1,25 Dihydroxyvitamin D Amplifies Type a Natriuretic Peptide Receptor Expression and Activity in Target Cells

Articles

1,25 Dihydroxyvitamin D Amplifies Type A Natriuretic Peptide Receptor Expression and Activity in Target Cells

Songcang Chen, Xi-Ping Ni, Michael H. Humphreys, and David G. Gardner Diabetes Center and Department of Medicine, University of California at San Francisco, San Francisco, California

1,25 dihydroxyvitamin D (VD) has been shown to exert a number of beneficial effects on cardiovascular function, including reduction in BP and inhibition of cardiac hypertrophy. In an effort to identify a possible mechanistic link between VD and these salutary effects, the role of VD in controlling the activity and expression of the type A natriuretic peptide receptor (NPR-A), a receptor that signals reductions in BP and suppression of cellular growth in the myocardium and vascular wall, was investigated. VD, as well as the nonhypercalcemic analogue RO-25-6760, increased NPR-A–dependent cyclic guanosine monophosphate production and NPR-A gene expression in cultured rat aortic smooth muscle cells. The increase in NPR-A expression was associated with an increase in NPR-A gene promoter activity that was critically dependent on the presence of a functional VD receptor response element located approximately 495 bp upstream from the transcription start site of the gene. This element was associated with the VD receptor/retinoid X receptor complex in vitro. Mutation of this element resulted in complete elimination of the VD-dependent induction of the NPR-A gene promoter but did not affect osmotic stimulation of the promoter. Treatment of rats with RO-25-6760 for 7 d increased the atrial natriuretic peptide–dependent excretion of sodium and cyclic guanosine monophosphate without affecting mean arterial BP or plasma calcium levels. This was associated with a twofold increase in NPR-A mRNA levels in the inner medulla. Amplification of NPR-A activity represents a plausible mechanism to account for at least some of the beneficial effects that VD exerts on cardiovascular function. J Am Soc Nephrol 16: 329–339, 2005. doi: 10.1681/ASN.2004090797

he natriuretic peptides constitute a family of closely receptor. Blockade of NPR-C has been shown to increase levels related vasoactive hormones that play an important of endogenous natriuretic peptides in circulating plasma (7). T role in the regulation of cardiovascular, renal, and en- Vitamin D is a seco-steroid that bears a number of important docrine homeostasis (1). Atrial natriuretic peptide (ANP) and similarities to hormonal ligands that act through nuclear mech- brain natriuretic peptide (BNP) are produced predominantly in anisms to control gene expression. Its most polar metabolite, the heart, whereas C-type natriuretic peptide (CNP) is pro- 1,25 dihydroxyvitamin D (VD), binds specifically and with high duced in endothelial cells of the vasculature (2), reproductive affinity to a member of the nuclear receptor gene family termed tissues (3), growth plate of long bones (4), and the central the vitamin D receptor (VDR). The liganded VDR, when com- nervous system (5). The natriuretic peptides exert their biologic plexed with its heterodimeric partner, the retinoid X receptor effects through association with specific membrane-bound re- (RXR), associates with specific regulatory elements on DNA ceptors. Natriuretic peptide receptor types A and B (NPR-A called vitamin D response elements (VDRE). One such VDRE, and NPR-B, respectively) each have a large extracellular ligand- DR3, is a direct repeat of the sequence AGGTCA separated by binding domain connected to the intracellular signaling do- a three-nucleotide spacer (AGGTCANNNAGGTCA). DR3 dis- main through a single membrane-spanning segment. The intra- plays selectivity in preferentially binding to the VDR-RXR het- cellular domain harbors a noncatalytic, kinase-like region erodimer versus other nuclear receptor dimeric complexes (8). linked to a particulate guanylyl cyclase at the carboxy terminus A number of studies have identified VD as a potentially of the molecule (1). NPR-A serves as the cognate receptor for important regulator of BP and cardiovascular homeostasis. The both ANP and BNP, whereas NPR-B binds selectively to CNP. mechanism underlying the inverse relationship between VD Most of the important physiologic activities of these peptides levels and BP is not completely understood; however, the re- seem to be mediated through activation of these guanylyl cy- cent studies of Li et al. (9) suggest that the renin-angiotensin clase–linked receptors. A third receptor subtype, NPR-C, is system may be involved. Using genetically engineered VDR thought to function largely, albeit not entirely (6), as a clearance knockout mice, they showed that these animals were hyperten- sive relative to their VDR ϩ/ϩ littermates and that this in-

Received September 27, 2004. Accepted October 20, 2004. crease in BP was accompanied by apparent activation of the plasma renin-angiotensin system (RAS). Published online ahead of print. Publication date available at www.jasn.org. The data are equally supportive for an important regulatory Address correspondence to: Dr. David G. Gardner, 1109 HSW, Diabetes Center, role for VD in the heart. Rats made vitamin D–deficient display University of California at San Francisco, 3rd and Parnassus Avenue, San Francisco, CA 94143-0540. Phone: 415-476-2729; Fax: 415-564-5813; E-mail: elevations not only in BP but also in ventricular hypertrophy [email protected] (10). The former seems to be related, at least in part, to the

Copyright © 2005 by the American Society of Nephrology ISSN: 1046-6673/1602-0329 330 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 329–339, 2005 attendant hypocalcemia because calcium supplementation cor- of the sample and standard using a commercial antibody and rects the hypertension; however, the cardiac hypertrophy per- [125I]cGMP as tracer. sists. Subsequent studies from the same group indicated that a major component of the hypertrophy seen in vitamin D–defi- RNA Isolation and Northern Blot Analysis cient animals reflects expansion of the interstitial compartment RASM cells were plated in 10-cm dishes, cultured, and treated with in the heart (predominantly cardiac fibroblasts) with increased VD or the nonhypercalcemic vitamin D analogue RO-25-6760 (20,21) at production of extracellular matrix proteins (11). VD has been different concentrations, as indicated in the individual figure legends. shown to reduce endothelin-stimulated ANP and BNP gene Total RNA was extracted from cells using the RNeasy minikit accord- expression (markers of myocyte hypertrophy) and transcrip- ing to instructions provided by the manufacturer. Total RNA was tion in cultured neonatal rat atrial (12) and ventricular (13) denatured and separated on a gel that contained 2.2% formaldehyde, transferred to nitrocellulose filters, and hybridized to radiolabeled myocytes and to suppress the hypertrophic response to endo- cDNA as described previously (22). A 1.2-kb EcoRI fragment of the rat thelin, a well known hypertrophic agonist in the cultured ven- NPR-A cDNA was isolated from vector sequence, radiolabeled using tricular myocyte model (13). It is interesting that this suppres- the Primer-it RMT kit (Stratagene), and separated from free nucleotide sive activity seems to require structural features of the VDR using NucTrap push columns (Stratagene). The membranes were pre- that are more typically associated with the activation function hybridized for 30 min at 68°C and hybridized with 32P-labeled NPR-A of this receptor (14,15). cDNA (107 cpm/2 ml of hybridization solution) for1hat68°C in Like the liganded VDR, activated NPR-A has been shown to hybridization solution provided by Stratagene. All membranes were be both antihypertrophic and antihypertensive (16,17) in a subsequently stripped and rehybridized with a radiolabeled 1150-bp number of different model systems. In the present study, we BamHI/EcoRI fragment of 18S rDNA to permit normalization among explored the possibility that VD might exert beneficial effects samples for differences in RNA loading and/or transfer to the filter. on the cardiovascular system through NPR-A. Our findings Hybridization signal was detected by autoradiography and quantified using the NIH Image program. support this hypothesis in demonstrating that VD increases both the activity of the NPR-A protein and the expression of the NPR-A gene in cultured rat aortic smooth muscle (RASM) cells. Plasmid Constructions Ϫ1575 rat NPR-A LUC was generated by cloning 1575 bp of NPR-A 5Ј flanking sequence into BamH1/NarI sites of pFOXLUC vector as Materials and Methods described previously (23). Six of the deletion mutants were generated Materials through the use of convenient restriction sites. In each case, the 3Ј ␣ 32 [ - P]dCTP and the cyclic guanosine monophosphate (cGMP) RIA linkage used the NarI site of the pFOXLUC vector. The 5Ј linkage used kit were purchased from Perkin-Elmer Life Sciences (Boston, MA). unique restriction sites (NdeI for Ϫ1290LUC, BclI for Ϫ716LUC, HindIII ANP was obtained from Phoenix Pharmaceuticals, Inc. (Mountain for Ϫ387LUC, SstII for Ϫ273LUC, and DraII for Ϫ77LUC) for placement View, CA). RNeasy minikit was from Qiagen Inc. (Santa Clara, CA). in the polylinker of the parent vector. The shortest construct, Ϫ9LUC Primer-it RMT kit, hybridization solution, and NucTrap push columns was generated using an upstream sense oligonucleotide (5Ј-GGGGAT were purchased from Stratagene (La Jolla, CA). RO-25-6760 was pro- CCTTCTGGCACACTCCTG-3Ј) that incorporated a BamHI site at its 5Ј vided by M. Uskokovic (Roche Pharmaceuticals). Other reagents were terminus and a downstream antisense oligonucleotide (5Ј-TAGAG- obtained through standard commercial suppliers. GATAGAATGGCGCCGG-3Ј) derived from 3Ј linker sequence of pFOXLUC containing the NarI site. A PCR product of appropriate size Culture of Vascular Smooth Muscle Cells was restricted with BamHI/NarI and cloned into compatible sites of pFOXLUC. All of the deletion constructs were sequenced to confirm the Neonatal RASM cells, originally isolated by P. Lynch at the Univer- predicted structure. sity of Southern California, were obtained at passage 20 from H. Ives at University of California San Francisco. Cells were cultured at 37°C in a

5% CO2-humidified incubator in DMEM supplemented with 10% FBS, Transfection and Luciferase Assay 100 units/ml penicillin, 100 ␮g/ml streptomycin, and 2% (vol/vol) RASM cells were transfected by electroporation at 250 V and 960 ␮F broth, tryptose phosphate (18). using 1 ␮gofϪ1575 NPR-A LUC with 0.2 ␮g of cytomegalovirus-␤- galactosidase. After transfection, cells were plated on six-well plates in growth medium for 24 h. Medium was then changed to DMEM/serum Measurement of ANP-Stimulated cGMP Levels substitute for the ensuing 24 h, at which point various concentrations of RASM cells were grown to approximately 80% confluence, then Ϫ VD, or 10 8 M VD at different time intervals, were added. Luciferase changed to serum-substitute media (19) for 24 h. At that point, cells Ϫ Ϫ activity was measured using the luciferase assay system (Promega). were treated with 10 10 to 10 7 M VD for 48 h. For measurement of ␤-Galactosidase activity was assayed using the Galactolight Plus ANP-stimulated cGMP accumulation, cells were washed three times chemiluminescence assay (Tropix, Bedford, MA). Luciferase levels with prewarmed PBS and incubated with 0.5 ml of DMEM that con- were normalized for ␤-galactosidase activity in the individual cultures. tained 0.5 mM isobutylmethylxanthine and 10 mM HEPES (pH 7.4) for Ϫ 10 min at 37°C. 10 7 M ANP was added to the medium, and the incubation was continued for another 10 min. The reaction was stopped Preparation of Nuclear Extracts by the removal of medium and addition of 0.3 ml of 12% TCA. The RASM cells were transfected with 5 ␮g of hVDR and hRXR or sham extraction was continued for 30 min at 4°C. The contents of the plate transfected using Lipofectin Reagent (InVitrogen) for 24 h. Cells were Ϫ were collected and centrifuged to pellet particulate material. The su- changed to serum substitute media and treated with vehicle or 10 8 M pernatant fraction was extracted four times with 0.5 ml of water- VD for 48 h. Nuclear extracts were prepared as described previously saturated ether. cGMP levels were determined by RIA after acetylation (24). J Am Soc Nephrol 16: 329–339, 2005 Vitamin D Stimulates NPR-A Gene 331

Electrophoretic Mobility Shift Assay from the upper respiratory tract. The left jugular vein was exposed, and 35S-labeled hVDR and unlabeled hVDR and hRXR␣ were translated a catheter (PE50) was placed. A solution of normal saline that contained in vitro using a cell-free translation kit (TNT T7 Quick Master Mix kit; 1 mg/ml BSA was infused continuously at 25 ␮l/min through this Promega) with [35S]methionine and unlabeled methionine, respec- catheter. The carotid artery was identified through a small paratracheal tively. Double-stranded DR3 (5Ј-AGCTTCAGGTCAAGGAGGTCA- incision, and a tapered polyethylene catheter (PE50) was inserted for GAG-3Ј; consensus VDRE half sites indicated in bold letters) and wild- measurement of arterial pressure with a P23id Statham BP transducer type or mutant oligonucleotides derived from NPR-A promoter attached to a direct writing recorder. A small catheter (PE10) was sequence spanning the region between Ϫ504 and Ϫ472 were generated inserted into the right ureter near the hilum through a right flank for use in the electrophoretic mobility shift assay (EMSA). Two differ- incision for collection of urine from the right kidney. A catheter (PE50 ent protocols were used for EMSA. tubing) was sutured into the dome of the urinary bladder via a supra- Protocol 1 pubic incision for the timed collection of urine from the left kidney. A Two microliters each of unlabeled recombinant VDR and hRXR␣ or curved 30-G needle attached to PE10 tubing was inserted into the left 3 ␮l of RASM cell nuclear extract (see above) were incubated in binding renal artery near its takeoff from the aorta through a left flank incision reaction buffer (10 mM HEPES [pH 7.9], 50 mM KCl, 0.2 mM EDTA, 2.5 under the rib cage. Prompt reflux of arterial blood into the catheter was mM DTT, 10% glycerol, and 0.05% NP-40) that contained 0.5 ␮gof used to indicate successful needle placement. After completion of this poly(dI-dC) and 32P-end-labeled, double-stranded oligonucleotide, and preparative surgery, the rats received an infusion of normal saline that 10 nM VD at room temperature for 30 min. For competition experi- contained 5% BSA (total volume equal to 0.7% of total body weight) ments, a 10- or 100-fold molar excess of unlabeled, double-stranded over 15 min to replace surgical fluid losses. BP was measured over the oligonucleotide was added to the binding reaction. All samples were subsequent 30 min, and an average value of the electrical mean was resolved on 5% nondenaturing polyacrylamide gels. Gels were dried taken to record mean arterial pressure. At that point, ANP (250 fmol/L and exposed to x-ray film. per min) was infused into the left renal artery for 20 min. Urine was Protocol 2 collected from both kidneys in the 20-min period before and during the Two microliters of 35S-labeled VDR (synthesized in vitro) was incu- 20-min period of intrarenal ANP infusion. At the end of study, the rats bated, in the presence of 2 ␮l of unlabeled hRXR␣ and 0.5 ␮g poly(dI- were killed, blood was collected and processed for measurement of dC), with 20 ng of unlabeled, double-stranded DR3 or NPR-A oligo- plasma calcium, and the renal inner medulla from both kidneys was nucleotides and 10 nM VD. Reactions were carried out in binding isolated for measurement of NPR-A mRNA. Urine volume was determined, and urinary sodium was measured by flame photometry buffer (10 mM NaHPO4 [pH 7.6], 0.25 mM EDTA, 0.5 mM MgCl2, and 5% glycerol) for 20 min at 23°C. DNA–protein complexes were resolved (Model 943; Instrumentation Laboratories, Lexington, MA). Urine sam- ␮ by electrophoresis on 5% nondenaturing polyacrylamide gels run in ples were then diluted 50-fold before measurement of cGMP (2- l TEA buffer (67 mM Tris-HCl [pH 7.5], 10 mM EDTA, and 33 mM sample) using the method described above. All procedures were ap- sodium acetate) at 240 V for3hat4°C. The gel was washed three times proved by the IAUCC at UCSF. with 30% methanol and 10% glacial acetic acid, amplified for 30 min (Amplifier, Amersham Pharmacia Biotech), dried, and subjected to Real-Time PCR Analysis autoradiography. Total RNA was prepared from frozen inner medullas of control and RO-25-6760–treated rats using the RNaeasy kit. Two micrograms of Site-Directed Mutagenesis total RNA was used to reverse transcribe cDNA using the Clontech Mutations (M1, M2) were created in Ϫ1575 NPR-A-Luc by site- reverse transcription kit. Rat NPR-A and glyceraldehyde phosphate directed mutagenesis using a commercial kit (Stratagene). Briefly, mix- dehydrogenase (GAPDH) (25) primer pairs and probes were synthe- tures that contained 50 ng of Ϫ1575 NPR-A-Luc, two mutagenic prim- sized by Applied Biosystems. Real-time PCR was performed using ers, dNTP, and Pfu DNA polymerase were added to the PCR buffer. Taqman Master mix (A&B Applied Biosystems) with an ABI Prism 7700 PCR was carried out for 18 cycles using 30 s of denaturation at 95°C, (A&B Applied Biosystems). Negative controls without input cDNA 1 min of annealing at 55°C, and 2 min/kb extension at 68°C. After PCR, were used to assess signal specificity. NPR-A transcript levels were 1 ␮lofDpn I was added to the reaction to cut parental DNA template, quantified and normalized for GAPDH transcript levels in each sample. and 5 ␮l of this digest was used for transformation. Several candidate clones were identified and characterized by DNA sequencing. Statistical Analyses Data were analyzed by ANOVA using Bonferroni test to assess Analysis of Renal ANP Activity In Vivo significance. Under anesthesia (intraperitoneal injection of Brevital sodium 50 mg/kg), a midscapular incision was made and Alzet osmotic Results minipump was implanted subcutaneously into adult male Sprague- To determine whether VD has a role in the regulation of Dawley rats. In the treatment group, the pumps were loaded with NPR-A expression/activity, we examined the functional re- ␮ ϩ RO-25-6760 (0.8 g/kg per d dissolved in 98% propylene glycol 2% sponse of the receptor to ligand (ANP-dependent cGMP gen- ethanol). In the control group, pumps were filled out with vehicle eration) in the presence or absence of VD. Forty-eight hours of alone. The wound was closed with sutures, and the rats were returned VD treatment had no effect on basal cGMP levels in RASM cells to their cages after recovery from anesthesia. Over the ensuing 7 d, they (Chen et al., data not shown). ANP treatment of control cells led were observed for any signs of calcium toxicity: Tremors, muscle spasms, seizures, anorexia, or weight loss. One week after pump im- to an approximately fivefold increment in cGMP levels, relative plantation, each rat received an intraperitoneal injection of Inactin (100 to basal, over a 10-min incubation period. VD pretreatment mg/kg) to induce anesthesia. Animals were placed on a heated oper- nearly doubled this response with a net stimulation of cGMP ating table. A tracheotomy was performed, and a segment of PE205 levels of approximately 10-fold after ANP treatment (Figure tubing was inserted into the trachea to facilitate clearance of secretions 1A). A similar amplification of the response to ANP was seen 332 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 329–339, 2005

after pretreatment with the nonhypercalcemic VD analogue RO-25-6760. This increment in NPR-A activity resulted, at least in part, from an increase in NPR-A gene expression. As shown in Figure 1B, VD pretreatment led to an approximately five- to sixfold increase in steady-state NPR-A mRNA levels. Of note, the nonhypercalcemic analogue RO-25-6760 led to a similar increment (approximately fivefold) in NPR-A mRNA levels in cultured RASM cells (Figure 1C). The increment in gene expression, in turn, seems to reflect an increase in NPR-A gene transcription. As shown in Figure 2, VD, as well as RO-25-6760, treatment of cultured RASM cells led to a significant increase in the activity of a transfected NPR-A promoter-luciferase reporter construct that was not seen with the transfected background vector (pGL3). This stimulation was ligand dependent (Figure 2A), VD receptor dependent (Figure 2B), and time-dependent (Figure 2B) and amplifi- able with co-transfected RXR, the heterodimeric partner of VDR (Figure 2C). A series of 5Ј-deletion mutants of the NPR-A promoter, extending from Ϫ1575 to Ϫ77 relative to the transcription start site, were generated and linked to the luciferase reporter. As shown in Figure 3, mutants that harbored as little as 706 bp of the 5Ј flanking sequence retained the ability to respond to VD; however, the Ϫ387 NPR-A-luciferase mutant displayed no induction with VD treatment. This suggests that the VD-sensitive regulatory element is positioned between Ϫ706 and Ϫ387 in the NPR-A gene promoter. Examination of the DNA sequence spanning this region for the presence of conventional VD regulatory elements revealed five potential candidates, each displaying moderate to high homology to the DR3 motif (two AGGTCA motifs arrayed as direct repeats and separated by three spacer nucleotides [8]; Figure 4). Double-stranded oligonucleotides spanning each of these regions were incubated with radiolabeled VDR, unla- Ϫ beled RXR, and the VD ligand (10 8 M) and fractionated on nondenaturing polyacrylamide gels to assess potential DNA– Figure 1. Vitamin D (VD) effects on natriuretic peptide receptor A protein interactions (i.e., conventional EMSA assay). As shown (NPR-A) activity and expression. (A) Rat aortic smooth muscle in Figure 5A, NPR-A-1, which displayed the highest degree of (RASM) cells were cultured in serum substitute medium for 24 h Ϫ Ϫ homology to the consensus VDRE (homologous at 9 of 12 and then treated with VD (10 10 to 10 7 M) or RO-25-6760 Ϫ positions), failed to interact with the VDR-RXR complex in (10 8 M) for 48 h. Cells were washed extensively and incubated Ϫ vitro. Similarly, NPR-A-2 and NPR-A-3 failed to associate with with 10 7 M atrial natriuretic peptide (ANP) in the presence of isobutylmethylxanthine (0.5 M) for 10 min. ANP-stimulated cyclic this receptor complex (Figure 5B); however, NPR-A-4 dis- guanosine monophosphate (cGMP) was determined as described played a level of VDR-RXR binding that was readily discern- in the Materials and Methods section. Pooled data from three ible, albeit at a fraction (approximately 50%) of the intensity independent experiments are shown. **P Ͻ 0.01 versus control in seen with the consensus DR3 motif. Binding to radiolabeled the presence of ANP. Control cGMP levels were 198 Ϯ NPR-A-4 was almost completely inhibited by a 100-fold molar 21 pmol/mg soluble protein. (B) RASM cells were treated with excess of unlabeled oligonucleotide (Figure 5C). Introduction of various concentrations of VD for 48 h. Total RNA was isolated, mutations into the downstream candidate site (NPR-A-4 M1) and Northern blot analysis was performed as described in the did not impair the ability of this oligonucleotide to inhibit materials and Methods section. Representative autoradiographs VDR-RXR binding to NPR-A-4 (Figure 6A) (i.e., mutated se- are shown. The bar graph displays findings (NPR-A mRNA/18S quence does not harbor the VDRE). Mutation of the upstream rRNA) from three separate experiments. **P Ͻ 0.01 versus control. site (NPR-A-4 M2), however, resulted in near complete abro- (C) RASM cells were exposed to various concentrations of RO-25- 6760 for 48 h. Total RNA was prepared, and Northern hybridiza- gation of the competition for NPR-A-4 binding to the het- tion was carried out. NPR-A mRNA levels were normalized to erodimeric receptor complex. The latter finding suggests that 18S rRNA. VD and RO-25-6760 concentrations in all three panels the liganded VDR-RXR complex binds to the upstream se- presented as log [M]. Pooled data from three independent exper- quence that is targeted by the NPR-A-4 M2 mutation. This was iments are shown. **P Ͻ 0.01 versus control. confirmed in direct binding studies as shown in Figure 6B. J Am Soc Nephrol 16: 329–339, 2005 Vitamin D Stimulates NPR-A Gene 333

Figure 2. VD upregulates NPR-A promoter activity in RASM cells. (A) RASM cells were transfected with Ϫ1595 NPR-A LUC or pFOX LUC, and cytomegalovirus-␤-galactosidase (CMV-␤-gal) and cultured for 24 h. Medium was then changed to DMEM/ serum substitute for the ensuing 24 h, at which point VD or RO-25-6760, at the concentrations indicated (all concentrations presented as log [M]), was added for the next 48 h. (B) Ϫ1595 NPR-A LUC and CMV-␤-gal were co-transfected with increasing Ϫ concentrations of VD receptor (VDR) expression vector into RASM cells. After transfection, cells were treated with 10 8 MVDfor indicated time intervals. (C) A total of 0.5 ␮g of VDR and retinoid X receptor (RXR) was co-transfected with Ϫ1595 NPR-A Ϫ LUC/CMV-␤-gal into RASM cells. The transfected cells were incubated with vehicle or 10 8 M VD for 48 h. In all cases, lysates were generated and luciferase and ␤-gal activities were measured as described in the Materials and Methods section. Experiments were repeated three to six times. **P Ͻ 0.01, *P Ͻ 0.05 versus corresponding control. 334 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 329–339, 2005

Figure 3. Deletion analysis of VD induction of NPR-A promoter activity. Different deletions of NPR-A-LUC were made as described in the Materials and Methods section and transfected into RASM cells. Twenty-four hours later, cells were incubated Ϫ in presence or absence of 10 8 M VD for 48 h before generating lysates for luciferase and ␤-gal measurements. The experiments Ͻ were repeated three to four times. **P 0.01 versus correspond- Figure 5. Specific VDR/RXR binding activity resides in NPR- ing control. A-4 sequence. (A and B) 35S-VDR and unlabeled RXR were incubated with unlabeled oligonucleotides encoding a consensus DR3 motif or one of four putative VDRE in the NPR-A Whereas wild-type NPR-A-4 and the M1 mutant associated promoter. All incubations were carried out in the presence of Ϫ with radiolabeled VDR in this assay, the M2 mutant was devoid VD (10 8 M). DNA-protein complexes were fractionated by of binding activity. electrophoretic mobility shift assays (EMSA) as described in the Binding of VDR to NPR-A-4 did not require ligand; binding Materials and Methods section. (C) Unlabeled VDR and RXR, 32 of recombinant VDR/RXR to NPR-A-4, if anything, was re- along with VD, were incubated with P-labeled NPR-A-4 oli- duced slightly in the presence of 1,25 dihydroxyvitamin D gonucleotide in the absence or presence of increasing concentrations of unlabeled oligonucleotide (10- or 100-fold molar (Figure 7A). It is interesting that this differed from receptor excess). The reaction complexes were resolved on 5% nonde- expressed in a transfected cell. As shown in Figure 7B, extracts naturing polyacrylamide gels. The experiments were repeated of cells that were transfected with hVDR plus hRXR showed a two to three times. Representative autoradiographs are shown.

modest increase in binding to NPR-A-4 versus untransfected cells; however, the inclusion of ligand resulted in a dramatic increase in binding of the extract to the oligonucleotide, suggesting that other proteins in the extract may serve to stabilize the DNA-VDR/RXR complex. Extending this analysis to the functional assay, mutation of the M1 site in the context of Ϫ1575 NPR-A luciferase failed to influence the VD-dependent stimulation of this promoter, relative to the wild-type sequence, whereas introduction of the M2 mutation into the same promoter context resulted in complete loss of the VD-dependent stimulation (Figure 8A). Noteworthy, the M2 mutation did not prevent induction of the NPR-A promoter by osmotic stimuli such as NaCl or sucrose (22,26) (Figure 8B), indicating that the targeted sequence selectively transduces the response to VD. To assess the physiologic relevance of the in vitro findings, Figure 4. Identification of candidate VD response elements we attempted to extend these analyses into an in vivo model. (VDRE) in NPR-A promoter. Consensus VDRE half sites in DR3 We chose a model that we have used previously to examine the motif are identified at the top. Candidate VDRE sequences in NPR-A promoter are identified in bold type. Termini identify effects of ANP infused directly into the renal artery on urinary positions of the DNA fragments relative to the transcription volume and sodium excretion (27). This model has the advan- start site; orientation is indicated by arrows positioned above. tage of identifying ANP-induced changes in urinary sodium Mutated bases in VDRE of NPR-A4 (M1 and M2) are indicated and water handling independent of changes in peripheral he- in lowercase letters. modynamics. As shown in Table 1, pretreatment of animals J Am Soc Nephrol 16: 329–339, 2005 Vitamin D Stimulates NPR-A Gene 335

Figure 7. VD promotes binding of VDR to VDRE in intact cells but not in cell-free extracts. (A) 35S-VDR and unlabeled RXR were incubated with unlabeled NPR-A-4 or consensus VDRE Ϫ oligonucleotide in the presence or absence of 10 8 MVDand separated by EMSA. (B) Six micrograms of nuclear extract from sham- versus VDR/RXR-transfected RASM cells in the presence or absence of VD were incubated with 32P-labeled NPR-A-4 oligonucleotide and separated by EMSA. Each experiment was repeated three times. Representative autoradiographs are shown.

with the nonhypercalcemic analogue RO-25-6760 for 7 d had no effect on basal mean arterial pressure relative to control. Serum calcium levels were measured in both the control and the analogue-infused animals. Calcium levels in the RO-25-6760– treated animals remained in the normal range and were not statistically different from the controls (Table 1). Infusion of ANP (250 fmol/kg per min) into the left renal artery had no effect on systemic arterial pressure (Table 1). It did, however, promote an increase in urinary sodium excretion that was not seen in the contralateral, noninfused kidney (Figure 9A). Pre- treatment with RO-25-6760 for 7 d resulted in a significant increase in ANP-stimulated urinary sodium excretion in the infused kidney, and this was accompanied by a significant increase in ANP-dependent urinary cGMP levels when com- pared with the control group (Figure 9B). In both instances, the

magnitude of the increases (in UNa and UcGMP) was larger in the RO-25-6760–treated versus vehicle-treated rats. A modest

elevation in UNaV and UcGMP was seen in urine that was collected from the contralateral kidney, presumably reflecting the effects of peptide that has either found its way into the Figure 6. Selective mutation of the VDRE in NPR-A-4. (A) systemic circulation or refluxed into the aorta from the infusion Ϫ Unlabeled VDR and RXR, along with VD (10 8 M), were incu- site. Collectively, these findings support our in vitro data in 32 bated with P-labeled NPR-A-4 oligonucleotide in the absence demonstrating that the VD analogue amplifies the activity of or presence of increasing concentrations (10- or 100-fold excess) the natriuretic peptides in target cells. of one of two mutant NPR-A oligonucleotides (see Figure 4 for To determine whether the increase in natriuretic peptide identification). A 100-fold molar excess NPR-A1 oligonucleo- activity resulted from increased NPR-A expression, as docu- tide was included as negative control. (B) 35S-VDR, unlabeled mented in the in vitro system above, we measured NPR-A RXR, and VD were incubated with the consensus DR3 element, wild-type NPR-A-1, wild-type NPR-A-4, or one of the two mRNA transcripts in inner medulla of control versus RO-25- mutants of NPR-A-4 identified in Figure 4. EMSA was per- 6760–treated rats using real-time PCR. As shown in Figure 9C, formed as described in the Materials and Methods section. treatment with RO-25-6760 for 7 d resulted in an approximately Experiments were repeated three times; representative results twofold increment in NPR-A mRNA levels in the inner me- are shown. dulla. Levels of the GAPDH transcript were essentially un- 336 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 329–339, 2005

identify an important mechanism through which this hormone may exert salutary effects in the cardiovascular system. Previ- ous studies have suggested regulation of NPR-C expression by glucocorticoids (28); however, transcriptional regulation was not documented in those studies, and a specific glucocorticoid responsive element was not identified. The current study dem- onstrates VD1-dependent stimulation of NPR-A activity as well as NPR-A gene expression. The latter derives from an increase in NPR-A promoter activity that is driven by a solitary VDRE located approximately 495 bp upstream from the transcription start site. Vitamin D activity is classically described in terms of its calcitropic properties in the intestine, kidney, and bone (29); however, more recently increased attention has been devoted to its ability to regulate physiologic events through pathways independent of those that it uses to control serum calcium and phosphorous concentrations. VD has immune-suppressant and anti-inflammatory properties (30), it reduces cell proliferation in a variety of in vitro models through mechanisms that are only partially understood (31,32), and it seems to play an important role in the growth and development of the epidermis and hair follicle (33,34). Data acquired through a number of different experimental and clinical studies indicate that VD also plays an important role in cardiovascular homeostasis. 25-Hydroxyvita- min D levels are inversely related to BP levels in selected forms of rodent (35) and human (36) hypertension, and administra- tion of VD restores BP to normal or near-normal levels (37,38). Of note, the recent study of Teng et al. (39) demonstrated improved survival of patients who had ESRD and were on hemodialysis when they were treated with the nonhypercalcemic analogue of VD, paricalcitol, versus VD itself, suggesting the existence of important salutary effects of the seco-steroid that are independent of its calcitropic properties. Because the majority of patients with ESRD die of cardiovascular causes, the inference would be that the benefit accrued secondary to the beneficial cardiovascular effects of these drugs. True vitamin D–deficiency in the rat is accompanied by acquired hypertension and associated ventricular hypertrophy Figure 8. Mutation of the VDRE in the NPR-A gene promoter (10), although in this model, the BP elevation abates with interferes with VD-induced NPR-A gene promoter activity. (A) correction of the attendant hypocalcemia. The recent studies of Wild-type Ϫ1595 NPR-A-LUC or one of two individual mu- Li et al. (9) suggest that hypertension results from defective VD tants of the NPR-A-4 sequence (M1 and M2) was co-transfected action in general, rather than solely from hypocalcemia. In VDR ␤ Ϫ with CMV- -gal into RASM cells. (B) 1595 NPR-A-LUC and null (Ϫ/Ϫ) mice, systolic BP was elevated approximately 20 M2 mutant along with CMV-␤-gal were co-transfected into Ϫ mmHg versus controls, and this was associated with activation RASM cells. After 24 h, the cells were treated with 10 8 MVD of the RAS. Because these animals were maintained on a high- for 48 h. Alternatively, 48 h after transfection, cells were treated calcium diet to prevent the development of hypocalcemia, it with 150 mM sucrose or 75 mM NaCl for 24 h. Luciferase and ␤-gal activity was measured. Experiment was repeated three seems likely that the BP elevation is at least partially reflective times. **P Ͻ 0.01 versus corresponding control. of the defect in VDR signaling. Li et al. also presented data suggesting a direct effect of the liganded VDR on the renin gene promoter, although a discrete functional VDRE was not iden- changed. This suggests that the increment in activity of the tified. liganded NPR-A noted above results from VD-dependent stim- NPR-A plays a major role in controlling BP and extracellular ulation of NPR-A gene expression. fluid volume (1). In fact, the studies of DuBois et al. (40) suggest that most, if not all, of the renal effects of ANP are mediated Discussion through this receptor. That it mediates vasodilation, increased The studies described above provide definitive evidence for urinary excretion of sodium and water, and suppression of transcriptional regulation of a cardiovascular gene by VD and myocardial hypertrophy and fibrosis (17) makes it an attractive J Am Soc Nephrol 16: 329–339, 2005 Vitamin D Stimulates NPR-A Gene 337

ϩ Table 1. Effect of RO-25-6760 on plasma Ca2 and mean arterial BP before and after infusion of ANP in vivoa

ϩ Group N Infusion of Vehicle Infusion of ANP Plasma Ca2

Control n ϭ 8 105 Ϯ 4.6 mmHg 106 Ϯ 5.2 mmHg 9.1 Ϯ 0.30 mg/dl RO-25-6760 n ϭ 8 102 Ϯ 4.4 mmHg 101 Ϯ 4.4 mmHg 9.7 Ϯ 0.59 mg/dl

aANP, atrial natriuretic peptide.

naling cascade (41), including inhibition of renin gene expression. Therefore, it is conceivable that at least a portion of the reduction in renin expression could result from enhanced NPR-A expression and activity rather than through a direct liganded VDR-dependent suppression of the renin gene promoter. In fact, Li et al. (42) reported that NPR-A gene expression is reduced in VDR Ϫ/Ϫ mice. Additional studies will be re- quired to assess whether the direct versus indirect model is operative here. Finally, our previous demonstration that VD suppresses expression and secretion of ANP (12,13,43) seems, at first glance, to be physiologically inconsistent with the data presented here, because a reduction in ligand levels would be predicted to offset the increase in receptor activity. However, it is important to point out that the VD-dependent inhibition of ANP expression is largely confined to hypertrophy-dependent expression (e.g., that induced by endothelin treatment) and reflects a global antagonism of the hypertrophic process rather than isolated inhibition of the ANP gene. Activation of the ANP gene in hypertrophy represents an attempt, oftentimes futile (44), to reverse the hemodynamic stimuli that elicit the increase in myocardial mass and through primary (16,17) as well as secondary (1) effects control further progression of the hypertrophic process. On the basis of its ability to increase NPR-A activity, VD would be predicted to support ANP effects in this setting. Thus, the VD effects may be viewed as consistently aligned toward suppression of growth and hypertrophy in the cardiovascular system. Figure 9. Effect of RO-25-6760 on ANP-stimulated UNaV, uri- In summary, we have shown that the vasodilatory, antimi- nary cGMP, and NPR-A mRNA levels in vivo. (A) Intrarenal togenic receptor NPR-A is under the transcriptional regulatory infusion of ANP (250 fmol/kg per min) for 20 min. Urine was control of 1,25 dihydroxyvitamin D. This effect may be respon- collected from infused and contralateral kidney of control and RO-25-6760–treated rats before and after ANP infusion. Urine sible for a major component of the salutary effects that vitamin volume and sodium were measured. (B) Urinary cGMP levels D exerts in the cardiovascular system. were measured in aliquots of urine as described in the Materi- Ϯ als and Methods section. Control cGMP levels were 533 53 Acknowledgments pmol/ml urine. (C) Real-time PCR analysis of NPR-A mRNA These studies were supported by RO1 HL45637 and DK58812 from levels in rat inner medulla. Paired inner medullas from control the National Institutes of Health. and RO-25-6760–treated rats were isolated and snap-frozen in We are grateful to M. Uskokovic of Roche Pharmaceuticals for pro- liquid nitrogen. Total RNA was prepared and reverse tran- viding RO-25-6760. scribed into cDNA. Real-time PCR was carried out as described in the Materials and Methods section. Pooled data (n ϭ 8) are ϩ shown. **P Ͻ 0.01, *P Ͻ 0.05 versus ANP-infused control; P Ͻ 0.01 versus corresponding control. References 1. Levin ER, Gardner DG, Samson WK: Natriuretic peptides. N Engl J Med 339: 321–328, 1998 2. Suga S, Itoh H, Komatsu Y, Ogawa Y, Hama N, Yoshimasa candidate as a target of and mediator for VD action. Although T, Nakao K: Cytokine-induced C-type natriuretic peptide this does not exclude a parallel role for VD in suppression of (CNP) secretion from vascular endothelial cells—Evidence the RAS, it is noteworthy that the liganded NPR-A is also for CNP as a novel autocrine/paracrine regulator from capable of suppressing the RAS at multiple points in the sig- endothelial cells. Endocrinology 133: 3038–3041, 1993 338 Journal of the American Society of Nephrology J Am Soc Nephrol 16: 329–339, 2005

3. Stepan H, Leitner E, Walter K, Bader M, Schultheiss H, mammary tumor cell lines and production of mouse mam- Faber R, Walther T: Gestational regulation of the gene mary tumor virus in a serum-free medium. In Vitro 12: expression of C-type natriuretic peptide in mouse repro- 558–563, 1976 ductive and embryonic tissue. Regul Pept 102: 9–13, 2001 20. Uskokovic MR, Norman AW, Manchand PS, Studzinski 4. Yasoda A, Ogawa Y, Suda M, Tamura N, Mori K, Sakuma GP, Campbell MJ, Koeffler HP, Takeuchi A, Siu-Caldera Y, Chusho H, Shiota K, Tanaka K, Nakao K: Natriuretic ML, Rao DS, Reddy GS: Highly active analogs of 1␣,25- peptide regulation of endochondral ossification. Evidence dihydroxyvitamin D(3) that resist metabolism through for possible roles of the C-type natriuretic peptide/guany- C-24 oxidation and C-3 epimerization pathways. Steroids lyl cyclase-B pathway. J Biol Chem 273: 11695–11700, 1998 66: 463–471, 2001 5. Barr CS, Rhodes P, Struthers AD: C-type natriuretic pep- 21. Pakkala S, de Vos S, Elstner E, Rude RK, Uskokovic M, tide. Peptides 17: 1243–1251, 1996 Binderup L, Koeffler HP: Vitamin D3 analogs: Effect on 6. Levin ER: Natriuretic peptide C-receptor: More than a leukemic clonal growth and differentiation, and on serum clearance receptor. Am J Physiol 264: E483–E489, 1993 calcium levels. Leuk Res 19: 65–72, 1995 7. Maack T, Suzuki M, Almeida FA, Nussenzveig D, Scarbor- 22. Chen S, Gardner DG: Osmoregulation of natriuretic pep- ough RM, McEnroe GA, Lewicki JA: Physiological role of tide receptor signaling in inner medullary collecting duct. silent receptors of atrial natriuretic factor. Science 238: 675– A requirement for p38 MAPK. J Biol Chem 277: 6037–6043, 678, 1987 2002 8. Umesono K, Murakami KK, Thompson CC, Evans RM: 23. Cao L, Wu J, Gardner DG: Atrial natriuretic peptide sup- Direct repeats as selective response elements for the thy- presses the transcription of its guanylyl cyclase-linked re- roid hormone, retinoic acid, and vitamin D3 receptors. Cell ceptor. J Biol Chem 270: 24891–24897, 1995 65: 1255–1266, 1991 24. Anderson HD, Rahmutula D, Gardner DG: Tumor necrosis 9. Li YC, Kong J, Wei M, Chen ZF Liu SQ, Cao LP: 1,25- factor-alpha inhibits endothelial nitric-oxide synthase gene Dihydroxyvitamin D(3) is a negative endocrine regulator promoter activity in bovine aortic endothelial cells. J Biol of the renin-angiotensin system. J Clin Invest 110: 229–238, Chem 279: 963–969, 2004 2002 25. Winer J, Jung CK, Shackel I, Williams PM: Development 10. Weishaar RE, Simpson RU: Vitamin D3 and cardiovascular and validation of real-time quantitative reverse transcrip- function in rats. J Clin Invest 79: 1706–1712, 1987 tase-polymerase chain reaction for monitoring gene ex- 11. Weishaar RE, Kim SN, Saunders DE, Simpson RU: Involve- pression in cardiac myocytes in vitro. Anal Biochem 270: ment of vitamin D3 with cardiovascular function. III. Ef- 41–49, 1999 fects on physical and morphological properties.AmJ 26. Chen S, Cao L, Intengan HD, Humphreys M, Gardner DG: Physiol 258: E134–E142, 1990 Osmoregulation of endothelial nitric-oxide synthase gene 12. Wu J, Garami M, Cao L, Li Q, Gardner DG: 1,25(OH)2D3 expression in inner medullary collecting duct cells. Role in suppresses expression and secretion of atrial natriuretic activation of the type A natriuretic peptide receptor. J Biol peptide from cardiac myocytes. Am J Physiol 268: E1108– Chem 277: 32498–32504, 2002 E1113, 1995 27. Chen XW, Ying WZ, Valentin JP, Ling KT, Lin SY, Wiede- 13. Wu J, Garami M, Cheng T, Gardner DG: 1,25(OH)2 vitamin mann E, Humphreys MH: Mechanism of the natriuretic D3, and retinoic acid antagonize endothelin-stimulated hy- action of gamma-melanocyte-stimulating hormone. Am J pertrophy of neonatal rat cardiac myocytes. J Clin Invest 97: Physiol 272: R1946–R1953, 1997 1577–1588, 1996 28. Herman JP, Dolgas CM, Marcinek R, Langub MC Jr: Ex- 14. Chen S, Costa CH, Nakamura K, Ribeiro RC, Gardner DG: pression and glucocorticoid regulation of natriuretic pep- Vitamin D-dependent suppression of human atrial natri- tide clearance receptor (NPR-C) mRNA in rat brain and uretic peptide gene promoter activity requires heterodimer choroid plexus. J Chem Neuroanat 11: 257–265, 1996 assembly. J Biol Chem 274: 11260–11266, 1999 29. Holick MF: Evolution and function of vitamin D. Recent 15. Chen S, Cui J, Nakamura K, Ribeiro RC, West BL, Gardner results. Cancer Res 164: 3–28, 2003 DG: Coactivator-vitamin D receptor interactions mediate 30. Hayes CE, Nashold FE, Spach KM, Pedersen LB: The im- inhibition of the atrial natriuretic peptide promoter. J Biol munological functions of the vitamin D endocrine system. Chem 275: 15039–15048, 2000 Cell Mol Biol (Noisy-le-grand) 49: 277–300, 2003 16. Holtwick R, van Eickels M, Skryabin BV, Baba HA, Bubikat 31. Godyn JJ, Xu H, Zhang F, Kolla S, Studzinski GP: A dual A, Begrow F, Schneider MD, Garbers DL, Kuhn M: Pres- block to cell cycle progression in HL60 cells exposed to sure-independent cardiac hypertrophy in mice with car- analogues of vitamin D3. Cell Prolif 27: 37–46, 1994 diomyocyte-restricted inactivation of the atrial natriuretic 32. Mitsuhashi T, Morris RC Jr, Ives HE: 1,25-Dihydroxyvita- peptide receptor guanylyl cyclase-A. J Clin Invest 111: min D3 modulates growth of vascular smooth muscle cells. 1399–1407, 2003 J Clin Invest 87: 1889–1895, 1991 17. Knowles JW, Esposito G, Mao L, Hagaman JR, Fox JE, 33. Kira M, Kobayashi T, Yoshikawa K: Vitamin D and the Smithies O, Rockman HA, Maeda N: Pressure-indepen- skin. J Dermatol 30: 429–437, 2003 dent enhancement of cardiac hypertrophy in natriuretic 34. Vegesna V, O’Kelly J, Uskokovic M, Said J, Lemp N, Saitoh peptide receptor A-deficient mice. J Clin Invest 107: 975– T, Ikezoe T, Binderup L, Koeffler HP: Vitamin D3 analogs 984, 2001 stimulate hair growth in nude mice. Endocrinology 143: 18. Wilson E, Mai Q, Sudhir K, Weiss RH, Ives HE: Mechanical 4389–4396, 2002 strain induces growth of vascular smooth muscle cells via 35. Thierry-Palmer M, Carlyle KS, Williams MD, Tewolde T, autocrine action of PDGF. J Cell Biol 123: 741–747, 199 Caines-McKenzie S, Bayorh MA, Emmett NL, Harris- 19. Bauer RF, Arthur LO, Fine DL: Propagation of mouse Hooker SA, Sanford GL, Williams EF: Plasma 25-hy- J Am Soc Nephrol 16: 329–339, 2005 Vitamin D Stimulates NPR-A Gene 339

droxyvitamin D concentrations are inversely associated with paricalcitol or calcitriol therapy. N Engl J Med 349: with blood pressure of Dahl salt-sensitive rats. J Steroid 446–456, 2003 Biochem Mol Biol 66: 255–261, 1998 40. Dubois SK, Kishimoto I, Lillis TO, Garbers DL: A genetic 36. Lind L, Hanni A, Lithell H, Hvarfner A, Sorensen OH, model defines the importance of the atrial natriuretic Ljunghall S: Vitamin D is related to blood pressure and peptide receptor (guanylyl cyclase-A) in the regulation other cardiovascular risk factors in middle-aged men. Am J of kidney function. Proc Natl Acad SciUSA97: 4369– Hypertens 8: 894–901, 1995 4373, 2000 37. Lind L, Wengle B, Wide L, Ljunghall S: Reduction of blood 41. Franco-Saenz R, Atarashi K, Takagi M, Mulrow PJ: Effect pressure during long-term treatment with active vitamin D of atrial natriuretic factor on renin and aldosterone. J Car- (alphacalcidol) is dependent on plasma renin activity and diovasc Pharmacol 13[Suppl 6]: S31–S35, 1989 calcium status. A double-blind, placebo-controlled study. 42. Li X, Zheng W, Li YC: Altered gene expression profile in Am J Hypertens 2: 20–25, 1989 the kidney of vitamin D receptor knockout mice. J Cell 38. Pfeifer M, Begerow B, Minne HW, Nachtigall D, Hansen C: Biochem 89: 709–719, 2003 Effects of a short-term vitamin D(3) and calcium supple- 43. Li Q, Gardner DG: Negative regulation of the human atrial mentation on blood pressure and parathyroid hormone natriuretic peptide gene by 1,25-dihydroxyvitamin D3. levels in elderly women. J Clin Endocrinol Metab 86: 1633– J Biol Chem 269: 4934–4939, 1994 1637, 2001 44. Charloux A, Piquard F, Doutreleau S, Brandenberger G, 39. Teng M, Wolf M, Lowrie E, Ofsthun N, Lazarus JM, Geny B: Mechanisms of renal hyporesponsiveness to ANP Thadhani R: Survival of patients undergoing hemodialysis in heart failure. Eur J Clin Invest 33: 769–778, 2003