Suppression of ANP Gene Transcription by Liganded Vitamin D Receptor
Involvement of Specific Receptor Domains
Abstract—We showed previously that liganded vitamin D receptor (VDR) effects a suppression of human atrial natriuretic peptide (hANP) gene-promoter activity in cultured neonatal rat atrial myocytes. In the present study, we have attempted to identify the structural domains of the VDR that are involved in mediating this suppression. We examined the effects of a series of VDR mutants on a cotransfected hANP promoter-driven chloramphenicol acetyltransferase (CAT) reporter. Neither the native VDR nor any of the mutants tested displayed inhibitory activity in the absence of the 1,25-dihydroxyvitamin D3 (VD3) ligand. Δ134, a deletant harboring solely the DNA binding region of the VDR, and L254G, a mutant shown to be defective in retinoid X receptor (RXR) heterodimer formation in other systems, were as effective as the native VDR in reducing promoter activity. HBD, a deletant containing only the hormone-binding domain of the VDR, and K246G, a point mutant that is defective in the activation function of the receptor, did not attenuate reporter activity. A similar activity profile was displayed when a positively regulated promoter containing a direct-repeat vitamin D responsive element (DR3-CAT) was examined in these cells. Liganded VDR, the Δ134 mutant, and liganded L254G effected increases in DR3-CAT activity of 2.5-, 2-, and 4-fold, respectively. Two nonhypercalcemic analogues of VD3 (RO 23–7553 and RO 25–6760) displayed the same inhibitory activity as VD3. These studies suggest that the inhibition of hANP promoter activity requires both the DNA binding and activation functions of the receptor but does not appear to require formation of a classic RXRα-VDR heterodimer.
The receptor for the secosteroid VD3 is a member of the nuclear receptor family of transcription factors. The VDR binds to specific DNA recognition sequences, termed vitamin D responsive elements, located in the promoter region of VD3 target genes. Liganded VDR associates with VDREs as a heterodimeric complex with RXR, or other related protein partners, to effect either activation or suppression of transcription of VD3-regulated genes.1
Although VD3 is traditionally thought to act primarily on bone and intestine to exert its biological effects in the whole animal, a growing number of studies have identified several “nontraditional” sites that possess VDR and respond to VD3 with alterations in gene expression. Vascular smooth muscle2 and myocardium3 4 5 are included in this latter group. Receptors for VD3 have been identified in cardiac myocytes,3 and occupancy of these receptors is associated with changes in the dynamics of calcium transport across the sarcolemmal membrane.4 An intriguing series of studies from Weishaar and Simpson5 suggested that VD3 deficiency is associated with elevations in blood pressure, alterations in cardiac contractility, and cardiac hypertrophy in a rodent model. The cardiac abnormalities were not linked to the elevations in blood pressure per se, since treatment with exogenous calcium corrected the hypertension but failed to reverse the cardiac hypertrophy.5
ANP is a cardiac hormone with potent natriuretic, diuretic, and vasorelaxant properties6 that render it an effective antagonist of other systems (eg, the renin-angiotensin system) that promote volume retention and elevations in blood pressure in the whole animal. ANP is produced in and secreted from atrial myocytes in the adult animal. Ventricular ANP gene expression is seen in late fetal and early neonatal life, but expression decays quickly in the postnatal period to the very low levels seen in the adult. Reactivation of ventricular ANP gene expression is, however, seen in a number of experimental models and clinical paradigms associated with hypertrophy of ventricular myocardium.7 This association is, in fact, so closely linked, that ANP gene expression has come to be viewed as one of the earliest and most reliable markers of the hypertrophic process. In neonatal rat cardiac myocytes, virtually every biochemical or physical perturbation that results in “hypertrophy” (ie, increased cell size, increased protein synthesis, and reorganization of sarcomeric structure) also leads to activation of ANP gene expression.
We recently have shown that liganded VDR at least partially inhibits ANP gene expression in cultured neonatal rat atrial8 9 and ventricular10 myocytes. In the latter instance, this inhibition is accompanied by reversal of a number of phenotypic changes associated with hypertrophy in this model. This, coupled with the well-documented association of vitamin D deficiency and myocardial hypertrophy in the rat,5 raises the possibility that VD3, or derivatives thereof, might be useful clinically in the management of a variety of cardiac disorders that are associated, often to their detriment, with hypertrophy of the ventricular myocardium. The present study was designed to identify those functional domains in the VDR that are responsible for its inhibitory effect at the level of the human ANP gene promoter.
[3H]acetyl CoA was purchased from DuPont–New England Nuclear Research Products. VD3 was obtained from Biomol Research Laboratory. 9-cis Retinoic acid was generously provided by A. Levin and RO 23–7553 and RO 25–6760 by M. Uskokovic, both of Hoffmann-LaRoche Inc. Other reagents were from standard commercial suppliers.
The construction of −1150 hANP-CAT11 and DR3-CAT12 as well as expression vectors for wild-type human VDR,13 RXRα,14 Δ134,15 K246G,16 and L254G,16 have been described. HBD was prepared as follows. A 30-mer oligonucleotide (GCCACTGCCAGGAATTCCCTGCCGGCTCAA) was used to create an internal EcoRI restriction site at amino acid residue 78 of the human VDR coding region. The resultant 1.8-kb cDNA, containing the hinge and hormone-binding domains only, was cloned into the EcoRI site of the pSG5 expression vector. The desired clone, which is transcribed and translated beginning with methionine 89 (HBD includes residues 89 to 427), was confirmed by both DNA sequencing and immunoblot analysis of expressed protein in extracts from transfected cells. Each of these mutants was used in cotransfection with the hANP-CAT reporter to identify which functional domains of the receptor are responsible for the observed inhibitory activity.
Cell Preparation, Transfection, and Culture
Atrial myocytes were collected from the upper one third of 1-day-old neonatal rat hearts by alternate cycles of trypsin digestion and mechanical disruption as described previously.17 Isolated cells were transfected by electroporation (280 V and 250 μF) with the DNA indicated in the individual figure legends before plating. All transfections were normalized for equivalent DNA content with pUC 18. Transfected cells were plated at a density of 2.5×106 cells per 10 cm2 and cultured in Dulbecco’s modified Eagle’s medium–H21 containing 10% enriched calf serum (Gemini Products) for 24 hours before switching to serum-substitute medium.18 Specific chemical additives were dissolved in ethanol and diluted 1:1000 from stock solutions into serum-free medium before addition to the cultures. Similar concentrations of ligand-free vehicle were used as controls and were without effect. Cells were harvested and lysed in 250 mmol/L Tris(hydroxymethyl)aminomethane/0.1% Triton X-100 72 hours after transfection. Protein concentration was measured using the Coomassie protein reagent (Pierce Biochemicals). Samples normalized for equivalent protein concentration were then assayed for CAT activity as described by Neumann et al.19
Data were analyzed using one-way ANOVA and the Newman-Keuls test for significance.
Liganded VDR effected an ≈50% reduction in the expression of a cotransfected −1150 hANP-CAT reporter in the atrial myocyte cultures. This reduction became statistically significant after 24 hours and was maximal after 48 hours, the longest time point examined in this study (Figure 1⇓). All future experiments were carried out at this 48-hour time point.
The reduction in hANP gene transcription was apparently independent of the hypercalcemic properties of VD3. As shown in Figure 2⇓, two nonhypercalcemic analogues of VD3 (ie, RO 23–7553 and RO 25–6760) were at least as effective as VD3 in inhibiting −1150 hANP-CAT expression, either alone or, even more dramatically, in the presence of cotransfected VDR. Both analogues and the natural VD3 compound yielded classic dose responses over the range of 10−8 to 10−10 mmol/L, with a suggestion that RO 25–6760 may be slightly more potent than the natural ligand.
In an attempt to define the structural requirements for the VDR-dependent inhibition, we compared the ability of four VDR mutants versus wild-type VDR to signal VD3-dependent transcriptional regulation in atrial myocytes. The schematic structure of each of these mutants is presented in Figure 3⇓. The Δ134 truncation contains the DNA-binding domain and a small portion of the hinge region from the human VDR,15 while HBD contains residues 89 to 427, lacking the DNA-binding domain of the receptor (Figure 3⇓) but including the hinge region and hormone-binding domain. K246G contains a lysine-to-glycine mutation at position 246 that generates a VDR that binds hormone normally and forms a heterodimeric complex (with RXRα) on the VDRE but transactivates at <10% of wild-type activity, ie, it is an activation function–deficient mutant.16 20 L254G contains a glycine-for-leucine substitution at position 254. This mutant is thought to be impaired in its ability to heterodimerize with RXRα or RXRβ on DR3-type, positive VDREs.16
The behavior of each of these mutants on a conventional VDRE (DR3-CAT) was examined in the transfected atrial myocyte cultures. As shown in Figure 4⇓, neither the wild-type receptor nor the mutants affected DR3-CAT in the absence of ligand. The addition of VD3 increased reporter activity slightly in the control samples, presumably because of low levels of endogenous VDR, and, more dramatically, in the VDR-, Δ134-, and L254G-transfected cells but had no effect on either HBD or K246G. Liganded RXRα alone led to only a minor increase in reporter activity (not statistically significant in this study). In contrast, RXRα amplified activation of the reporter gene by the liganded VDR and, to a lesser extent, by liganded HBD and K246G but had no effect on Δ134 (relative to VD3 treatment alone). Interestingly, the response to L254G decreased rather than increased after introduction of the liganded RXRα.
Turning our attention to the hANP gene promoter, the cotransfected wild-type receptor, as well as each of the mutants, had only a modest and statistically insignificant effect on −1150 hANP-CAT expression in the unliganded state (Figure 5⇓). In the presence of ligand, the wild-type VDR, Δ134, and L254G each displayed a VD3-dependent reduction in reporter activity of roughly equivalent magnitude. The HBD and K246G mutants, on the other hand, were devoid of such activity and even appeared to blunt slightly the suppressive effect of endogenous VDR.
Cotransfection of unliganded RXRα had little effect on the VD3-treated control cultures, implying that endogenous RXR activity is near optimal in the basal state, but it amplified the inhibition seen in those cultures transfected with wild-type VDR (Figure 5⇑). Unliganded RXRα provided no additional inhibition (versus VD3 alone) with the Δ134 or L254G mutants but did appear to effect an inhibition in the presence of the HBD or K246G mutants (as noted above, VD3 alone effected no inhibition in the presence of these mutants). The addition of 9-cis RA (a ligand for RXR) to the RXRα-transfected cultures led to a suggestion of a further increase in the level of inhibition (versus RXRα alone) in the control, Δ134-, HBD-, and L254G-transfected cultures; however, this difference was significant only in the L254G-transfected group (P<0.05 versus VD3+RXRα alone).
The data presented herein demonstrate the following: (1) the DNA-binding domain of the VDR is critical for its inhibitory effect on the hANP gene promoter, (2) preservation of the activation function of the VDR is also important for the inhibition, and (3) retention of the capacity for heterodimerization between the VDR and RXRα/RXRβ may not be critical for the inhibition. The latter finding suggests a novel mechanism for VDR-mediated suppression of hANP gene transcription, perhaps dependent on cardiac cell-specific factors, one of which may serve as a unique partner for VDR in DNA binding. Finally, inhibition of hANP gene transcription by occupied VDR also does not appear to require those molecular properties of the ligand that lead to hypercalcemia in vivo because oxacalcitriol9 10 and now RO 23–7553 and RO 25–6760, two compounds that also lack the hypercalcemic properties of VD3, were as efficacious as VD3 in promoting the repression (Figure 2⇑). This finding should provide impetus for further examination of these agents for potential efficacy in the control (ie, inhibition) of hypertrophy, without attendant hypercalcemic side effects, in vivo.
The precise mechanism underlying the liganded VDR–dependent suppression of hANP promoter activity remains elusive. Using gel mobility shift analyses, we (Jurutka et al, unpublished data, 1998) have as yet been unable to identify traditional binding of enriched VDR/RXRα heterodimers to the 1150 bp of the hANP promoter shown to be actively repressed in functional studies. Liganded VDR has been shown to use nontraditional mechanisms to regulate gene expression in a negative fashion21,22; however, in virtually every case, some evidence for association of the receptor with DNA has been forthcoming. It is conceivable that a low-abundance accessory protein required for stable complex assembly was not present at sufficient concentrations to promote VDR-DNA association in the in vitro binding studies performed to date. Alternatively, this may imply that the ANP gene is not a primary target of liganded VDR (ie, VD3, through VDR-RXR, positively regulates a second gene product, which in turn downregulates ANP gene-promoter activity). In favor of this possibility is the slow kinetics of the reduction in reporter activity (Figure 1⇑), the absence of VDR/RXRα binding activity alluded to above, and qualitative similarities in the effects of the various VDR mutants in inhibiting the ANP gene promoter (Figure 5⇑) versus activation of DR3-CAT (Figure 4⇑).
The use of the VDR mutants has provided us with some additional clues as to how the receptor may regulate transcription in the intact myocyte. On a DR3-CAT reporter, liganded VDR displayed the predicted response, effecting a significant increase in reporter activity above that seen with VD3 alone. The observation that the Δ134 mutant can support hormone-dependent transactivation (Figure 4⇑) and repression (Figure 5⇑) in an RXR-independent manner implies that Δ134 may be capable of binding to one of the VDRE half sites (presumably the one usually occupied by RXR) and promoting, or at least allowing for, hormone-dependent activation/repression by endogenous, wild-type VDR. This unusual, and nonphysiological, synergism between Δ134 and wild-type VDR has also been observed with other VDRE-driven reporter constructs (Hsieh et al, unpublished data, 1998). It is unlikely that the DBD acts independently, either as a monomer or dimer, to activate the VDRE or suppress the hANP gene promoter because each of these activities is ligand dependent and the DBD does not bind ligand. Furthermore, it lacks the activation function in the ligand-binding domain that is required for both stimulatory16 and inhibitory (Figure 5⇑) activity.
The results presented in Figure 4⇑ indicate that the HBD and K246G mutants were devoid of activity on DR3-CAT, as expected; however, the heterodimerization-defective mutant L254G was, if anything, more effective than wild-type receptor in activating the reporter. This would suggest that within the context of the atrial myocyte and under the experimental conditions used here, heterodimerization may not be required for activation of this VDRE. Alternatively, since the predominant RXR isoform in heart is RXRγ23 rather than RXRα or RXRβ as in most VD3 target tissues, it is conceivable that VDR heterodimerization in the myocyte (ie, with RXRγ) may not require the L254 residue for effective complex formation. This latter scenario would also explain why the L254G VDR mutant displays decreased activity after addition of liganded RXRα. Because heterodimerization of the RXRα isoform with L254G has been shown to be impaired,16 we assume that the transfected RXRα diverts a portion of RXRγ, the putative functional partner of native VDR, to form inactive RXRα-RXRγ “homodimers.”
With −1150 hANP-CAT (Figure 5⇑), the efficacy of the individual VDR mutants was much the same as that seen with DR3-CAT, albeit with an inhibitory rather than stimulatory effect. While none of the VDR mutants displayed any activity in the unliganded state, the wild-type VDR, Δ134, and L254G each reduced reporter activity in a ligand-dependent fashion. Again, this implies that formation of conventional heterodimers (ie, between VDR and RXRα) is not an absolute prerequisite for the inhibition, since the two mutants cited (Δ134 and L254G) are incapable of such associations. Our data do not, however, preclude the involvement of VDR-RXR heterodimeric complexes in this inhibition, particularly if RXRγ retains the ability to associate with L254G (see above). The ability of unliganded RXRα to amplify the effect of liganded wild-type VDR and its inability to amplify those of “liganded” Δ134 and L254G are supportive of this selective heterodimerization hypothesis.
In contrast, unliganded RXRα significantly promotes inhibition in the presence of the HBD and K246G mutants. The absence of inhibition in the presence of VD3 alone in these latter instances presumably reflects sequestration of endogenous RXR (eg, RXRγ) complexes by these mutant VDR proteins, each of which would be expected to retain its capacity for dimerization in the intact cell. Consistent with this interpretation is the observation (Figure 5⇑) that supplemental coexpression of RXRα results in significant VD3-dependent repression, in essence “rescuing” the endogenous VDR effect for the HBD and K246G mutants.
The addition of 9-cis RA to the RXRα-transfected cultures accentuated the inhibition seen with RXRα alone only in the presence of the L254G mutant (Figure 5⇑). This observation implies that the bulk of the RXR effects seen in Figures 4⇑ and 5⇑ are the result of heterodimeric interactions of apoRXRα with VDR and its mutants. There is likely a second RXRα-dependent effect that results from formation of hANP-inhibitory, ligand-dependent homodimeric complexes (ie, RXRα-RXRα), an effect that is largely obscured when VDR-dependent activity is fully expressed.
In conclusion, liganded VDR is a significant negative regulator of ANP gene expression in cardiac myocytes, and on the basis of this and earlier studies,5 8 9 10 we hypothesize that this receptor could play a broader role in the control of cardiac growth and hypertrophy. The molecular functioning of hormone-occupied VDR in governing ANP gene transcription differs from that in traditional bone mineral homeostatic systems in that neither the specific calcemic properties of the ligand nor classic heterodimerization with RXRα or RXRβ are required, yet both the DNA-binding domain and the activation function within the ligand-binding domain of VDR are obligatory. Parallel activity profiles of various VDR mutants for mediating the positive (ie, DR3-CAT) and negative (hANP-CAT) effects of VD3, combined with the slow onset of hANP promoter repression and our own inability to demonstrate association of VDR-RXRα complexes with the hANP gene promoter in vitro suggest that transcription may be suppressed through a secondary mechanism involving the induction of an intermediary repressor. Interestingly, this action of VDR in myocytes appears to be unique in that VDR may be operating in a novel heterodimeric complex with RXRγ or an as yet uncharacterized cardiac transcription factor. The excitement of the future will revolve around the identification of the myocyte-specific heterodimeric partner of VDR, as well as the putative induced repressor that, if the model is correct, binds to and attenuates transcription of the hANP promoter.
Selected Abbreviations and Acronyms
|hANP||=||human atrial natriuretic peptide|
|RXR||=||retinoid X receptor|
|VD3||=||1,25(OH)2 vitamin D3|
|VDR||=||vitamin D receptor|
|VDRE||=||vitamin D responsive element|
This study was supported in part by grants HL 35753, DK 33351, and DK 49604 from the National Institutes of Health. The authors are grateful to Karl Nakamura for preparation of the cardiac myocytes.
- Received January 20, 1998.
- Revision received January 29, 1998.
- Accepted February 3, 1998.
Haussler MR, Haussler CA, Jurutka PW, Thompson PD, Hsieh J-C, Remus LS, Selznick SH, Whitfield GK. The vitamin D hormone and its nuclear receptor: molecular actions and disease states. J Endocrinol. 1997;154:S57–S73.
Walters MR, Ilenchuk TT, Claycomb WC. 1,25-Dihydroxyvitamin D3 stimulates 45Ca2+ uptake by cultured adult rat ventricular cardiac muscle cells. J Biol Chem. 1987;262:2536–2541.
Weishaar RE, Simpson RU. Vitamin D3 and cardiovascular function in rats. J Clin Invest. 1987;79:1706–1712.
Brenner BM, Ballermann BJ, Gunning ME, Zeidel ML. Diverse biological actions of atrial natriuretic peptide. Physiol Rev. 1990;70:665–699.
Chien KR, Knowlton KU, Zhu H, Chien S. Regulation of cardiac gene expression during myocardial growth and hypertrophy: molecular studies of an adaptive physiologic response. FASEB J. 1991;5:3037–3046.
Li Q, Gardner DG. Negative regulation of the human atrial natriuretic peptide gene by 1,25-dihydroxyvitamin D3. J Biol Chem. 1994;269:4934–4939.
Wu J, Garami M, Cao L, Li Q, Gardner DG. 1,25(OH)2 D3 suppresses expression and secretion of atrial natriuretic peptide from cardiac myocytes. Am J Physiol. 1995;268:E1108–E1113.
Baker AR, McDonnell DP, Hughes M, Crisp TM, Mangelsdorf DJ, Haussler MR, Pike JW, Shine J, O’Malley BW. Cloning and expression of full-length cDNA encoding human vitamin D receptor. Proc Natl Acad Sci U S A. 1988;85:3294–3298.
Kliewer SA, Umesono K, Heyman RA, Mangelsdorf DJ, Dyck JA, Evans RM. Retinoid X receptor-COUP-TF interactions modulate retinoic acid signaling. Proc Natl Acad Sci U S A. 1992;89:1448–1452.
Whitfield GK, Hsieh J-C, Nakajima S, MacDonald PN, Thompson PD, Jurutka PW, Haussler CA, Haussler MR. A highly conserved region in the hormone binding domain of the human vitamin D receptor contains residues vital for heterodimerization with retinoid X receptor and for transcriptional activation. Mol Endocrinol. 1995;9:1166–1179.
Wu JP, LaPointe MC, West BL, Gardner DG. Tissue-specific determinants of human atrial natriuretic factor gene expression in cardiac tissue. J Biol Chem. 1989;264:6472–6479.
Neumann JR, Morency CA, Russian KO. A novel rapid assay for chloramphenicol acetyltransferase gene expression. Biotechniques. 1987;5:444–447.
Henttu PM, Kalkhoven E, Parker MG. AF-2 activity and recruitment of steroid receptor coactivator 1 to the estrogen receptor depend on a lysine residue conserved in nuclear receptors. Mol Cell Biol. 1997;17:1832–1839.
Demay MB, Kiernan MS, DeLuca HF, Kronenberg HM. Sequences in the human parathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor and mediate transcriptional repression in response to 1,25-dihydroxyvitamin D3. Proc Natl Acad Sci U S A.. 1992;89:8097–8101.
Alroy I, Towers TL, Freedman LP. Transcriptional repression of the interleukin-2 gene by vitamin D3: direct inhibition of NFATp/AP-1 complex formation by a nuclear hormone receptor. Mol Cell Biol. 1995;15:5789–5799.
Mangelsdorf DJ, Borgmeyer U, Heyman RA, Zhou JY, Ong ES, Oro AE, Kakizuka A, Evans RM. Characterization of three RXR genes that mediate the action of 9-cis retinoic acid. Genes Dev. 1992;6:329–344.