values obtained at day 0. At day 0 the relative abundance of VDR mRNA was not significantly different whether culture conditions were with normal or stripped FBS.

A VDR mRNA variant, newly identified in the mouse Ob17 cells, cannot explain the discrepancies between VDR mRNA and VDR site modulation patterns

In the rat, a VDR mRNA variant (rVDR1) that differs from the canonical VDR (VDR0) in its 3'-coding and 3'-non-coding terminus has recently been described (Ebihara et al. 1996). This rVDR1 mRNA is generated through alternative splicing at codon 337 and the retention of intron 8. We demonstrate here that a similar VDR1 mRNA variant is also expressed in mouse cells. Primers were designed on each side of the putative rat-equivalent exon 8–exon 9 junction in the mouse VDR cDNA sequence (Fig. 3A, primers 1 and 2). These primers allowed the production of a 1.35 kb DNA fragment by PCR amplification of mouse genomic DNA. The sequence of this mouse intron is closely related to that of rat intron 8 (78.7% of identity) (Fig. 3C). An intronic primer 3 was designed. Both 1–2 and 1–3 sets of primers were used in separate RT-PCR assays on Ob17 cell mRNA and each one detected a single band at the expected size (189 bp and 167 bp for primers 1–2 and 1–3 respectively) and with the expected sequence. Assuming that a TGA codon in mouse intron 9 is a stop codon, the deduced mVDR1 protein sequence (370 amino acids, molecular mass 41.6 kDa). is closely related to rVDR1, with only two amino acid changes and 13 additional amino acids at the C-terminus (Fig. 3D). A VDR1-type mRNA variant is thus expressed in the Ob17 cells. This variant was also detected in other mouse tissues (kidney, liver; data not shown). In duplex RT-PCR amplifications of both mVDR0 and mVDR1 from Ob17 cell mRNAs, the VDR1 transcript level appeared to be markedly lower than that of VDR0 transcript (five- to sevenfold lower; Fig. 3B).

A comparative study of the VDR0 and VDR1 mRNA levels was then performed at different stages during adipose differentiation of Ob17 cells (Fig. 4A). The initially selected set of primers a–b, and the new sets 1–2 and 1–3 were used in semiquantitative RT-PCR assays for the specific estimation of VDR(0+1), VDR0 and VDR1 respectively. In this study, in order to obtain a better evaluation of early events in preadipocytes, the Ob17 preadipocytes were synchronised in their late proliferation stage and further cultured in the presence of normal or stripped FBS. The stripped culture medium was also supplemented with either 1,25-D3 (0.2 nM) T₃ or (1.5 nM) at day – 1 to trigger adipogenesis. In agreement with the results presented in Fig. 2, the first series of panels in Fig. 4A shows that levels of the VDR(0+1) transcript were poorly and not significantly modulated during adipogenesis and under the action of 1,25-D3. In contrast, and as suggested from results in Fig. 2, T₃ always decreased the level of these transcripts. The negative effect of T₃ was progressive during differentiation, from moderate in early stages to high and significant at day 14 (60% decrease). It was also significant when comparing treated and untreated cells at day 14 (Fig. 4A; P < 0.01). The second and third series of panels in Fig. 4A indicate that the expression of distinct VDR0 and VDR1 transcripts roughly follows the same patterns as VDR(0+1) during adipose differentiation and under the action of 1,25-D3 and T₃. The extent of adipose differentiation was checked through the increase in C/EBP transcripts, which is correlated to the progressive accumulation of lipid droplets within the cells (fourth series of panels in Fig. 4A). C/EBP expresssion confirmed that adipogenesis was blunted in the presence of stripped FBS and restored by the early addition of either T₃ or 1,25-D3.

View larger version (26K): [in this window] [in a new window]   Figure 4 Expression of mRNA for (A) VDR(0+1), VDR0, VDR1 and C/EBP and (B) TRß1, throughout Ob17 cell differentiation and under the action of 1,25-D3 and T3. Ob17 cells were synchronised using a serum deprivation method and grown in a medium containing 10% normal FBS or adipogenic agent-stripped FBS. One day before confluence (day –1), 1,25-D3 (0.2 nM) or T3 (1.5 nM) was added to the culture medium containing stripped FBS. Total RNA was extracted from cells on various days from day –1 to day 14 as indicated and was used in semiquantitative RT-PCR assays. (A) The sets of primers used for VDR(0+1), VDR0 and VDR1 cDNA amplifications are a–b, 1–2 and 1–3 respectively (see Fig. 3). The extent of Ob17 cell adipose differentiation was estimated by following the accumulation of C/EBP mRNA throughout, as shown in the fourth series of panels. (B) Temporal changes in the relative abundance of TRß1 transcripts in Ob17 cells during differentiation or under the influence of 1,25-D3 and T3. The level of each mRNA was estimated and expressed relative to the level of ß-actin mRNA. The ratio for the level of each transcript relative to ß-actin was normalised to 1 on day –1. Results are expressed as means ± S.E. of three independent cell series (each analysed at least four times). *Significant differences in VDR or TRß1 mRNA levels relative to that on day – 1 (P < 0.05; ANOVA with Student–Newman–Keuls test).

VDR protein is more abundant but less immunoreactive within high molecular mass nuclear oligomeric complexes in preadipocytes than in adipocytes

In western blot analyses of cellular and nuclear proteins of Ob17 cells, anti-VDR antibodies, directed against VDR sequences either C-terminal (Ab C-t) or internal (mAb), detected the same 52 kDa band (Fig. 5A). This immunoreactive band migrated in the same way as did a recombinant hVDR and an in vitro-synthesised ³⁵S-labelled hVDR (Fig. 5A, lanes 6 and 18). The abundance of the VDR in nuclei was enhanced when cells were pretreated with 1,25-D3, which triggered the nuclear translocation of VDR as expected (Racz & Barsony 1999), and this occurred both at confluence and at day 9 (1,25-D3 added at day – 2 for 2 or 11 days or added at day 7 for 2 days) (Fig. 5A, lane 2 compared with lane 1 and lane 9 compared with lane 8 in preadipocyte nuclei; lanes 4 and 5 compared with lane 3 and lanes 11 and 12 compared with lane 10 in adipocyte nuclei). Remarkably, when equal amounts of nuclear or cell extracts were analysed (Fig. 5B), the immunodetected VDR was always found to be more abundant in preadipocytes (day 0) than in adipocytes (day 9); this was also seen in nuclei when 1,25-D3 was present during the cell culture (Fig. 5A). The modulations of VDR sites during the differentiation described above may thus reflect, at least in part, variations in the abundance of the VDR protein. The potential presence of the VDR1 protein isoform was also sought using the anti-VDR mAb that is directed against a common epitope in VDR0 and VDR1 (see Fig. 3A), but VDR1 was not detected within the expected size range (i.e. 41.6 kDa).

View larger version (15K): [in this window] [in a new window]   Figure 5 Western blot immunodetection of mouse VDR protein in Ob17 cells and nuclei during differentiation and under the action of 1,25-D3. Ob17 cells were grown in standard medium and harvested at the preadipocyte (PAd) stage (day 0; lanes 1, 2, 8, 9, 13, 14) or the adipocyte (Ad) stage (day 9; lanes 3–5, 10–12, 15–17). Cells were treated with vehicle or with 0.2 nM 1,25-D3 added 2 days before confluence (lanes 2 and 4, 9 and 11, 14 and 16) or 7 days after confluence (lanes 5, 12 and 17). (A) Nuclear proteins or total cell proteins (50 µg DNA equivalent/sample) were separated by 10% SDS-PAGE and immunodetected by western blot using rabbit polyclonal IgGs directed against a C-terminal sequence (20 amino acids) in VDR (Ab C-t) or a monoclonal antibody mAb with its epitope localised immediately C-terminal to the VDR DNA binding domain (amino acids 89–105). Lanes 6 and 7 show the detection of recombinant hVDR and the absence of detection of hTRß1, both synthesised in vitro using a reticulocyte lysate system. Lanes 18 and 19 show in vitro-synthesised 35S-labelled hVDR and TRß1 respectively, detected by electronic analysis of the membrane with a Fuji Bas 2000. The arrow indicates the position of VDR. (B) Total proteins in each nuclear or cell extract were controlled, after SDS-PAGE and transfer, by staining the membrane with Ponceau red, and are here shown for lanes 8–17. Molecular weight markers are shown on the left. The data obtained with preadipocyte and adipocyte cell extracts when using Ab C-t were similar to those presented here and obtained with mAb. Results are representative of four different cell series.

View larger version (38K): [in this window] [in a new window]   Figure 6 EMSAs of Ob17 cell nuclear extracts or recombinant receptors using the osteopontin VDRE and analysis of reactivity to anti-VDR antibodies. (A) Nuclear extracts (2.5 µg protein/lane) from preadipocytes (day 0) or adipocytes (day 9), pre-treated with 1,25-D3 from day -2, were incubated with a mouse osteopontin [32P]VDRE. For VDR immunodetection, antibodies (Ab) against VDR were added to the nuclear protein incubate before the addition of [32P]VDRE osteopontin. Incubation and electrophoretic separations of anti-VDR Ab (mAb and Ab C-t as described in Fig. 5 legend) were performed as described in Materials and methods. Controls were performed in the presence of equivalent protein amounts of either non-relevant ascitic fluid (Asc) or preimmune IgGs. A 250-fold excess of unlabelled VDRE osteopontin was added as competitor (Ab C-t+VDRE). On the left are shown the main numbered shifted bands in addition to the immuno-supershifted band (ssh). In the third series of lanes, recombinant hVDR and rRXR (8 fmol each, produced by in vitro transcription–translation in reticulocyte lysate) were incubated with [32P]VDRE osteopontin. The total amount of reticulocyte lysate was equalised in each assay by the addition of unprogrammed reticulocyte lysate (URL). The position of VDR–RXR heterodimers is indicated by the arrow. (B) Band density changes under the action of anti-VDR antibodies applied as described in (A). The density of each main individual band (ssh, 1, 2) was expressed as relative to total density in VDRE-shifted bands and normalised between paired control and assay lanes for equal total density including unbound VDRE. The specific antibody effect illustrated here is the difference between the data in control and paired assay lanes, for the supershifted band (increased by Ab) and for the main individual bands 1 and 2 (decreased by Ab). Data are expressed as percent of control. ‘Total’ decrease was estimated for bands 1+2+2b. PAd, Ad, preadipocyte and adipocyte nuclear extract respectively. Values are the means ± S.E. of five or six independent series. Significant differences compared with controls without antibody: *P < 0.05, **P < 0.02 (paired t-test).

As shown above, T₃ downmodulates the VDR sites at a pretranslational level. Conversely, as we reported previously (Lenoir et al. 1996), the downmodulation of the TR sites by adipogenic concentrations of 1,25-D3 could not be attributed to decreased amounts of TR1 mRNA and protein. A post-translational event through changes in potential TR–VDR heterodimeric interactions was then sought. EMSA studies were performed using Ob17 nuclear extracts with osteopontin VDRE and applying anti-TR1 antibodies as described above. Immunologically-related TR1 appeared to be present within the oligomeric band 1: anti-TR1 antibodies produced in preadipocytes, but not in adipocytes, a moderate but significant attenuation of band 1 density (decrease 90.6 ± 0.9% of control assays with unrelated IgGs; n=3, P < 0.001) (data not shown). Nevertheless, further EMSA studies using recombinant VDR and TRs with different VDREs or TREs in combination with anti-VDR or anti-TR antibodies did not bring any evidence that formation of TR–VDR heterodimers could have occurred, whether or not 1,25-D3 or T₃ (or both) were added (data not shown).

Subsequently, the possible occurrence of alterations in the ability of TR to bind T₃ as a result of thepresence of liganded VDR was sought in in vitro hormone binding assays. Recombinant TR1 or TRß1 was co-incubated with VDR and with or without one of the respective radio-inert ligands and then with the alternate radiolabelled ligand for Scatchard analyses of binding parameters. These studies did not detect any significant alteration in the binding parameters attributable to TR–VDR co-incubations, when considering either the affinity for the ligand or the maximum binding capacity (Fig. 7A, C). In parallel assays, nuclei purified from Ob17 cells were incubated with [¹²⁵I]T₃, with or without the co-addition of 1,25-D3 at different concentrations (0.25–250 nM); 1,25-D3 did not produce any decrease in levels of the T₃ binding site (Fig. 7B).

View larger version (14K): [in this window] [in a new window]   Figure 7 Absence of T3 binding site modulation by 1,25-D3 in co-incubations of recombinant TR and VDR and in nuclei of Ob17 cells. (A) hTR1 produced in Escherichia coli or in vitro synthesised hTRß1 were incubated with equimolar amount of in vitro synthesised hVDR (4 fmol) supplemented or not with 2.5 nM 1,25-D3. (B) Nuclei purified from Ob17 preadipocytes (50 µg DNA per assay) were incubated with a nearly saturating concentration of [125I]T3 (1 nM) in the presence or not of 1,25-D3 at the indicated concentrations. Specific binding of T3 was estimated as indicated in Materials and Methods and expressed relatively to control T3 binding in absence of 1,25-D3. (C) VDR was incubated with TR1 in the presence or not of 10 nM T3 (A, C). After co-incubations (30 min, 0 °C), [125I]T3 or [3H]1,25-D3 were added to determine binding parameters in Scatchard analyses performed as described in Materials and Methods. Values for TR and VDR sites in receptor co-incubations are expressed as relative to those obtained without addition of the alternate receptor and ligand.

1,25-D3 triggers, as efficiently as T₃, the early transient expression of TRß1 that precedes the burst of C/EBP gene expression during preadipocyte differentiation

A transient expression of the TRß gene and of TRß1 protein was previously demonstrated in Ob17 preadipocytes at an early step of their differentiation, preceding the expression of several adipogenic transacting factors such as PPAR and C/EBP. This expression of TRß1 was obliterated under non-adipogenic culture conditions (stripped FBS). Addition of T₃ to preconfluent cells partly restored the level of TRß1 transcript estimated at day 2 (Dace et al. 1999). As shown in Fig. 4B, which illustrates the RT-PCR analyses of TRß1 transcript expression, and most interestingly, not only T₃ (1.5 nM) but also 1,25-D3, applied at the suboptimal concentration of 0.2 nM, robustly triggered the early transient expression of TRß1 that was blunted in their absence. Under the same conditions, 0.5 nM 1,25-D3 or 0.2 nM 1,25-D3 plus 1.5 nM T₃ also triggered the expression of TRß1 analysed at day 4, both giving similar increments over the amount on day – 1 (2.8 ± 0.2-fold and 2.6 ± 0.6-fold respectively, n=2), these increments being similar to that obtained with 0.2 nM 1,25-D3 (3.0 ± 0.5-fold, n=3; Fig. 4B). In a concentration of 1 nM, 1,25-D3 was less efficient at increasing the expression of TRß1 (2.0 ± 0.2-fold, n=2), but it then produced, as we described previously, a great loss of cells and a decrease in the extent of morphological differentiation. In Fig. 4A, the fourth series of panels shows that the extent of this TBß1 expression was well correlated with the extent of expression of C/EBP, which is analysed as a control of cell commitment toward adipocytes. The chronic addition of 1,25-D3 (0.2 nM) markedly amplified, as did T₃ (1.5 nM), the level of C/EBP transcript in maturing adipocytes (Fig. 4A, fourth series of panels), whereas greater concentrations of 1,25-D3 (0.5 nM, 1 nM) were less efficient (data not shown), in agreement with the previously described potential adipogenic role of 1,25-D3. Regardless of these details, the ability of 1,25-D3 to trigger early expression of TRß1 discloses an additional molecular level at which the cross talk between the 1,25-D3 and T₃ signalling pathways can occur.

	Discussion

Top Abstract Introduction Materials and methods Results Discussion References

 
Expression of the VDR gene in Ob17 cells was initially established by the identification of VDR transcripts. The presence of VDR protein is now clearly identified by the detection of high-affinity 1,25-D3 binding sites and of immunoreactive protein using two different anti-VDR antibodies. The abundance of the VDR site and its affinity for 1,25-D3 appeared to be consistent with those reported for several other 1,25-D3-responsive tissues and cells (Brehier & Thomasset 1988, Sato & Hiragun 1988, Gensure et al. 1993, Zhao & Feldman 1993, Lee et al. 1994, Bland et al. 1997, Li et al. 1999). Interestingly, the present study, further, has demonstrated for the first time that the Ob17 cells (i.e. mouse cells) co-express the canonical VDR (VDR0) and a VDR1 mRNA variant that until now had been identified only in the rat and was here found to be closely related to the rat VDR1. This splicing variant may produce a VDR protein isoform shortened in its ligand binding domain, unable to bind 1,25-D3 and bearing the potential ability to antagonise some VDR0 transcriptional actions as reported for the rat VDR1 (Ebihara et al. 1996). These properties are reminiscent of those reported for a c-erbA2 splicing variant of TR1 mRNA, derived from the TR gene and acting as an antagonist of TR action (Lazar 1993). Nevertheless, unlike c-ErbA2, mRNA for which is generally more abundant than that of TR1, VDR1 mRNA is expressed to a markedly lesser extent than is that of VDR0 in both rat and mouse. Furthermore, here, the presence of a VDR1 protein could not be detected in the mouse Ob17 cells. Ebihara et al.(1996) were also unable to identify it in rat tissues, and its true biological role remains undefined.

The abundance of cell VDR sites was found to be downmodulated in a marked and sustained manner during adipose differentiation, consistent with what was reported for several other cell types that undergo a differentiation process in culture (Sato & Hiragun 1988, Zhao & Feldman 1993, Lee et al. 1994). Furthermore, the chronic addition of physiological concentrations of 1,25-D3 to early Ob17 preadipocytes largely prevented this loss of VDR sites; delayed addition of 1,25-D3 was less efficient. The variations in the levels of both VDR0 and VDR1 mRNAs remained moderate and not significant throughout the differentiation process with or without 1,25-D3, and were thus not correlated to changes in the numbers of VDR sites. Conversely, a correlation was found, for the most part at least, between variations in the number of VDR sites and in the abundance of nuclear VDR protein analysed in western blots. Prevention of the decrease in numbers of VDR sites by 1,25-D3 may reflect a stabilisation of the VDR protein upon 1,25-D3 binding, as already reported (Arbour et al. 1993). This may involve conformational changes hindering proteolytic cleavage sites (Nayeri & Carlberg 1997) and may implicate the ubiquitin–proteasome pathway, action of which on the VDR was found to be prevented by 1,25-D3 binding (Li et al. 1999). In Ob17 cells, changes such as a decrease in band 1 oligomeric complexes and the appearance of faster migrating bands occurred in the distribution of VDR-containing complexes on DNA during the differentiation process. A decreased band 1 in adipocytes may reflect changes in the extent or nature of interactions between the VDR and some co-factors or some components of a degradation machinery. Changes in the type of VDR–protein interactions may also be implicit in the observed limitation of VDR stabilisation by 1,25-D3 when added later during adipose differentiation.

In Ob17 cells, the VDR sites were also found to be modulated by T₃, another adipogenic agent in these cells. Unlike 1,25-D3 however, T₃ downmodulated the number of VDR sites beyond the usual downmodulation observed during differentiation. T₃ also largely impaired the protective effect exerted on these sites by the addition of 1,25-D3. This loss of VDR sites under the action of T₃ could roughly be correlated to a decrease in the abundance of VDR mRNAs (both VDR0 and VDR1), which occurred progressively during adipose differentiation and became significant at late stages. The negative effect of T₃ on the level of VDR mRNA would thus involve a pretranslational, most probably post-transcriptional, action.

As we reported previously, 1,25-D3 and T₃ both behave as adipogenic agents in mouse Ob17 preadipocytes, each one giving an optimal response in a physiological concentration range. Furthermore, each one interfered with the other’s action when applied simultaneously, exerting first synergy and then antagonism as their concentration was enhanced (Lenoir et al. 1996, Dace et al. 1997). The present study has now brought evidence for the existence, in these cells, of a cross-downmodulation of the respective receptor sites as a possible basis for the observed interplay between the adipogenic actions of 1,25-D3 and T₃. We have shown here that T₃ downmodulates the levels of VDR sites (and transcript). Alternatively, 1,25-D3 downmodulates the T₃ receptor sites (essentially the TR1 type), implying a post-translational event with apparent conservation of the abundance of TR1 protein (Lenoir et al. 1996). A sequestration of TR sites through a potential TR–VDR heterodimerisation event after binding of 1,25-D3 to VDR, as suggested previously (Schrader et al. 1994), could not be sustained by several of our present data. This was also refuted by other groups who attempted to explain the transrepression that a liganded TR or VDR exerts on the transcriptional activity of the alternate receptor (Garcia-Villalba et al. 1996, Raval-Pandya et al. 1998, Thompson et al. 1999). It is thus worth considering that the TRs, like the VDR and several other nuclear receptors, are generally engaged as heterodimers with RXR and within oligomeric complexes with transcriptional co-factors that can be shared between different receptors (Mangelsdorf & Evans 1995, Lemon & Freedman 1999). Certain co-factors may allow and stabilise receptor conformation that is optimal for ligand binding. This was suggested in previous analyses of purified recombinant VDR (Nakajima et al. 1993) and TR1 (Daadi et al. 1995), the ligand-binding property of which was markedly and specifically amplified by nuclear extracts. The possibility that a RXR had such a role was suggested, but could not be demonstrated, for TR (Daadi et al. 1995). Thus an as yet undefined nuclear co-factor shared between VDR and TR may be involved. A titration of this factor by VDR upon 1,25-D3 binding would then limit its availability for the TR and produce a loss of TR sites after the addition of 1,25-D3 to the cells. In a second type of hypothesis, the observed modulations of VDR and TR sites may involve the proteasome degradation pathway. This pathway has been implicated in the degradation of unliganded VDR (see above); it also plays a part in the degradation of the TRs, but in that case when in the liganded state (Dace et al. 2000). These degradative actions were mainly ascribed to conformational changes between liganded and unliganded status, exposing or not ubiquitinylation or cleavage sites; such events could be related once more to interaction or loss of interaction with shared protein factors. Moreover, our present findings further indicate that 1,25-D3 was unable to downmodulate the TR sites when applied to incubates of isolated Ob17 cell nuclei. This suggests that the decline in TR sites caused by the application of 1,25-D3 to whole cells requires functional nuclear–cytoplasmic interrelations. In this regard, several recent reports have described the existence of a rapid nucleocytoplasmic shuttling of both VDR (Prufer & Barsony 2002) and TRs (Baumann et al. 2001, Bunn et al. 2001). As TR-type T₃ binding sites could never be detected in Ob17 cell cytosols (Lenoir et al. 1996), this shuttling may represent a step at which a TR degradation could be initiated with a loss of T₃ binding activity.

In addition, and remarkably, in the interplay that is exerted between the 1,25-D3 and T₃ signalling pathways for preadipocyte differentiation, another molecular event may play a part: as we reported recently (Dace et al. 1999), 1,25-D3, in a physiological dose range, is able to trigger, as efficiently as T₃, the transient expression of the TRß gene that occurs at a very early preadipocyte step in differentiating Ob17 cells. The molecular level at which 1,25-D3 exerts its action to enhance the level of the TRß1 transcript remains to be defined in the light of better knowledge of the mouse TRß1 promoter. To date there has been no report as to any action of 1,25-D3 on the expression of TR genes. This action of 1,25-D3 may explain the synergistic effect that low concentrations of 1,25-D3 exert in the presence of T₃ (Dace et al. 1997).

In conclusion, the present findings demonstrate that, in Ob17 preadipocytes, the 1,25-D3/VDR signalling pathway is present and can interfere with that used by T₃, another adipogenic factor in these cells. Our data demonstrate the existence of a cross talk in Ob17 cells between the 1,25-D3/VDR and the T₃/TR pathways, which may be a basis for interpreting the interplay that 1,25-D3 and T₃ exert in their adipogenic action. Our present data also demonstrate that the cross talk can involve different mechanisms, including cross-modulations of ligand binding capacities by the alternate ligand acting at either a pretranslational level (action of T₃ on VDR sites) or a post-translational level (action of 1,25-D3 on TR sites), and cross-modulations of the expression of receptor transcripts, such as a sustained decrease of VDR by T₃ and an early transient increase in ß1-type TR induced by 1,25-D3.

	Acknowledgements

 
The authors are indebted to C Malezet-Desmoulin for sequence analyses and to J Bonne for protein electrophoresis.

	Funding

This study was supported by contract grant sponsors Conseil Régional Provence–Alpes–Côte d’Azur, Centre d’Etudes et d’Informations sur les Vitamines (Laboratoires Roche) and Fondation pour la Recherche Médicale.

	References

Top Abstract Introduction Materials and methods Results Discussion References

 
Arbour NC, Prahl JM & DeLuca HF 1993 Stabilization of the vitamin D receptor in rat osteosarcoma cells through the action of 1,25-dihydroxyvitamin D3. Molecular Endocrinology 7 1307–1312.[Abstract]

Baumann CT, Maruvada P, Hager GL & Yen PM 2001 Nuclear cytoplasmic shuttling by thyroid hormone receptors. Multiple protein interactions are required for nuclear retention. Journal of Biological Chemistry 276 11237–11245.[Abstract/Free Full Text]

Bismuth J, Lenoir C, Teboul M, Daadi M, Giorgilli G, Bonne J, Macchia E, Planells R & Torresani J 1995 Antibodies against restricted sequences in human c-ErbA hinge domain recognize differentially natural mammalian alpha- or beta-type triiodothyronine receptors and interfere differently with hormone binding. European Journal of Endocrinology 132 347–356.[Abstract]

Bland R, Sammons RL, Sheppard MC & Williams GR 1997 Thyroid hormone, vitamin D and retinoid receptor expression and signalling in primary cultures of rat osteoblastic and immortalised osteosarcoma cells. Journal of Endocrinology 154 63–74.[Abstract]

Brehier A & Thomasset M 1988 Human colon cell line HT-29: characterisation of 1,25-dihydroxyvitamin D3 receptor and induction of differentiation by the hormone. Journal of Steroid Biochemistry 29 265–270.[CrossRef][ISI][Medline]

Bunn CF, Neidig JA, Freidinger KE, Stankiewicz TA, Weaver BS, McGrew J & Allison LA 2001 Nucleocytoplasmic shuttling of the thyroid hormone receptor alpha. Molecular Endocrinology 15 512–533.[Abstract/Free Full Text]

Byrne IM, Flanagan L, Tenniswood MP & Welsh J 2000 Identification of a hormone-responsive promoter immediately upstream of exon 1c in the human vitamin D receptor gene. Endocrinology 141 2829–2836.[Abstract/Free Full Text]

Daadi M, Lenoir C, Dace A, Bonne J, Teboul M, Planells R & Torresani J 1995 Nuclear factors specifically favor thyroid hormone binding to c-ErbA alpha 1 protein (thyroid hormone receptor alpha) over-expressed in E coli. FEBS Letters 358 137–141.[CrossRef][ISI][Medline]

Dace A, Martin-el Yazidi C, Bonne J, Planells R & Torresani J 1997 Calcitriol is a positive effector of adipose differentiation in the OB 17 cell line: relationship with the adipogenic action of triiodothyronine. Biochemical and Biophysical Research Communications 232 771–776.[CrossRef][ISI][Medline]

Dace A, Sarkissian G, Schneider L, Martin-El Yazidi C, Bonne J, Margotat A, Planells R & Torresani J 1999 Transient expression of c-erbAbeta1 messenger ribonucleic acid and beta1 thyroid hormone receptor early in adipogenesis of Ob 17 cells. Endocrinology 140 2983–2990.[Abstract/Free Full Text]

Dace A, Zhao L, Park KS, Furuno T, Takamura N, Nakanishi M, West BL, Hanover JA & Cheng S 2000 Hormone binding induces rapid proteasome-mediated degradation of thyroid hormone receptors. PNAS 97 8985–8990.[Abstract/Free Full Text]

Dani C, Amri EZ, Bertrand B, Enerback S, Bjursell G, Grimaldi P & Ailhaud G 1990 Expression and regulation of pOb24 and lipoprotein lipase genes during adipose conversion. Journal of Cellular Biochemistry 43 103–110.[CrossRef][ISI][Medline]

Demay MB, Gerardi JM, DeLuca HF & Kronenberg HM 1990 DNA sequences in the rat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and confer responsiveness to 1,25-dihydroxyvitamin D3. PNAS 87 369–373.[Abstract/Free Full Text]

Ebihara K, Masuhiro Y, Kitamoto T, Suzawa M, Uematsu Y, Yoshizawa T, Ono T, Harada H, Matsuda K, Hasegawa T, Masushige S & Kato S 1996 Intron retention generates a novel isoform of the murine vitamin D receptor that acts in a dominant negative way on the vitamin D signaling pathway. Molecular and Cellular Biology 16 3393–3400.[Abstract]

Garcia-Villalba P, Jimenez-Lara AM & Aranda A 1996 Vitamin D interferes with transactivation of the growth hormone gene by thyroid hormone and retinoic acid. Molecular and Cellular Biology 16 318–327.[Abstract]

Gensure RC, Fish JM & Walters MR 1993 Dexamethasone downregulates vitamin D receptors in rat kidney, unmasking a high affinity binding site. Biochemical and Biophysical Research Communications 195 1139–1144.[CrossRef][ISI][Medline]

Hermey DC & Popoff SN 1995 Expression of the vitamin D receptor in the intestine and kidney of osteopetrotic rats. Endocrinology 136 4558–4564.[Abstract]

Jehan F & DeLuca HF 1997 Cloning and characterization of the mouse vitamin D receptor promoter. PNAS 94 10138–10143.[Abstract/Free Full Text]

Jones G, Strugnell SA & DeLuca HF 1998 Current understanding of the molecular actions of vitamin D. Physiological Reviews 78 1193–1231.[Abstract/Free Full Text]

Kamei Y, Kawada T, Fukuwatari T, Ono T, Kato S & Sugimoto E 1995 Cloning and sequencing of the gene encoding the mouse vitamin D receptor. Gene 152 281–282.[CrossRef][ISI][Medline]

Labarca C & Paigen K 1980 A simple, rapid, and sensitive DNA assay procedure. Analytical Biochemistry 102 344–352.[CrossRef][ISI][Medline]

Lazar MA 1993 Thyroid hormone receptors: multiple forms, multiple possibilities. Endocrine Reviews 14 184–193.[CrossRef][ISI][Medline]

Lee S, Clark SA, Gill RK & Christakos S 1994 1,25-Dihydroxyvitamin D3 and pancreatic beta-cell function: vitamin D receptors, gene expression, and insulin secretion. Endocrinology 134 1602–1610.[Abstract]

Lemon BD & Freedman LP 1999 Nuclear receptor cofactors as chromatin remodelers. Current Opinion in Genetics and Development 9 499–504.[CrossRef][ISI][Medline]

Lenoir C, Dace A, Martin C, Bonne J, Teboul M, Planells R & Torresani J 1996 Calcitriol down-modulates the 3,5,3' triiodothyronine (T3) receptors and affects, in a biphasic manner, the T3-dependent adipose differentiation of Ob 17 preadipocytes. Endocrinology 137 4268–4276.[Abstract]

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. Molecular Endocrinology 13 1686–1694.[Abstract/Free Full Text]

Liu M & Freedman LP 1994 Transcriptional synergism between the vitamin D3 receptor and other nonreceptor transcription factors. Molecular Endocrinology 8 1593–1604.[Abstract]

MacDougald OA & Lane MD 1995 Transcriptional regulation of gene expression during adipocyte differentiation. Annual Review of Biochemistry 64 345–373.[CrossRef][ISI][Medline]

Malloy PJ, Pike JW & Feldman D 1999 The vitamin D receptor and the syndrome of hereditary 1,25-dihydroxyvitamin D-resistant rickets