| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,25-Dihydroxyvitamin D3 Transrepresses Retinoic Acid Transcriptional Activity via Vitamin D Receptor in Myeloid Cells
Laboratoire de Biologie Cellulaire Hématopoïétique, Institut National de la Santé et de la Recherche Médicale Equipe Mixte Inserm 00-03, Institut Universitaire d’Hématologie, Paris 7, Hôpital Saint-Louis, 75010 Paris, France
Address all correspondence and requests for reprints to: Christine Chomienne, M.D., Ph.D., Laboratoire de Biologie Cellulaire Hématopoïétique, INSERM Equipe Mixte Inserm 00–03, Institut Universitaire d’Hématologie, Paris 7, Hôpital Saint-Louis, 75010 Paris, France. E-mail: christine.chomienne{at}sls.ap-hop-paris.fr.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-dihydroxyvitamin D3). Here we show that the stimulatory effect of 1,25-dihydroxyvitamin D3 on the production of monocytic colonies (CFU-M) is accompanied by a repression of granulocytic colony (CFU-G) production. We further demonstrate that in bipotent HL-60 myeloid cells as in purified human CD34+ myeloid progenitor cells, the 1,25-dihydroxyvitamin D3-induced monocytic differentiation is concomitant with a direct inhibition of the RA-transcriptional activity. Indeed, a transrepression of the RARß RA-target gene promoter via formation of a nuclear complex involving VDR was identified in vitro and in vivo. The fact that binding of RXR-RAR on DR3 is not observed suggests that contrary to RA-induced granulocytic differentiation, 1,25-dihydroxyvitamin D3-mediated monocytic differentiation requires positive and negative transcriptional controls both likely mediated by the RXR-VDR heterodimer. These novel findings implicate that 1,25-dihydroxyvitamin D3 exerts a dominant negative effect on the RA-dependent granulocytic commitment of human bone marrow cells via repression of the RA-target gene promoters. Hence, the transcriptional response to RA and 1,25-dihydroxyvitamin D3 in myeloid cells depends on a complex combinatory pattern of interaction among different nuclear receptors with DNA. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-dihydroxyvitamin D3, have profound effects on cellular differentiation during embryonic development and adult tissue differentiation (1, 2, 3). Moreover, 1,25-dihydroxyvitamin D3 controls calcium homeostasis and bone metabolism (4, 5). RA and 1,25-dihydroxyvitamin D3 bind to specific ligand-inducible transcription factors [retinoic acid receptors (RARs) and vitamin D receptor (VDR)] that activate or repress expression of target genes. RARs and VDR are members of the steroid/thyroid hormone nuclear receptor superfamily that, together with retinoid X receptors (RXRs), form a subgroup of structurally homologous proteins with overlapping functions due to their ability to form functional heterodimers (6). All nuclear receptors exhibit a common modular structure consisting of six distinct domains with a highly conserved DNA-binding domain and a moderately conserved ligand-binding domain. In addition to ligand recognition, these receptors are involved in ligand-dependent transactivation and bind to response elements as homo- or heterodimers (7, 8). Heterodimers bind to response elements with a specific polarity (7, 8, 9). Nucleotide spacing between the half-sites is equally determinant. Thus, VDR, thyroid hormone receptors (TRs), and RARs bind in vitro preferentially to direct repeats (DRs) spaced by three, four, or five nucleotides, respectively (8, 9). Naturally occurring response elements are usually composed of at least two copies of the consensus AGGTCA half-site motif arranged as DRs, palindromes, or inverted repeats. Although the orientation and spacing of the half-sites can determine selective transcriptional responses, specificity is not total, and response elements have been found to bind different heterodimers with high affinity (10); thus, RXR-VDR binds not only DR3 but also DR4, DR5, and even palindromes both in vivo and in vitro (11, 12, 13, 14, 15). The specific transcriptional control by nuclear receptors equally involves chromatin remodeling including acetylation of histone proteins as directly controlled by nuclear receptor-associated proteins, also referred to coactivators or corepressors (16, 17). Thus, the repressive activity of apo-receptors is related to their interactions with corepressors such as nuclear receptor corepressor and silencing mediator of retinoid and thyroid hormone receptor (SMRT) which recruit histone deacetylases (18, 19). Ligand binding triggers a conformational change of nuclear receptors causing the release of corepressors (nuclear receptor corepressor and SMRT) followed by the recruitment of various coactivator proteins, some of which harbor intrinsic histone acetyltransferase activity, allowing further modification of chromatin and recruitment of various protein complexes including the basal transcription machinery (16).
Normal hematopoiesis involves multiple steps of differentiation from the pluripotent stem cell to the determined progenitor cells, which in turn differentiate along lineage-specific pathways to produce the mature circulating blood cells. These different steps are regulated by a precise combinatory control involving specific transcription factors in response to a specific combination of extracellular signals (20). One of these progenitor cells gives rise to a monocytic or a granulocytic precursor, which then differentiates into mature monocytes or granulocytes. Commitment to either granulocytic or monocytic lineage is triggered by specific extracellular signals involving cytokines [granulocyte-colony stimulating factor (G-CSF), granulocyte-macrophage (GM)-CSF, and macrophage (M)-CSF] or nuclear receptor vitamin metabolites (RA) and 1,25-dihydroxyvitamin D3. These signals act on the synthesis and activation of transcription proteins driving the myeloid precursor cell to either the granulocytic or monocytic lineage (21, 22). Although specific growth and differentiation factors of these lineages have been identified (22, 23), elucidation of the molecular control that determines the switch to either the monocytic or granulocytic differentiation is only at the beginning. Specific transcription factors such as CCAAT-enhancer binding proteins (C/EBPs), PU.1, GATA-1, and acute myeloid leukemia 1 are essential for hematopoiesis and are controlled by specific patterns of protein expression and interactions (22, 24). For RA and 1,25-dihydroxyvitamin D3, much remains unclear, although their participation in the control of myeloid differentiation both in normal myeloid progenitors (1, 2, 25, 26, 27, 28) as in leukemic cell lines (29, 30, 31, 32, 33) has been well established. 1,25-Dihydroxyvitamin D3 induces monocytic differentiation in the HL-60 myeloblastic cell line (34, 35, 36, 37, 38, 39, 40) and the U-937 monoblastic cell line (41, 42, 43, 44, 45). By contrast, RA induces granulocytic differentiation of the HL-60 cell line (34, 35, 36, 40, 46, 47) and monocytic differentiation of the U-937 cell line (38, 48, 49). Thus, interactions between 1,25-dihydroxyvitamin D3 and RA signaling pathways at the common myeloid progenitor are likely important for commitment to either monocytic or granulocytic differentiation.
So far, no studies have addressed the molecular mechanisms involved in the myelo-monocytic switch by nuclear receptors. Here, we investigate the molecular mechanism responsible for the differential commitment of signaling pathways toward the monocytic or granulocytic lineage mediated by either 1,25-dihydroxyvitamin D3 or RA. To this end, we used purified normal human bone marrow (BM) cells and examined the effect of 1,25-dihydroxyvitamin D3 on RA-induced colony forming unit-granulocyte (CFU-G) or colony forming unit-macrophage (CFU-M) growth. We then studied the effects of these hormones both at the molecular and cellular level. 1,25-Dihydroxyvitamin D3 acts on monocytic differentiation through 1,25-dihydroxyvitamin D3-target gene promoters via RXR-VDR, and RA induces granulocytic differentiation through RA-target gene promoters through RXR-RAR. In this study, we further show that the 1,25-dihydroxyvitamin D3-induced monocytic differentiation is concomitant with a direct inhibition of the RA-transcriptional activity through a transrepression of RA-target gene promoters via VDR. Interestingly, RA-mediated granulocytic differentiation does not act as a transrepressor of the 1,25-dihydroxyvitamin D3 transcriptional activity. These novel findings suggest that 1,25-dihydroxyvitamin D3 exerts a dominant negative effect on the RA-dependent granulocytic commitment of human BM cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-Dihydroxyvitamin D3 Exerts a Dominant Negative Effect on the RA-Dependent Granulocytic Commitment of Normal Human BM Cells,25-dihydroxyvitamin D3), and the number and type of colony formed were determined after 14 d of culture (Fig. 1A
). The assay was performed in the presence of a specific conditioned medium that allows clear discrimination of colony forming unit-granulocyte (CFU-G) and colony forming unit-macrophage (CFU-M) colonies (50). In the presence of culture medium alone, BMC spontaneously produce more CFU-G than CFU-M (
3-fold). However, upon addition of nanomolar concentrations of 1,25-dihydroxyvitamin D3, a striking decrease of CFU-G colonies is noted (40% of control) although the number of CFU-G still remains superior to CFU-M (Fig. 1A). A concentration greater than 5 nM 1,25-dihydroxyvitamin D3 is needed to switch the bone marrow myeloid progenitors from granulocytic to monocytic differentiation. This concentration corresponds to the pivotal differentiation switch between CFU-M and CFU-G determination. At 10 nM, 50 nM, or more, 1,25-dihydroxyvitamin D3 provokes a significant increase of CFU-M formation, which then becomes superior to the number of CFU-G. These results suggest that 1,25-dihydroxyvitamin D3 is a potential factor of the cellular switch between granulocytic and monocytic differentiation, acting both as a biological repressor of CFU-G formation and an enhancer of CFU-M formation. On the other hand, RA is known to stimulate the formation of CFU-G and inhibit CFU-M formation, also in a dose-dependent fashion (1, 25, 50). To further investigate the myelo-monocytic differentiation switch triggered by 1,25-dihydroxyvitamin D3 and RA in human BMC, different concentrations of RA were added to a constant dose of 1,25-dihydroxyvitamin D3 and the number and type of colonies were calculated (Fig. 1, B and C). As depicted in Fig. 1B, when 1,25-dihydroxyvitamin D3 or RA is added alone to the culture medium, formation of granulocytic colonies (CFU-G) is inhibited by 1,25-dihydroxyvitamin D3 and enhanced by RA whereas formation of monocytic colonies (CFU-M) is enhanced by 1,25-dihydroxyvitamin D3 and slightly inhibited by RA (Fig. 1C). Combination of both RA and 1,25-dihydroxyvitamin D3 results in CFU-G inhibition, to an extent similar to that observed with 1,25-dihydroxyvitamin D3 alone. This latter result demonstrates that RA cannot overcome the 1,25-dihydroxyvitamin D3-dependent inhibition of CFU-G formation even at a high pharmacological concentration of 1 µM (Fig. 1B). Conversely, RA exerts a weaker inhibitory effect on 1,25-dihydroxyvitamin D3-dependent CFU-M formation (Fig. 1, B and C). Hence, 1,25-dihydroxyvitamin D3 is responsible for the predominant effect observed in CFU-G formation inhibition when both ligands are combined. These findings suggest that 1,25-dihydroxyvitamin D3 and RA have a distinct activity on myeloid commitment in which RA favors mainly commitment toward the granulocytic lineage by a stimulatory effect on CFU-G formation whereas 1,25-dihydroxyvitamin D3 is implicated in the commitment of the monocytic lineage by exerting both a repression of the RA-mediated granulocytic differentiation and a direct stimulatory effect on CFU-M formation.
|
,25-Dihydroxyvitamin D3 Represses the RA Signaling Pathway at the Transcriptional Level in Myeloid Cells,25-dihydroxyvitamin D3 in human myeloid progenitors, we first carried out transactivation assays in the bipotent myeloid HL-60 cell line. Like BMC cells, HL-60 cells differentiate into granulocytic cells in the presence of RA (34, 35, 36, 40, 46, 47) (Fig. 2B) or into monocytic cells in the presence of 1,25-dihydroxyvitamin D3 (34, 35, 36, 37, 38, 39, 40) (Fig. 2C) or RA and 1,25-dihydroxyvitamin D3 (Fig. 2D) (34, 35). Differentiation was assessed on morphological and functional criteria. Slides were stained with May-Grünwald-Giemsa stain. The nitroblue tetrazolium (NBT) reduction test, which measures both functional monocytic or granulocytic differentiation, was performed using standard methodology. The percentage of cells containing intracellular-reduced black formazan deposits was determined. HL-60 cells were treated with either RA or 1,25-dihydroxyvitamin D3 alone or in combination, and differentiation toward monocytic or granulocytic was assessed by the acquired capacity to generate an oxidative burst (NBT test), morphological criteria, and the expression of a specific monocytic differentiation antigen (CD14 antigen). As expected, after only 3 d of incubation with a low concentration of RA, the HL-60 cell population was committed toward the granulocytic lineage according to morphological criteria associated with the NBT reduction test (Fig. 2, A and B). A monocytic differentiation occurred upon addition of 1 nM or 10 nM 1,25-dihydroxyvitamin D3 according to morphological criteria associated with the CD14 antigen measurement (Fig. 2C). Only granulopoiesis was enhanced by RA alone (35% NBT-positive cells) whereas monocytic differentiation was enhanced by 1,25-dihydroxyvitamin D3 alone, from 16% NBT-positive cells (35% of CD14-positive cells) to 95% NBT-positive cells in the presence of both drugs (56% of CD14-positive cells) (Fig. 2, C and D). Thus, in HL-60 cells as in normal BMC cells, combination of RA and 1,25-dihydroxyvitamin D3 enhances and determines monocytic and not granulocytic commitment.
|
). In HL-60 cells, endogenous RARs allow RA to enhance transactivation of the RARß gene promoter in a dose-dependent manner (from 8-fold at 0.1 µM to 26-fold at 1 µM) [Fig. 3A and as previously reported (47, 52)]. 1,25-Dihydroxyvitamin D3 has no transactivation effect on the RARß gene promoter. However, when combined with RA, 1,25-dihydroxyvitamin D3 exerts an antagonistic effect on the RA-mediated transactivation. Indeed, 10 nM 1,25-dihydroxyvitamin D3 decreased luciferase activity measured in the presence of 1 µM RA alone (from 26-fold induction to a 11-fold induction) (Fig. 3A). Thus, in a bipotential myeloid cell, 1,25-dihydroxyvitamin D3 prevents the transcription of an RA-target gene.
|
,25-dihydroxyvitamin D3 response element (VDRE) (6xDR3) cloned in front of a luciferase reporter gene. At 0.1 µM and 1 µM, RA had no effect on the VDRE reporter gene, whereas at 10 nM 1,25-dihydroxyvitamin D3 enhanced dramatically the luciferase activity (Fig. 3B
). This result suggested that the 1,25-dihydroxyvitamin D3-induced monocytic differentiation of HL-60 cells could be mediated via VDRE. Interestingly, addition of a high concentration of RA to 1,25-dihydroxyvitamin D3 did not inhibit the transcriptional activity of 1,25-dihydroxyvitamin D3 on VDRE (Fig. 3B).
Using a real-time quantitative PCR (TaqMan, ABI Prism 7700, Sequence Detector, Applera, Applied Biosystems, Courtaboeuf, France) and normalization of the mRNA values to 18S rRNA, we observed a correlation between the transcriptional activity of the RARß promoter and the endogenous RARß expression (Fig. 3C). Indeed, the expression of RARß increased 10-fold within 6 h of RA treatment of HL-60 cells (Fig. 3C), whereas 1,25-dihydroxyvitamin D3 had no effect on RARß expression. Interestingly, combination of both ligands decreased the RARß expression (Fig. 3C). Thus, the observed effects of RA and 1,25-dihydroxyvitamin D3 on the RAREß-mediated transcription of the RARß promoter in HL-60 are equally reflected in the overall transcription of the RARß gene itself.
To confirm these findings in a more physiological cell context, we determined the necessary conditions to transfect the pluripotential CD34+ stem cells (Fig. 4). As in HL-60 cells, RA enhanced the transactivation of the RARß promoter gene by 10-fold (Fig. 4A). No luciferase activity was detected in the presence of 1,25-dihydroxyvitamin D3 (Fig. 4A). When 1,25-dihydroxyvitamin D3 and RA were combined, an antagonistic effect on RA-mediated transactivation was again observed (Fig. 4A). Likewise, only 1,25-dihydroxyvitamin D3 enhanced transactivation from the VDRE promoter gene (Fig. 4B) albeit less than in HL-60 cells, which are probably more committed than CD34+ cells. Last, RA did not antagonize the 1,25-dihydroxyvitamin D3-mediated transactivation (Fig. 4B). Altogether, these data show that in bipotent myeloid precursors, 1,25-dihydroxyvitamin D3 inhibits the RA-mediated transactivation by a transcriptional repression of RA-target genes, and RA-mediated granulocytic differentiation does not act as a transrepressor of the 1,25-dihydroxyvitamin D3 transcriptional activity.
|
,25-dihydroxyvitamin D3 induces differentiation toward the monocytic lineage by inhibiting signaling through the RARE and activates gene transactivation via VDRE.
VDR Is Part of the DR5-Containing Promoter RARß in HL-60 Cells Treated by 1,25-Dihydroxyvitamin D3 in Vitro and in Vivo
To document the transcriptional control of 1,25-dihydroxyvitamin D3 on either VDRE or ßRARE promoter sequences, nuclear extracts from HL-60 cells were used for gel retardation assays. Nuclear extracts of HL-60 cells were prepared after 6 h incubation with either RA alone, 1,25-dihydroxyvitamin D3 alone, or combination of RA and 1,25-dihydroxyvitamin D3 (Fig. 5, A–D). When the 10 nM 1,25-dihydroxyvitamin D3 HL-60-treated nuclear extracts were incubated with a 32P-labeled VDRE (DR3) oligonucleotide substrate, a novel retarded complex was detected (Fig. 5A, lane 4). The signal corresponding to this new retarded complex remained present even if weaker when 1 µM or 0.1 µM RA was combined with 10 nM 1,25-dihydroxyvitamin D3 (Fig. 5A, lanes 6 and 8). Lower concentrations of 1,25-dihydroxyvitamin D3 (1 nM) did not modify the pattern observed in control extracts (Fig. 5A, lanes 5, 7, and 9).
|
,25-dihydroxyvitamin D3-dependent retarded complex (Fig. 5A, lanes 4, 6, and 8; and Fig. 5B, lane 1, indicated by a star) contains at least VDR and RXR as it was completely supershifted by an anti-RXR antibody (Fig. 5B, lane 2, indicated by the arrow) and by an anti-VDR antibody (Ref.47 and Fig. 5B, lane 3, indicated by two arrows). As expected, the complex did not comprise RAR, as an anti-RAR antibody did not modify the pattern (Fig. 5B, compare lanes 4 and 1, indicated by a star).
When the nuclear extracts were incubated with the 32P-labeled ßRARE (DR5) oligonucleotide probe, at least four distinct retarded complexes were revealed (Fig. 5C, lane 1). As reported earlier, compared with control, RA does not induce additional complexes because of a spontaneous activation of the endogenous receptors by discrete RA concentrations present in the culture medium (Fig. 5C, lanes 2 and 3, and Ref.47). Interestingly, treatment with 10 nM 1,25-dihydroxyvitamin D3 alone (Fig. 5C, lane 4) induces a novel distinct slower migrating retarded complex, still present albeit weaker when in combination with RA (Fig. 5C, lanes 6 and 8) (indicated by a star). Here again, as observed with the DR3, lower concentrations of 1,25-dihydroxyvitamin D3 did not induce this complex (Fig. 5C, lanes 5, 7, and 9).
The RARE-bound complexes are shifted or inhibited, respectively, by anti-RXR or anti-RAR antibodies, confirming the binding of RXR and RAR to the RARE (Fig. 5D, lanes 2 and 3). VDR is not found bound to the RARE when HL-60 cells are treated with RA alone, as incubation with an anti-VDR antibody did not modify the pattern (Fig. 5C, compare lanes 4 and 1). The novel complex observed when HL-60 cells are treated with 1,25-dihydroxyvitamin D3 alone (Fig. 5C, lane 4; and Fig. 5D, lane 5, indicated by a star) corresponds to a RARE complex of which VDR, RXR, and RAR are part. Treatment with VD3 in combination with RA also revealed the presence of the same nuclear receptors (data not shown). Indeed, incubation with an anti-RXR antibody (Fig. 5D, lane 6) supershifts all complexes (indicated by arrow 1) whereas the anti-VDR shifts only the specific 1,25-dihydroxyvitamin D3-induced complex (arrow 3) as shown by its disappearance [arrow 2 (Fig. 5D, lane 7)]. The RARE bound complexes are inhibited by anti-RAR antibody, indicating the possibility of the presence of RAR in this complex (arrows 4) (Fig. 5D, lane 8). Interestingly, the pattern obtained with the anti-VDR antibody is similar whether the 1,25-dihydroxyvitamin D3-treated nuclear extracts are bound to an RARE (DR5) or a VDRE (DR3) (indicated by the two arrows 3, Fig. 5D, lane 7; and by the arrow, Fig. 5B, lane 3), further stressing that VDR is indeed part of this RARE-bound complex. Thus, we cannot rule out sequestration of the free RXR by VDR.
Because EMSA is accomplished between isolated nuclear proteins and DNA oligonucleotides, and is thus not a direct measure of in vivo protein-DNA interaction in the presence of complex chromatin structures, we performed chromatin immunoprecipitation (ChIP) assays using two specific antibodies against VDR and specific primers of the RARß promoter in whole-HL-60 cell lysates. DNA fragments that coprecipitated with VDR were purified upon reversal of protein-DNA cross-links and used as templates for amplification with the specific primers. The resulting PCR-DNA products were fractionated by agarose electrophoresis and visualized by ethidium bromide staining. Positive control corresponding to a DNA immunoprecipitation using an anti-transcription factor IIB (TFIIB) and a PCR performed with PCR primers flanking the TFIIB site of the constitutively active glyceraldehyde-3-phosphate dehydrogenase (GAPDH) promoter locus (e.g. DNA immunoprecipitated with anti-TFIIB) indicated the efficacy of reagents and protocol (data not shown). As shown in Fig. 5Ea, chromatin prepared from 1,25-dihydroxyvitamin D3-treated HL-60 cells immunoprecipitated with antibodies against VDR (lanes 3 and 4) was enriched for RARß promoter sequences by 2- to 3-fold compared with levels from untreated cells (lanes 1 and 2). This band migrated on an agarose gel to a position identical to the genomic DNA control (input) (Fig. 5Ea; lanes 5 and 6). Negative controls indicated that purification of immunoprecipitated DNA from untreated cells was relatively less efficient than purification of immunoprecipitated DNA from 1,25-dihydroxyvitamin D3-treated cells (Fig. 5b,c). ChIP analysis confirms that VDR is recruited to the RARß promoter upon 1,25-dihydroxyvitamin D3 stimulation in living HL-60 cells.
To quantify the protein level of RAR and VDR present in the HL-60 nuclear extracts used for the EMSA and ChIP studies, Western blot experiments were performed (Fig. 5F). Compared with controls, 6 h incubation with 1,25-dihydroxyvitamin D3 dramatically increased VDR nuclear levels (5-fold) which are maintained when RA is added to 1,25-dihydroxyvitamin D3 (Fig. 5F, lanes 4 and 6). In these conditions, RAR nuclear levels were constant and little modified by either RA, 1,25-dihydroxyvitamin D3, or the combination (Fig. 5F, lanes 2–6). By immunoblotting, the RXR nuclear protein level was not detected after 6 h incubation (Fig. 5F). After 18 h incubation, however, RXR was detected but not modulated by RA or 1,25-dihydroxyvitamin D3, alone or in combination (data not shown). In cytoplasmic extracts from HL-60 cells treated by RA and 1,25-dihydroxyvitamin D3 for 6 h, only VDR was detected but the level was not modulated by either drug alone or in combination (data not shown). These results suggest that induction of VDR by 1,25-dihydroxyvitamin D3 corroborates the presence of VDR in the nuclear extracts and in the nuclear complex bound to either DR3 or DR5. Thus, the 1,25-dihydroxyvitamin D3-induced VDR increase is probably a crucial factor in the molecular switch.
Our findings indicate that VDR interferes at the transcriptional level with the RA-dependent nuclear complex by interacting with the RARE transcriptional complex in myeloid cells treated with 1,25-dihydroxyvitamin D3.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-dihydroxyvitamin D3 active metabolites). RA and 1,25-dihydroxyvitamin D3 both act on the myelo-monocytic progenitor to determine granulocytic or monocytic differentiation, respectively. To date, the transcriptional basis of their molecular mechanisms in myelo-monocytic commitment remains poorly understood. In the present study, we have shown that the 1,25-dihydroxyvitamin D3-induced monocytic differentiation is correlated to an inhibition of the RA transcriptional activity via VDR, suggesting that 1,25-dihydroxyvitamin D3 exerts a dominant negative effect on the RA-dependent granulocytic commitment of myeloid cells (Fig. 6). As the opposite is not observed, these data suggest that, contrary to RA-induced granulocytic differentiation, 1,25-dihydroxyvitamin D3-mediated monocytic differentiation requires positive and negative transcriptional controls, both likely mediated by the RXR-VDR heterodimer.
|
Presence of DR3 and/or DR5 on the promoters can bring selectivity and flexibility to target gene activation and have important functional consequences. Indeed, in contrast to the usual polarity of 5'-RXR-VDR on a DR3, binding of RXR-VDR on DR5 results in two heterodimeric combinations of VDR with RXR, the 5'-RXR-VDR and 5'-VDR-RXR orientations. Binding occurs in both orientations on a DR5, suggesting that VDR and RXR interactions could generate two distinct dimerization interface or one very flexible dimerization interface (12). Nevertheless, the fact that no transactivation effect is observed in 1,25-dihydroxyvitamin D3-treated HL-60 cells when transfected by the DR5-containing promoter, RARß indicates that binding of either RXR-VDR or VDR-RXR to a DR5 could prevent the in vivo recruitment of the transcriptional machinery and thus abolish transcription.
VDR- or TR-mediated repression may also implicate titration of a common associated protein or co-activator(s). RXR is the first candidate as shared by numerous nuclear receptors such as RAR and TR. The molecular choice of RXR to heterodimerize and bind on the DR3 may be dictated by the relative ratio of the partner (i.e. VDR vs. RAR). In the breast cancer cell line MCF-7, VDR repression of T3-mediated transcription was suggested to result either from competitive binding on the TRE or RXR sequestration (65). In myeloid HL-60 cells, similar mechanisms may be involved because 1,25-dihydroxyvitamin D3-mediated monocytic differentiation requires positive and negative transcriptional controls both mediated by VDR, and sequestration of the free RXR may result in the differentiation switch. However, in 1,25-dihydroxyvitamin D3-treated HL-60 cells, VDR and RAR complexes are identified with similar intensities, suggesting that free RXR is sufficient for the formation of both RAR and VDR heterodimers.
A time-dependent increase of VDR protein levels of VDR was noted and paralleled the VDR levels bound to either the RARE or the VDRE. On the contrary, RAR or RXR protein levels were stable. The increase of VDR was 1,25-dihydroxyvitamin D3 dose dependent and not affected by RA treatment. None of the previous studies on VDR repression of transactivation have underlined the increased expression of VDR. Contrary to the RAR gene, which is directly controlled by RA through an RARE, no VDRE has been identified in the VDR gene (67, 68, 69), and posttranscriptional controls, such as inhibition of VDR degradation, may be involved.
In Cos or HeLa cells, RXR and VDR, in their apo-form, repress in vitro and in vivo the transcription of RA-target promoters, by recruiting corepressor complexes (NCo-R and SMRT) containing histone deacetylases that, in turn, induce a modification of chromatin not permissive for transcription (70, 71). VDR was also shown to bind additional corepressors such as Alien (72) or hairless (73). Independently of the VDR-corepressor complex function, a repressive function of VDR was identified via either direct binding of VDR to DNA sequences within promoter target genes (74, 75, 76) or competitive transcriptional activities resulting from heterodimer formation between VDR and other transcription factors (60, 77) such as acute myeloid leukemia 1 (70) or Pit-1 (62). Protein-protein interactions are often the key element in the hematopoietic cell determination such as Hox proteins and PU.1 cross-talks with other transcription factors involved in the control of erythropoiesis or myelopoiesis (21, 78, 79, 80). For example, GATA-2 at the stem cell level and GATA-1 in erythroid precursor cells repress PU.1-mediated transcription by inhibiting PU.1-cJUN interactions (46, 60, 80). In eosinophilic development, C/EBP
acts as a repressor of GATA-1-mediated transactivation and GATA-1-PU.1 synergy via protein-protein interactions through the C/EBP and/or GATA-binding sites (54). Furthermore, in Langerhans cell differentiation, C/EBP or C/EBPß acts as repressors of the PU.1 activity (55). C/EBP blocks the expression of c-JUN by inhibiting the transcriptional activity of c-JUN by protein-protein interaction and preventing its binding to its own promoter in granulocytic differentiation (58). In lymphoid differentiation, VDR was reported to exert a repressive function on IL-2 and GM-CSF genes (2, 68). In this context, as a monomer, VDR represses GM-CSF expression by competitive binding on overlapping NFAT-1 (nuclear factor of activated T cells 1) and activator protein 1 binding sites and interaction with c-JUN (68).
In conclusion, our findings provide another example of repression in myeloid cells in which 1,25-dihydroxyvitamin D3 signaling interferes molecularly through VDR at the transcriptional level with the RA transcription signaling pathway by forming a complex located on the RARE. A direct interaction occurs between RA- and 1,25-dihydroxyvitamin D3-signaling pathways that is dependent on VDR and the accessibility of 1,25-dihydroxyvitamin D3 and RA target genes. Identifying specific target genes in stem cells is essential to fully characterize both RA and 1,25-dihydroxyvitamin D3 signaling pathways. Such findings set the stage for future identification of molecular mechanisms of transcriptional and posttranscriptional repression regulating hematopoietic cell proliferation and lineage determination.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-Dihydroxy-vitamin D3 (1,25D) was obtained from Hoffmann-La Roche (Basel, Switzerland), dissolved in absolute ethanol at a concentration of 10–3 M, and stored at –20 C. Antibody (Coulter Corp., Hialeah, FL) directed against CD14 (IgG2a isotype) is a phycoerythrin-conjugated antibody. IgG2a is used as negative control in immunophenotyping assays. Antibodies anti-RAR and anti-RXR were kindly provided by Dr. C. Rochette-Egly and Pr. P. Chambon (52). The two polyclonal anti-VDR antibodies (sc-1008 and sc-1008x) provided by Santa Cruz Biotechnology, Inc. (Santa Cruz, CA) were used for Western blot and EMSA, respectively. The monoclonal anti-VDR antibody used for ChIP assay was provided by Chemicon International (Temecula, CA), and the polyclonal anti-VDR antibody was provided by Santa Cruz Biotechnology (sc1008). Rabbit polyclonal antibody directed against actin (A-2066) was from Sigma-Aldrich (Saint-Quentin Fallavier, France).
Cell Sources and Cell Culture
Normal bone marrow samples were obtained from the iliac bone of patients undergoing orthopedic surgery, with the donor’s informed consent. The mononuclear cell (MNC) fraction was obtained as previously described (50). For CD34+ cells, human BMCs were obtained from discarded filter sets from patients with solid tumors without BM involvement, undergoing BM harvest for transplantation. The CD34+ cells were separated using the immunomagnetic Miltenyi MACS system (Miltenyi Biotec France, Paris, France). The mean purity of the CD34+ cells from PBSC (peripheral blood stem cells) was 97.1 ± 1.5%. CD34+ cells were maintained in Iscove’s modified Dulbecco’s medium (Life Technologies, Gaithersburg, MD) supplemented with 15% heat-inactived fetal calf serum (BioWhittaker, Inc., Walkersville, MD) and penicillin (100 U/ml) (Life Technologies), streptomycin (100 µg/ml) (Life Technologies), and 2 mM L-glutamine (Life Technologies). Human myeloblastic HL-60 cells were maintained in RPMI 1640 medium (Life Technologies) supplemented with 15% heat-inactived fetal calf serum (BioWhittaker, Inc.) and penicillin (100 U/ml), streptomycin (100 µg/ml), and 2 mM L-glutamine. All cultures were incubated at 37 C in a 5% CO2 humidified atmosphere.
Clonogenic Progenitor Assay
Mononuclear cells were plated at 2 x 105 cells/ml in 3-mm Petri dishes. Cultures were prepared in 1 ml mixture containing Iscove’s modified Dulbecco’s medium with 15% fetal calf serum (BioWhittaker), 1% glutamine (Life Technologies), 1% antibiotics (Life Technologies), and 0.3% agar (Difco Laboratories, Detroit, MI). Colony-stimulating factors were provided with 5 µl human placenta-conditioned medium (Terry Fox Laboratory, Vancouver, British Columbia, Canada). Ligands were added at the beginning of the culture. The culture plates were incubated at 37 C and in 5% CO2 atmosphere. At d 14, colonies were scored with an inverted microscope, and different colony types (CFU-G and CFU-M) were morphologically discriminated.
Differentiation Assays
HL-60 cells were plated at 1 x 106 cells/ml and incubated in the presence or absence of RA and 1,25-dihydroxyvitamin D3, alone or in combination. Differentiation was assessed on morphological and functional criteria. Cell morphology was analyzed on cytospin slide preparations stained with May-Grünwald-Giemsa. NBT reduction was detected by formation of blue-black formazan deposits in the cell. Two hundred cells were counted to assess the percentage of NBT-positive cells. For immunophenotyping, 1 x 106 cells were harvested and washed twice with ice-cold PBS, resuspended in 200 µl of PBS containing 0.5% BSA. One microliter of phycoerythrin-conjugated CD14 antibody corresponding to 1:200 was added to the cell suspension; an isotype-matched antibody control was included for each sample as a control of nonspecific binding. Cells were incubated in the dark on ice for 30 min, and then washed twice with ice-cold PBS, 1% BSA and resuspended in 0.2 ml of this same buffer. Stained cells were collected by a FACScalibur 3CA (Becton Dickinson and Co., Franklin Lakes, NJ). The number of cells that were CD14 positive was expressed as a percentage of total cell number.
Transactivation Assay
HL-60 cells were transfected by electroporation, as previously described (52). In contrast to HL-60 cells, CD34+ cells were electroporated at 300 V, 950 µF. Cells were transfected with the luciferase reporter gene hRARß2-Luc or VDRE-Luc in the presence or absence of RA and 1,25-dihydroxyvitamin D3. All transfections were performed with 1 µg of the Tk-ß-galactosidase expression vector used as an internal standard. Cells were harvested 18 h after transfection, and a luciferase assay was performed by a standard procedure. All results are expressed as fold induction based on the basal activity of the reporter gene (arbitrarily set at 1) observed in the absence of any ligand. Consistent results were obtained in more than three similar experiments.
Real-Time RT-PCR Quantification
Total RNA was extracted from HL-60 cells. Briefly, total RNA was prepared by ultracentrifugation of a guanidine-isothiocyanate lysate through a cesium chloride cushion. The pellet was washed with 70% ethanol and dissolved in diethylpyrocarbonate-treated water. RNA was reprecipitated by 10% sodium acetate (3 M) and 2.5 vol ethanol, centrifuged, washed with 70% ethanol, redissolved in diethylpyrocarbonate-treated water, and then stored at –80 C. Integrity of RNA was assessed by migration on a formaldehyde-agarose gel. After RNA isolation, cDNA was prepared from each sample as described previously (81). cDNAs and an internal reference gene (18S) were quantified using a fluorescence-based real time detection method [ABI Prism 7700 Sequence Detection System (TaqMan); Applied Biosystems], as previously described (81). The primers and probe sequences used were as follows: for RARß: primers, 5'-CTAAATACACCACGAATTCCAGTGCTGA-3' and 5'-CAGACGTTTAGCAAACTCCACGATCTTA-3'; probe 6FAM-5'-TCCGACTGGACCTGGGCCTCTGGG-3'-TAMRA; the 18S primers and probe were provided by Applied Biosystems. For each sample, parallel TaqMan PCR were performed for each gene of interest and the 18S reference gene to normalize for input cDNA. The ratio between the values obtained provided relative gene expression levels for RARß.
EMSA
The EMSA procedures used were similar to those previously described (52). In addition to the nuclear extracts (2–5 µg), reaction mixtures contained 20 µl of binding buffer and the double-stranded DR5 oligonucleotide probe (37 bp) (30 ng) corresponding to the RARE of the RARß2 natural promoter (5'-GATCAGGGTTCACCGAAAGTTCACTCGCATATATTAG-3') (52) or the double stranded OPVDRE (5'-GATCGCTCGGGTAGGGTTCACGAGGTTCACTCGACT-3'). When indicated, antibodies corresponding to RAR, RXR, and VDR were added.
ChIP Assays
ChIP assays were performed according to the instruction described by Active Motif (Active Motif Europe, Rixensart, Belgium) in a ChIP assay kit. Briefly, HL-60 cells were grown at 0.5 million per ml and treated with ligands the day before the experiment. The cells (1 million per ml) were fixed with 1% formaldehyde in media for 8 min at room temperature to cross-link DNA with proteins. Then, the cells were washed twice with 1x PBS. Cells were resuspended in lysis buffer and sonicated. Cell debris was removed by centrifugation. After incubation with protein G saturated with salmon sperm DNA, the supernatants underwent immunoprecipitation with 3 µg VDR antibodies at 4 C overnight. Antibodies were recovered with protein G saturated with salmon sperm DNA and then washed. DNA-protein complexes were eluted with 100 µl elution buffer (1% SDS, 50 mM NaHCO3). The elutants of each sample were treated with 200 mM NaCl and RNase A at 65 C overnight to reverse protein-DNA cross-link and to remove RNA. Elutants were then deproteinized by proteinase K at 42 C for 2 h. Free DNAs were extracted by the DNA purification minicolumns from the kit. DNA was eluted in 100 µl H2O. Five microliters of each DNA were amplified in PCR by employing Taq DNA polymerase (Applied Biosystems, Foster City, CA) and primers forward (5'-TCATTCTGTGTGACAGAAGTAGTAGGAA-3') and reverse (5'-CCCCCCTTTGGCAAAGAATA-3') to detect the human RARß gene promoter. We adjusted the number of PCR cycles to 36 as to remain within the linear range of PCR amplification as recommended by Active Motif. As a negative control for ChIP, an IgG was used for Immunoprecipitation, and the RARß promoter was amplified by using specific primers with 40 cycles of PCR. As another negative control for ChIP, an IgG was used for immunoprecipitation, and a region of genomic DNA between the GAPDH gene and the chromosome condensation-related SMC-associated protein (CNAP1) gene was amplified by using specific primers, forward (5'-ATGGTTGCCACTGGGGATCT-3') and reverse (5'-TGCCAAAGCCTAGGGGAAGA-3') with 40 cycles of PCR, as described by Active Motif. As a positive control for ChIP, a TFIIB antibody was used for Immunoprecipitation, and the GAPDH promoter locus was amplified by using specific primers as described by Active Motif. As another positive control (input), sonicated lysates underwent reverse cross-link and DNA purification and were subject to PCR. PCR products were fractionated and visualized on 2% agarose-TAE (Tris acetate EDTA) gel containing ethidium bromide.
Western Blot Analysis
Nuclear extracts from HL-60 cells, obtained as previously described, were run in 12% polyacrylamide gels (52). Proteins were quantified by the BCA protein assay (Pierce Chemical Co., Rockford, IL). The proteins were transferred onto a nitrocellulose membrane and incubated either with a 1:200 anti-VDR or 1:500 anti-RAR antibody, respectively. Equivalent loading of lanes was controlled by a 1:4000 antiactin antibody. The proteins were identified by chemiluminescence.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
J.-N.B. and N.B. contributed equally to this work and should both be considered first authors.
Present address for F.G.: Division of Medical and Molecular Genetics, Guy’s, King’s and St. Thomas’ School of Medicine, London SE1 9RT, United Kingdom.
Abbreviations: BM cell, Bone marrow cell; BMC, bone marrow mononuclear cell; ChIP, chromatin immunoprecipitation; C/EBP, CCAAT-enhancer binding protein; CFU-G, colony forming unit-granulocyte; CFU-M, colony forming unit-macrophage; 1,25D, 1,25-dihydroxyvitamin D3; DR, direct repeat; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GM-CSF, granulocyte macrophage colony-stimulating factor; NBT, nitroblue tetrazolium; RA, all-trans-retinoic acid; RAR, retinoic acid receptor; RARE, retinoic acid response element; RXR, retinoid X receptor; SMRT, silencing mediator of retinoid and thyroid hormone receptor; TFIIB, transcription factor IIB; TR, thyroid hormone receptor; VDR, vitamin D receptor; VDRE, vitamin D response element.
Received for publication October 22, 2003. Accepted for publication July 22, 2004.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
,25-dihydroxyvitamin D3. Proc Natl Acad Sci USA 78:4990–4993, 25-dihydroxyvitamin D3 co-operate to promote differentiation of the human promyeloid leukemia cell line HL-60 to monocytes. Leukemia 8:806–815[Medline]
-dihydroxyvitamin D(3) or all-trans-retinoic acid. Exp Cell Res 266:126–134[CrossRef][Medline]
teau MT, Cabane S, Marti J 1991 Synergistic effect of retinoic acid and 1,25-dihydroxyvitamin D3 on the differentiation of the human monocytic cell line U-937. Leuk Res 15:1145–1152[CrossRef][Medline]
