All-trans retinoic acid reduces the transcriptional regulation of intestinal sodium-dependent phosphate co-transporter gene (Npt2b)

Abstract

Inorganic phosphate (Pi) homeostasis is regulated by intestinal absorption via type II sodium-dependent co-transporter (Npt2b) and by renal reabsorption via Npt2a and Npt2c. Although we previously reported that vitamin A-deficient (VAD) rats had increased urine Pi excretion through the decreased renal expression of Npt2a and Npt2c, the effect of vitamin A on the intestinal Npt2b expression remains unclear. In this study, we investigated the effects of treatment with all-trans retinoic acid (ATRA), a metabolite of vitamin A, on the Pi absorption and the Npt2b expression in the intestine of VAD rats, as well as and the underlying molecular mechanisms. In VAD rats, the intestinal Pi uptake activity and the expression of Npt2b were increased, but were reduced by the administration of ATRA. The transcriptional activity of reporter plasmid containing the promoter region of the rat Npt2b gene was reduced by ATRA in NIH3T3 cells overexpressing retinoic acid receptor (RAR) and retinoid X receptor (RXR). On the other hand, CCAAT/enhancer-binding proteins (C/EBP) induced transcriptional activity of the Npt2b gene. Knockdown of the C/EBP gene and a mutation analysis of the C/EBP responsible element in the Npt2b gene promoter indicated that C/EBP plays a pivotal role in the regulation of Npt2b gene transcriptional activity by ATRA. EMSA revealed that the RAR/RXR complex inhibits binding of C/EBP to Npt2b gene promoter. Together, these results suggest that ATRA may reduce the intestinal Pi uptake by preventing C/EBP activation of the intestinal Npt2b gene.

Introduction

Inorganic phosphate (Pi) homeostasis in mammals is strictly controlled through the balance of intestinal absorption and renal excretion/reabsorption [1,2]. The uptake of Pi is mediated by type II sodium-dependent phosphate co-transporters (Npt2) in the brush-border membrane of the small intestine and renal proximal tubule [1,2]. Npt2a and Npt2c are responsible for most Pi reabsorption in the kidney and can be regulated by dietary Pi, 1α,25-dihydroxyvitamin D₃ [1,25(OH)₂D₃], all-trans retinoic acid (ATRA), and hormonal factors such as parathyroid hormone (PTH), fibroblast growth factor 23 (FGF23), and tumor necrosis factor-α (TNFα) [1–6]. Npt2b plays a critical role in intestinal Pi absorption and its expression is regulated by dietary Pi, 1,25(OH)₂D₃, FGF23, epidermal growth factor (EGF), TNFα, and nuclear factor 1 (NF1) [1,7–14].

Disturbance of Pi homeostasis can cause important clinical disorders. Hyperphosphatemia, which is associated with the pathophysiology of chronic kidney disease (CKD), can lead to vascular calcification, which has been linked to increased cardiovascular morbidity and mortality [15]. The regulation of the intestinal Pi absorption mediated by Npt2b is a possible therapeutic target to control hyperphosphatemia in CKD [16]. In fact, Schiavi et al. [17] suggested that targeting Npt2b in addition to using dietary Pi binders may be a therapeutic approach to modulate serum Pi in CKD. Aside from this, the expression of Npt2b has also been detected in various tissues, including the small intestine, lung, kidney, testis, and liver [18]. Npt2b is also involved in the reuptake of phosphate for the synthesis of surfactant proteins in the lung [19,20]. Mutations in the Npt2b gene are linked to pulmonary alveolar microlithiasis (PAM), an autosomal recessive disorder characterized by the deposition of calcium phosphate microlith, and testicular microlithiasis (TM) — a disease that is associated with cancer and infertility [21].

Transcription factors precisely control the diversity and specificity of the complex patterns of gene regulation. The transcription factor CCAAT/enhancer-binding proteins (C/EBPs), of which there are six members (C/EBPα, C/EBPβ, C/EBPδ, C/EBPε, C/EBPγ, and C/EBPζ), are involved in adipocyte differentiation, energy metabolism, immunity, and inflammation [22,23]. In addition, C/EBP homolog protein (CHOP), which is activated by endoplasmic reticulum (ER) stress, mediates TNFα- and CKD-induced vascular calcification [24–26]. We recently reported that C/EBPβ contributes to vascular calcification via the up-regulation of the expression of type III sodium-dependent phosphate co-transporters (PiT1 and PiT2) [27]. The double-knockout Cebpa/Cebpb mice generate an early embryonic lethal phenotype [28]. On the other hand, Shibasaki et al. [29] demonstrated that the developmental deletion of the Npt2b gene leads to an embryonic lethal phenotype. Interestingly, Xu et al. suggested that C/EBPα controls surfactant lipid homeostasis by regulating the transcription of the Npt2b gene [30].

The physiological actions of ATRA, a metabolite of vitamin A, are mediated by specific nuclear receptors, including retinoic acid receptors (RARs) and retinoid X receptors (RXRs). These receptors are members of the steroid/thyroid hormone nuclear receptor superfamily, which act as ligand-dependent transcriptional factors. RARs and RXRs regulate the transcription of target genes by binding to retinoic acid-response elements (RAREs) in their promoters [31,32]. We previously reported that the renal expression of Npt2a and Npt2c is decreased in vitamin A-deficient (VAD) rats, and that the transcriptional activity of human Npt2a and Npt2c genes is up-regulated by ATRA and its receptors [4]. However, the effect of ATRA on the intestinal expression of Npt2b and the underlying molecular mechanism remain unclear. Furthermore, in the previous study, we did not examine the effects of ATRA treatment on phosphate homeostasis in VAD rats.

In the present study, we used VAD rats to investigate the effects of ATRA on the expression of Npt2b in the small intestine. We also characterized the Npt2b gene promoter with regard to transcriptional regulation through RAR/RXR and C/EBP.

Experimental procedures

Chemicals and reagents

ATRA, DMSO, and mouse anti-β-actin monoclonal antibody were purchased from Sigma–Aldrich (St. Louis, MO, U.S.A.). Buprenorphine hydrochloride was purchased from Otsuka Pharmaceutical Co., Ltd. (Tokyo, Japan). Pentobarbital sodium salt was purchased from Tokyo Kasei Co., Ltd. (Tokyo, Japan). Anti-Npt2b antibody was purchased from Alpha Diagnostics (San Antonio, TX, U.S.A.). Goat anti-rabbit IgG(H + L)-HRP conjugate was purchased from Bio-Rad (Hercules, CA, U.S.A.). ECL Plus system and poly(dI-dC) were purchased from GE Healthcare (Buckinghamshire, U.K.). QuikChange^® site-directed mutagenesis kit was purchased from Stratagene (La Jolla, CA, U.S.A.). 4-[E-2-(5, 6, 7, 8-Tetrahydro-5, 5, 8, 8-tetra-methyl-2-naphtalenyl)-1-propenyl] benzoic acid (TTNPB) was purchased from Biomol Research Laboratories (Boston, MA, U.S.A.). Double-strand Stealth RNAi oligos for C/EBPβ and negative control were purchased from Thermo Fisher Scientific (Waltham, MA, U.S.A.). [γ−³²P] ATP was purchased from ICN (Costa Mesa, CA, U.S. A.). T_NT^® Quick Coupled Transcription/Translation System was purchased from Promega Corporation (Madison, WI, U.S.A.). T₄ polynucleotide kinase was purchased from Takara (Shiga, Japan). C/EBP consensus oligonucleotide (cebp; catalog number sc-2525) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA, U.S.A.).

Animal experiments

The animal work took place in Division for Animal Researches and Genetic Engineering Support Center for Advanced Medical Sciences, Institute of Biomedical Sciences, Tokushima University Graduate School. The animals were housed in pathogen-free conditions and maintained under a standard 12 h light-dark cycle with free access to water. Briefly, pregnant Wistar rats ( Japan SLC, Shizuoka, Japan) were fed an altered VAD AIN93-G diet (Oriental Yeast, Osaka, Japan) containing 0.6% Pi and 0.6% Ca (VAD) or an altered AIN93-G diet (Oriental Yeast) containing 0.6% Pi and 0.6% Ca (control) from the 14th day of gestation until the pups were weaned. Pups (male) were continued on each diet as discussed above until they were killed at 7 weeks of age. The VAD group was randomly divided into two groups and intraperitoneally administrated a total of 1 mg/kg body weight of ATRA or DMSO prepared in 500 μl sterile saline. The control group was intraperitoneally administrated the same dose of DMSO. Each group (n = 4 per group) of rats was fasted for 18 h in metabolic cages with water ad libitum before sacrifice with a total of 0.1 mg/kg body weight of buprenorphine hydrochloride and a total of 50 mg/kg body weight of pentobarbital sodium salt, and the removal of tissues. Composition of the diets was described previously [4]. The present study was approved by the Animal Experimentation Committee of Tokushima University School of Medicine (animal ethical clearance No. T28–24) and was carried out in accordance with guidelines for the Animal Care and Use Committee of Tokushima University School of Medicine.

Plasma and urine parameters

Concentrations of Pi, Ca, creatinine (Cr), and vitamin A (retinol) were determined as described previously [4]. Concentrations of plasma 1,25(OH)₂D, PTH, and FGF-23 were determined as described previously [6]. Metabolic cages were used for 18 h urine collection. The fractional excretion indexes for Pi (FEI Pi) and for Ca (FEI Ca) were calculated as urine Pi or Ca/(urine Cr × plasma Pi or Ca).

Feces Pi extraction

Feces samples were collected during the period of 47–49 days to determine intestinal Pi and Ca excretion. The feces were first dried at 110°C for 12 h then micropulverized, from which 100 mg samples were ashed at 250°C for 3 h, at 350°C for 3 h, and 550°C for 24 h as previously described [33]. These samples were heated at 100°C for 15 min with 25 ml of 1% HCl. Extracted Pi and Ca were measured using a standard molybdate assay [34] and inductively coupled with plasma-mass spectrometry (ICP-MS).

Preparation of brush border membrane vesicles (BBMVs) and Pi uptake

BBMVs were prepared from rat small intestine and kidney by the Ca²⁺ precipitation method as described previously [3]. The uptake of ³²P into BBMVs was measured by a rapid filtration technique. Ten μl of vesicle suspension was added to 90 μl of incubation solution that was composed of 100 mM NaCl or choline chloride, 100 mM mannitol, 20 mM HEPES/Tris, and 0.1 mM $\mathrm{K}{\mathrm{H}}_2^{32}\mathrm{P}{\mathrm{O}}_4$, and the preparation was incubated at 20°C. Na⁺-dependent and Na⁺-independent Pi uptake were measured as described previously [35].

Western blot analysis

Protein samples were heated at 95°C for 5 min in sample buffer in the presence of 5% 2-mercaptoethanol and subjected to SDS–PAGE. The separated proteins were transferred by electrophoresis to polyvinylidene difluoride transfer membranes (Immobilon-P, Millipore, MA, U.S.A.). The membranes were treated with diluted affinity-purified anti-Npt2a (1 : 5000), and anti-Npt2c (1 : 500), and anti-Npt2b (1 : 2000) antibody [8]. Mouse anti-β-actin monoclonal antibody was used as an internal control. Goat anti-rabbit IgG(H + L)-HRP conjugate (1 : 2000) was utilized as the secondary antibody, and signals were detected using the ECL Plus system.

Quantitative PCR analysis

Extraction of total RNA, cDNA synthesis, and real-time PCR were performed as described previously [4]. The primer sequences (Npt2b, C/EBPα, and C/EBPβ) for PCR amplification are shown in Supplementary Table S1. Other primer sequences (Npt2a, Npt2c, PiT1, PiT2, and β-actin) were described previously [4]. Amplification products were then analyzed by a melting curve, which confirmed the presence of a single PCR product in all reactions (apart from negative controls). The PCR products were quantified by fit-point analysis, and results were normalized to that of β-actin.

Reporter plasmid construction

Luciferase reporter plasmids prNp2b-1.8k, prNp2b-800, prNp2b-67, prNp2b-55, pmNp2b-1.7k, pmNp2b-700, pmNp2b-39, and phNp2b-1.5k were constructed by PCR amplification of rat, mouse, or human genomic DNA as a template using gene-specific primers (Supplementary Table S2). These PCR products were subcloned into a pGL-3 or pGL-4.19 vector (Promega, Madison, WI, U.S.A.). Reporter plasmid prNp2b-180 was cloned by the digestion of prNp2b-800 using XhoI restriction enzyme. Reporter plasmids phNp2b-1.1k, phNp2b-200, and phNp2b+17 were cloned by the digestion of phNp2b-1.5k using SacI/HindIII, SacI/ApaI, and SmaI restriction enzymes, respectively. Mutated reporter plasmids prNp2b-180-Mut-GC-box (rMut-G), prNp2b-180-Mut-C/EBP (rMut-C), prNp2b-180-Mut-E-box (rMut-E), pmNp2b-700-Mut-GC-box (mMut-G), and pmNp2b-700-Mut-C/EBP (mMut-C) were constructed with the QuikChange^® site-directed mutagenesis kit using the oligonucleotides shown in Supplementary Table S2. The β-galactosidase expression vector pCMV-β (CLONTECH, Palo Alto, CA, U.S.A.) was used as an internal control. Each plasmid was purified with PureYield™ Plasmid Midiprep System (Promega).

Transfection and luciferase assay

NIH3T3 cells were obtained from Riken Cell Bank, Tokyo, Japan. NIH3T3 cells were cultured in DMEM at 37° C in an atmosphere containing 5% CO₂. The growth medium was supplemented with 10% (v/v) FBS, 100 units/ml penicillin, 100 μg/ml streptomycin, and 1 mM sodium pyruvate. Mouse RARα, RARβ, and RARγ expression vectors ( pSG5-RARα, pSG5-RARβ, and pSG5-RARγ), and mouse RXRα expression vector ( pSG5-RXRα) were kindly provided by Dr. P. Chambon. Mouse C/EBPα and C/EBPβ expression vectors ( pcDNA3.1-C/EBPα and pcDNA3.1-C/EBPβ) were constructed as previously described [36]. Transfection was performed as described previously [4]. Cells were then treated with several concentrations of ATRA, TTNPB or ethanol as vehicle control for an additional 16 h. Luciferase assay was performed as described previously [4].

Stealth RNAi

For the application of RNA interference technology, double-strand Stealth RNAi oligos designed using RNAi designer software (https://rnaidesigner.thermofisher.com/rnaiexpress/) were synthesized by Thermo Fisher Scientific. The target sequence for C/EBPβ (NM_009883) is as follows: 5′-AGACCCATGGAAGTGGCCAACTTCT-3′. For the control, Stealth RNAi Negative Control Duplexes were used.

Coupled transcription/translation assays

Each C/EBPβ, RARβ, and RXRα proteins were synthesized with the T_NT^® Quick Coupled Transcription/Translation System at 30°C for 90 min in the presence of 20 μM methionine. Generated proteins were used for EMSAs.

EMSA

EMSA was performed as described previously [3]. Double-stranded nucleotides for Npt2b-C/EBP and Npt2b-C/EBP-Mutant (Mut) were synthesized (Supplementary Table S3). Purified DNA fragments were radiolabeled with [γ−³²P] ATP (110 TBq/mmol) using T₄ polynucleotide kinase. Nuclear extracts (RARα, pSG5-RXRα, and C/EBPβ) were prepared as described previously [4]. Briefly, the NIH3T3 cells were cultured in 35-mm dishes to 90% confluence and transfected with pSG5-RARα, pSG5-RXRα, and pcDNA3.1-C/EBPβ. Prepared nuclear extracts (15 μg) were incubated with the radiolabeled probe in binding buffer [10 mM (Tris–HCl), pH7.5, 1 mM DTT, 1 mM EDTA, 10% Glycerol, 1 mM MgCl₂, 0.25 mg/ml bovine serum albumin, 2.5 μg/ml salmon sperm DNA and 2 μg poly(dI-dC)] in a final volume of 20 μl for 30 min at room temperature. Specificity of the binding reaction was determined with a 100-fold molar excess of the indicated cold competitor oligonucleotide. The reaction mixture was then subjected to electrophoresis on a 5% polyacrylamide gel with 0.25 × TBE running buffer for 2 h at 150 V. The gel was dried and analyzed with a Fujix Bio-imaging analyzer (BAS-1500, Fuji-film, Tokyo, Japan).

Statistical analysis

Data were collected from more than two independent experiments and were reported as the means ± S.E.M. Statistical analysis for two-group comparison was performed using a two-tailed Student’s t-test, or one-way ANOVA with a Student-Newman post-hoc test for multi-group comparison. All data analysis was performed using GraphPad Prism 5 software (Graphpad Software, San Diego, CA, U.S.A.). P < 0.05 was considered statistically significant.

Results

ATRA alleviates VAD-induced hyperphosphaturia through the up-regulation of the Npt2 gene expression in the rat kidney

Previously, we reported that VAD rats had decreased Pi uptake, however, the effect of ATRA on the intestinal Pi uptake remains unclear [4]. Furthermore, we did not examine the effects of ATRA treatment on phosphate homeostasis in VAD rats. We made VAD rats and these rats were treated with a total of 1 mg/kg body weight of ATRA. Because ATRA cannot be converted to retinol, plasma levels of retinol were undetectable in both VAD and VAD + ATRA rats, which was in line with our expectations (control: 511 ± 12.2 μg/dl). Although the plasma Pi and Ca levels were not changed among three groups of rats, ATRA treatment significantly reduced the urine Pi/Cr and Ca/Cr ratios, which had been increased by VAD (Table 1). Levels of plasma Pi-regulating hormones (1,25[OH]₂D₃, PTH, and FGF23) did not differ between VAD and control rats. In VAD rats, the plasma PTH level was not affected by ATRA, whereas the plasma 1,25(OH)₂D₃ and FGF23 levels were significantly reduced by ATRA (Table 1). The Na⁺-dependent Pi uptake activity — but not the Na⁺-independent Pi uptake activity — in renal BBMVs was markedly decreased in VAD rats, which was significantly increased by ATRA treatment (Figure 1A). As we previously reported, Western blotting revealed that the expression levels of renal Npt2a and Npt2c proteins in VAD rats were significantly decreased in comparison with controls. Furthermore, the decreased expression of Npt2c protein — but not Npt2a — in the kidney of VAD mice was partially restored by ATRA treatment (Figure 1B). Next, we performed real-time PCR to measure the renal Npt2a and Npt2c mRNA expression. The decreased mRNA levels of renal Npt2a and Npt2c in VAD rats were significantly increased by ATRA treatment (Figure 1C).

Table 1.

Effects of vitamin A on plasma, urine Pi levels, and fecal Pi excretion

	Control	VAD	VAD + ATRA
Plasma
retinol (μg/dl)	511 ± 12.2	U.D.	U.D.
Pi (mg/dl)	5.58 ± 0.21	5.84 ± 0.37	5.24 ± 0.17
Ca (mg/dl)	11.1 ± 0.13	10.9 ± 0.46	10.7 ± 0.56
PTH (pg/ml)	124 ± 9.73	128 ± 10.7	135 ± 6.74
FGF23 (pg/ml)	296 ± 17.1	258 ± 20.6	164 ± 24.8#
1,25(OH)2D3 (pg/ml)	544 ± 19.4	575 ± 48.6	423 ± 5.44#
Urine
Pi/Cr	1.34 ± 0.30	4.71 ± 0.60*	2.82 ± 0.23#
Ca/Cr	0.26 ± 0.10	0.87 ± 0.03*	0.47 ± 0.04#
FEI Pi	0.25 ± 0.06	0.78 ± 0.10*	0.51 ± 0.05
FEI Ca	0.02 ± 0.01	0.08 ± 0.01*	0.04 ± 0.01#
Fecal
Pi (mg/day)	46.4 ± 2.09	35.3 ± 2.07*	N.D.
Ca (mg/day)	78.7 ± 6.71	62.3 ± 9.45*	N.D.

Value are mean ± S.E.M. (n = 3–4).

^*P <0.05 vs. control

^#P <0.05 vs. VAD. (one-way ANOVA with a Student-Newman post-hoc test).

U.D.: Undetectable. N.D.: not determined.

Figure 1. Effects of ATRA treatment on the expression of renal Npt2a and Npt2c in VAD rats.

Seven-week-old male rats with VAD induced by a VAD diet were randomly divided into two groups and treated with DMSO (VAD) or ATRA (VAD + ATRA). (A) The Na⁺-dependent and Na⁺-independent Pi transport activity were assessed by measuring the uptake of Pi in renal BBMVs. (B) Western blotting of Npt2a and Npt2c in renal BBMVs. Each lane was loaded with 25 μg of BBMVs. β-actin was used as an internal control. (C) The Npt2a and Npt2c mRNA levels in renal BBMVs were analyzed by quantitative PCR. β-actin was used as an internal control. Values are the mean ± S.E.M. (n = 3–4). * P < 0.05, ** P < 0.01 (one-way ANOVA with a Student-Newman post-hoc test).

ATRA reduces VAD-induced Pi absorption through the expression of the Npt2b gene in the rat jejunum

Because Pi homeostasis is strictly controlled by intestinal absorption and renal excretion, we next investigated whether ATRA regulates the Pi absorption and Npt2b expression in the intestine. We used the jejunum for this analysis because the levels of Npt2b protein and mRNA in the jejunum are higher than those in the duodenum and ileum of rats [10]. As expected, the fecal Pi excretion in VAD rats was significantly decreased in comparison with controls (Table 1). The Na⁺-dependent Pi uptake activity in intestinal BBMVs was markedly increased in VAD rats, and this was significantly reduced by ATRA treatment (Figure 2A and Supplementary Figure S1). Western blotting revealed that ATRA treatment blocked the VAD-induced Npt2b protein expression in the jejunum (Figure 2B). As shown in Figure 2C, real-time PCR demonstrated that the VAD-induced jejunal Npt2b mRNA levels were significantly suppressed by ATRA. However, the jejunal PiT1 and PiT2 mRNA expression levels were not changed among the three groups (data not shown).

Figure 2. Effects of vitamin A on the expression of intestinal and extraintestinal Npt2b in rats.

(A) Na⁺-dependent and Na⁺-independent Pi transport activity was assessed by measuring the uptake of Pi in intestinal BBMVs. (B) Western blotting of Npt2b in BBMVs. Each lane was loaded with 25 μg of BBMVs. β-actin was used as an internal control. (C) The mRNA expression levels of Npt2b in various tissues (e.g. duodenum, jejunum, ileum, lung, liver, kidney, and spleen) of rats were determined by RT-PCR (lower) and quantitative PCR (upper). β-actin was used as an internal control. Values are the mean ± S.E.M. (n = 3–4). * P < 0.05, ** P < 0.01 (one-way ANOVA with a Student-Newman post-hoc test).

Because the expression of the Npt2b gene has been detected in various tissues, including the small intestine, lung, kidney, and liver [18], we also investigated the effects of ATRA on the Npt2b mRNA expression in several extraintestinal organs by real-time PCR. Unlike the jejunum, the Npt2b mRNA expression in the duodenum and ileum was low, and was not changed among three groups of rats (Figure 2C). In the lung, the Npt2b mRNA expression was considerably high, but was not changed among three groups. The hepatic Npt2b mRNA expression was regulated by ATRA, similarly to the jejunum. ATRA did not affect the reduced expression of the renal Npt2b mRNA in VAD rats. The expression of Npt2b mRNA was almost undetectable in the spleen.

ATRA down-regulates the transcriptional activity of the rat Npt2b gene promoter

We previously reported that the transcriptional activities of human Npt2a and Npt2c genes are controlled by ATRA and its receptors [4]. To investigate the molecular mechanisms underlying the regulation of the Npt2b gene expression by ATRA, we examined the responsiveness of rat Npt2b gene promoters to ATRA using a luciferase assay. Because we thought that the effects of ATRA on the Npt2b gene expression differ among tissues, we used NIH3T3 cells for a luciferase assay to eliminate tissue specific factor. prNp2b-1.8k, phNp2b-1.5k, and pmNp2b-1.7k reporter constructs, which respectively contained the promoter and exon 1 fragments of the rat, human, and mouse Npt2b gene, were utilized for a luciferase assay in NIH3T3 cells. While ATRA had little impact on the transcriptional activity of prNp2b-1.8k without the overexpression of RAR/RXR in NIH3T3 cells, its activity was markedly inhibited by co-overexpressing RARs (RARα, RARβ, or RARγ)/RXR (Figure 3A). Furthermore, ATRA additively reduced the promoter activity of rat Npt2b that was reduced by the overexpression of RARs/RXR in NIH3T3 cells (Figure 3A). Next, ATRA dose-dependently reduced the rat Npt2b gene promoter activity in NIH3T3 cells overexpressing RARα/RXR, as well as TTNPB, an RAR-specific agonist (Figure 3B). The phNp2b-1.5k and pmNp2b-1.7k reporter constructs exhibited similar responses to ATRA and TTNPB, as did prNp2b-1.8k (Supplementary Figure S2). A histone modification, a covalent post-translational modification (PTM), such as histone acetylation and methylation, is involved in regulating the transcription [37]. To test whether the inhibition of the Npt2b gene promoter activity by ATRA is associated with the effects of histone acetylation and methylation, we determine the luciferase activity of prNp2b-1.8k with Trichostatin A (histone deacetylase inhibitor, TSA) or 5-aza-2deoxycytidine (methylation inhibitor, 5--Aza-2dc). Neither TSA nor 5-Aza-2dc affected the suppression of the luciferase activity of prNp2b-1.8k by ATRA (Figure 3C).

Figure 3. Suppression of rat Npt2b gene promoter by ATRA and its receptors in NIH3T3 cells.

(A) A Schematic illustration of the rat Npt2b gene promoter in the upper panels. prNp2b-1.8k and pCMV-β were transfected with pSG5-RAR (α, β, γ) and pSG5-RXRα, or empty vector and incubated in the presence of 100 nM ATRA or ethanol as a vehicle control for 24 h in NIH3T3 cells. * P < 0.05 vs. empty vector. ^#P < 0.05 vs. vehicle (one-way ANOVA with a Student-Newman post-hoc test). (B) NIH3T3 cells were transfected with prNp2b-1.8k, pSG5-RARα, pSG5-RXRα, and pCMV-β and treated with the indicated concentrations of ATRA (white circles) or TTNPB (black circles) for 24 h. (C) NIH3T3 cells were transfected with prNp2b-1.8k, pSG5-RARα, pSG5-RXRα, and pCMV-β and incubated in the presence of 100 nM ATRA with 100 nM TSA or 5-Aza-2dc (5-Aza). Each point represents the average of quadruplicate analyses ± S.E.M. normalized for β-gal activity. Similar results were obtained from three independent experiments. * P < 0.05, N.S. = not significant (one-way ANOVA with a Student-Newman post-hoc test).

The deletion analysis of the Npt2b gene promoter

In the search for conserved putative regulatory elements, the sequence of the rat Npt2b gene promoter region (−207 to +33) relative to the transcriptional start site was compared with the corresponding regions of the human (−146 to +83) and mouse (−172 to +49) sequences. As shown in Figure 4A, highly conserved nucleotide sequences were determined among human, rat, and mouse Npt2b gene promoters with a minimal sequence similarity of 73%. Interestingly, a search for transcription factor binding motifs within this region suggested some potential consensus binding site such as GC-box, C/EBP, and E-box (Figure 4A). In order to understand the molecular mechanisms underlying the responsiveness of these Npt2b genes to ATRA, several reporter constructs lacking portions of the 5′-promoter region of the human, rat, and mouse Npt2b genes were tested in NIH3T3 cells overexpressing RAR/RXR, with or without ATRA. These deletion analyses suggest that the C/EBP binding site in the Npt2b gene promoter is involved in the down-regulation of the transcriptional activity of the Npt2b gene by ATRA and its receptors (Figure 4B).

Figure 4. Nucleotide sequence of 5’-flanking region in the Npt2b gene and the deletion analysis.

(A) The sequences of the human, rat, and mouse Npt2b gene promoter regions (human, −146 to +83 bp; rat, −207 to +33 bp; mouse: −172 to +49 bp). The boxes indicate the putative binding sites for various transcription factors (STAT, GC-box, C/EBP, TATA-box, E-box). Asterisk indicates homology of sequences in the Npt2b gene promoter among these mammalians species. (B) Transcriptional activity of deletion constructs of human, rat, and mouse Npt2b gene promoters ( phNp2b-1.5k, prNp2b-1.8k, and pmNp2b-1.7k). Deletion constructs are illustrated in the panels on the left. NIH3T3 cells were transfected with the indicated human, rat, or mouse Npt2b gene reporter constructs and pSG5-RARα, pSG5-RXRα, and pCMV-β and treated with 100 nM ATRA or ethanol (NT) for 24 h. Each point represents the average of quadruplicate analyses ± S.E.M. normalized for β-gal activity. Similar results were obtained from three independent experiments. * P < 0.001 vs. NT (two-tailed unpaired Student’s t-test).

RAR/RXR blocks the binding of C/EBP to the Npt2b gene promoter

To further test whether the C/EBP binding site is the first target of the down-regulation of the Npt2b gene promoter by ATRA and its receptors, we determined the luciferase activity of prNp2b-Mut-GC-box (rMut-G), prNp2b-Mut-C/EBP (rMut-C), and prNp2b-Mut-E-box (rMut-E), the sequences and constructs of which are shown in Figure 5A. These mutation analyses showed that the C/EBP binding site — but not GC-box or E-box — was essential for the repression of the Npt2b gene promoter activity by ATRA (Figure 5B and Supplementary Figure S3C). Next, we investigated how the transcriptional factor C/EBP actually contributes to the suppression of the transcriptional activity of the Npt2b gene by ATRA and its receptors. The overexpression of C/EBPα or C/EBPβ each increased the promoter activity of the rat Npt2b construct to more than double the original level. However, ATRA additively diminished the transcriptional activity of the Npt2b gene, which was reduced by the overexpression of RAR/RXR (Figure 5C). Next, we analyzed levels of mRNA expression of C/EBPα and C/EBPβ in NIH3T3 cells using qPCR analysis with the absolute standard curve method. Unlike C/EBPβ, C/EBPα gene was not expressed at all in NIH3T3 cells. Therefore, we selected C/EBPβ for gene knockdown experiments, but not C/EBPα. We generated NIH3T3 cells with the knockdown of C/EBPβ, endogenous C/EBPβ mRNA levels of which were reduced by more than 50%, using C/EBPβ-specific siRNA (data not shown). As shown in Figure 5D, the luciferase activity of prNp2b-180 was reduced by the siRNA-mediated knockdown of C/EBPβ in NIH3T3 cells. Furthermore, C/EBPβ siRNA significantly ameliorated the reduction in the Npt2b gene promoter activity by ATRA (Figure 5E). Next, to elucidate how ATRA and its receptors downregulate the transcriptional activity of the Npt2b gene through the action of C/EBP, we examined whether RAR/RXR affects the binding of C/EBP to the Npt2b gene promoter by an EMSA analysis. As shown in Figure 5F, a radiolabeled oligonucleotide containing an Npt2b-C/EBP probe detected a band in nuclear extracts prepared from NIH3T3 cells overexpressing C/EBPβ, but not RAR/RXR. Although these complexes are susceptible to competition with unlabeled Npt2b-C/EBP, consensus C/EBP, and an antibody against C/EBPβ, unlabeled mutated oligonucleotide (Mut: mutated Npt2b-C/EBP) did not compete with these complexes. The formation of these complexes with C/EBPβ was inhibited in the presence of nuclear extracts prepared from NIH3T3 cells overexpressing RAR/RXR (Figure 5F). Likewise, although in vitro synthesized C/EBPβ recombinant protein bound to this probe, this DNA–protein complex was inhibited by RARβ/RXRα recombinant protein and an antibody against C/EBPβ (Supplementary Figure S4).

Figure 5. C/EBP binds to the C/EBP binding site in the Npt2b gene promoter and the mutation analysis of the Npt2b gene.

(A) Transcription factor binding sites mutated in the rat Npt2b gene promoter region are underlined. rMut-G, rMut-C, and rMut-E targeted the binding sites for transcription factors GC-box, C/EBP, and E-box, respectively. A schematic diagram showing the wild-type (rWT) rat Npt2b reporter plasmid as well as reporter plasmid with mutations (rMut-G, rMut-C, and rMut-E) in transcription factors sequences, shown with an X through the mutant sequences. (B) Each rat Npt2b reporter plasmid (rWT, rMut-G, rMut-C, and rMut-E) was transfected with pSG5-RARα, pSG5-RXRα, and pCMV-β into NIH3T3 cells. Cells were treated with vehicle (NT: ethanol) or 100 nM ATRA and cell lysates were assessed for β-gal and luciferase activities 24 h later. * P < 0.01 vs. NT (two-tailed unpaired Student’s t-test). (C) NIH3T3 cells were transfected with prNp2b-180, pCMV-β, pSG5-RARα, pSG5-RXRα, pcDNA3.1-C/EBP (α or β), or empty vector and incubated in the presence of 100 nM ATRA or ethanol and cell lysates were assessed for β-gal and luciferase activity 24 h later. * P < 0.01 (one-way ANOVA with a Student-Newman post-hoc test). (D) NIH3T3 cells were transfected with prNp2b-180 and pCMV-β 48 h after transfection with C/EBPβ siRNA (10 or 100 pmol) or control (Cont.) and cell lysates were assessed for β-gal and luciferase activity 24 h later. *P < 0.01 vs. Cont. (two-tailed unpaired Student’s t-test). (E) NIH3T3 cells were transfected with prNp2b-180, pCMV-β, pSG5-RARα, pSG5-RXRα, or empty vector 48 h after transfection with C/EBPβ siRNA (100 pmol) or control and incubated in the presence of 100 nM ATRA or ethanol and cell lysates were assessed for β-gal and luciferase activity 24 h later. *P < 0.01 (one-way ANOVA with a Student-Newman post-hoc test). (F) EMSAs using ³²P-labelled Npt2b-C/EBP as probes. EMSAs were performed with nuclear extracts (N.E.) from NIH3T3 cells overexpressing C/EBPβ, RARβ, and RXRα, with the addition of unlabeled competitor oligonucleotides as indicated. A 100-fold molar excess of each competitor was used. The location of the DNA–protein complex band is indicated by an arrowhead. cebp, C/EBP-binding sequence; Mut, mutated Npt2b-C/EBP; αC/EBP, C/EBPβ-specific antibody. Each point represents the average of quadruplicate analyses ± S.E.M. normalized for β-gal activity. Similar results were obtained from three independent experiments.

Discussion

In the present study, we have determined that the reduction in the intestinal Pi uptake activity and the Npt2b expression in VAD rats were ameliorated by ATRA treatment. Furthermore, we revealed that ATRA reduced the transcriptional activity of the Npt2b gene by inhibiting the binding of C/EBP to the Npt2b gene promoter. Previously, we reported that VAD induced hyperphosphaturia through the down-regulation of the Npt2a and Npt2c gene expression in the kidney, without changing the plasma Pi levels [4]. From these inconsistent results, we hypothesized that ATRA might affect not only renal Pi reabsorption but also intestinal Pi absorption. In this study, we found that ATRA increased the renal Pi uptake through the induction of the expression of the Npt2a and Npt2c genes, the levels of which are reduced by VAD, whereas the VAD-induced uptake of Pi via the Npt2b gene expression in the jejunum of VAD rats was reduced by the administration of ATRA. Together, these results suggested that ATRA did not change the plasma Pi levels because of the opposite effects of ATRA on the intestinal Pi uptake and the renal Pi uptake. This is the first report to demonstrate the presence of a Pi regulating factor that has opposite effects on the expression patterns of Npt2a and Npt2c in the kidney and Npt2b in the intestine. However, intestinal Pi absorption is mediated by two pathways: sodium-dependent Pi transport by Npt2b and PiT1/2 via a transcellular pathway; and sodium-independent Pi transport via a paracellular pathway [16]. It was recently reported that tenapanor, an inhibitor of the sodium/hydrogen exchanger (NHE3), significantly reduced intestinal Pi absorption in healthy volunteers and improved hyperphosphatemia in both rodents and humans with CKD [16]. The conformational change of the tight junctions in the enterocytes by NHE3 inhibition leads to increased transepithelial electrical resistance, which contributes to the reduction in permeability to phosphate. Interestingly, it has been reported that retinoic acid can enhance the intestinal epithelial barrier by increasing tight junction protein levels [38,39]. It was suggested that ATRA reduces intestinal Pi absorption through not only a reduction in the Npt2b expression but also the enhancement of the tight junction function in intestinal enterocytes. These data and reports suggest that ATRA regulates Pi homeostasis in the body through the positive and negative regulation of Pi (re)absorption in the kidney and jejunum. Interestingly, it has been suggested that a gut-derived factor called intestinal phosphatonin is released in response to ingestion of dietary Pi and rapidly modulates renal Pi reabsorption to prevent large post-prandial fluctuations in serum phosphate levels [16]. Although a large part of intestinal phosphatonin remains unclear, ATRA may affect this unknown factor.

Intestinal Pi absorption is modulated by 1,25(OH)₂D₃ as a positive regulator and by FGF-23 as a negative regulator [9,11]. In the present study, the levels of plasma 1,25(OH)₂D₃ and FGF-23 did not differ between VAD and control rats, as previously reported [4], whereas the treatment of VAD rats with ATRA reduced the plasma 1,25(OH)₂D₃ and FGF-23 levels. We considered that ATRA might reduce the plasma FGF23 levels in VAD rats through the reduction in the plasma 1,25(OH)₂D₃ levels, because 1,25(OH)₂D₃ is positive regulator of the FGF23 gene expression [40]. However, it also remains unclear why ATRA reduced the plasma 1,25 (OH)₂D₃ levels in VAD rats. Interestingly, an experiment to investigate the effects of retinol and retinoic acid on metabolism of 25(OH)D (which is converted to 1,25[OH]₂D₃ by CYP27B1 and which is converted to 24,25 (OH)₂D by CYP24A1) in kidney cell culture may answer to this question. Treatment with retinoic acid — but not retinol — for 6 h reduced the production of 1,25(OH)₂D₃ and increased the production of 24,25(OH)₂D in kidney cells [41]. These findings suggest that the administration of ATRA to VAD rats might reduce plasma 1,25(OH)₂D₃ levels through the reduction in the conversion of 25(OH)D to 1,25(OH)₂D₃. Furthermore, these observations suggest that the VAD-mediated up-regulation of the Npt2b gene expression was independent of the plasma 1,25(OH)₂D₃ and FGF-23 levels, whereas the ATRA-mediated down-regulation of the Npt2b gene expression in VAD rats could be partially induced by decreased plasma 1,25(OH)₂D₃. Although it has been reported that 1,25(OH)₂D₃ induces Npt2b gene promoter activity, a classical vitamin D responsive element (VDRE) in the Npt2b gene promoter has not been identified [42].

We reported that the transcriptional activity of human Npt2a and Npt2c genes was up-regulated by ATRA and its receptors [4]. Thus, we hypothesized that ATRA might control the expression of the Npt2b gene through the regulation of transcription. While ATRA had little impact on the transcriptional activity of prNp2b-1.8k without the overexpression of RAR/RXR in NIH3T3 cells, its activity was markedly inhibited by co-overexpressing RARs/RXR without the addition of ATRA. Furthermore, ATRA additively reduced the promoter activity of rat Npt2b that was reduced by the overexpression of RARs/RXR in NIH3T3 cells. These results may suggest that RAR/RXR is more important for the down-regulation of the transcriptional activity of rat Npt2b gene than ATRA. On the other hand, the promoter activity of the Npt2b gene was increased by either C/EBPα or C/EBPβ. Xu et al. reported that regulation of mouse Npt2b gene transcription was activated by C/EBPα in a lung epithelial cell line using transient transfection promoter assays [30]. Schwarz et al. reported that liganded RAR inhibits adipogenesis by blocking the C/EBPβ-mediated induction of downstream genes [43]. Although we have investigated whether ATRA regulates the C/EBP expression in the small intestine of rats, the jejunal C/EBPα and C/EBPβ mRNA expression levels remained unchanged in VAD rats and VAD rats treated with ATRA (data not shown). Furthermore, ATRA treatment did not affect the C/EBPβ mRNA expression in NIH3T3 cells (data not shown). Next, although we have investigated whether RAR could displace C/EBP from binding to a canonical C/EBP binding site, an EMSA analysis showed that RAR/RXR could not bind to oligonucleotide containing an Npt2b-C/EBP probe. Surprisingly, it has been reported that C/EBPβ could bind to the glucocorticoid receptor and RAR [44,45]. We also demonstrated that RAR/RXR inhibited the binding of C/EBP to the promoter of the Npt2b gene using an EMSA analysis. These data suggest that liganded RAR/RXR inhibits the binding C/EBPα or C/EBPβ to the C/EBP binding site in the Npt2b gene promoter, resulting in the suppression of the transcription activity of the Npt2b gene. Interestingly, we found that ATRA has a distinct effect on the expression patterns of Npt2b mRNA in different organs. C/EBPs (C/EBPα, β, δ, ε, γ, and ζ), which are expressed at different levels in each tissue, respectively regulate the diversity and specificity of the complex patterns of gene regulation [22,23]. That is to say, the difference in the C/EBPs gene expression levels in each tissue may contribute to a distinct effect on the Npt2b mRNA expression patterns by VAD and ATRA treatment in these tissues. We also reported that a complex of C/EBPβ and activating transcription factor-4 (ATF4) activates the expression of type III sodium-dependent Pi co-transporters (PiT1 and PiT2) through transcriptional regulation [27]. Because C/EBP might be involved in the regulation of Pi homeostasis in various ways, further studies are necessary to fully elucidate the underlying mechanisms.

Hyperphosphatemia, which is associated with the pathophysiology of CKD, can lead to vascular calcification, which has been linked to increased cardiovascular morbidity and mortality [15]. From studies on the regulation of intestinal Pi transport and metabolism, researchers have targeted Npt2b for alleviating the hyperphosphatemia of CKD [17]. Niacin and its derivative nicotinamide have been reported to reduce the Npt2b expression and lower the plasma Pi levels in CKD animals [46,47]. However, the exact mechanism through which these compounds reduce intestinal Pi absorption has not been completely elucidated. Interestingly, Fujimori et al. [48] showed that niacin reduced the expression of C/EBPβ mRNA in the early phase of adipogenesis. In other words, the molecular mechanism through which niacin diminishes the expression of Npt2b may be associated with the reduction in the C/EBPβ expression by niacin. In the present study, we indicated that ATRA reduces the transcriptional activity of the Npt2b gene by inhibiting the binding of C/EBP to the Npt2b gene promoter. That is to say, ATRA and C/EBP may become therapeutic targets for the prevention of hyperphosphatemia in CKD. Interestingly, elevated plasma retinol, ATRA, and retinol binding protein 4 (RBP4) levels have been observed in CKD patients and animal models [49,50]. However, the effects of ATRA on intestinal Pi absorption focusing on the expression of Npt2b through the action of C/EBP in CKD are not understood. The clarification of these effects would be valuable for advancing the treatment of CKD.

In conclusion, our findings reveal that the down-regulation of the intestinal Npt2b gene and the up-regulation of the renal Npt2a and Npt2c gene expression by ATRA contributes to Pi homeostasis in rats. Furthermore, it is clear that ATRA and its receptors can suppress the transcriptional activity of Npt2b gene promoters by blocking the C/EBP-mediated induction of its gene promoter activity.

Abbreviations

1,25(OH)₂D₃

1α,25-dihydroxyvitamin D₃

ATRA

all-trans-retinoic acid

BBMV

brush-border membrane vesicle

C/EBP

CCAAT/enhancer-binding protein

CKD

chronic kidney disease

Cr

creatinine

EMSA

electrophoretic mobility-shift assay

FBS

fetal bovine serum

FEI

fractional excretion index

FGF23

fibroblast growth factor 23

ICP-MS

inductively coupled with plasma-mass spectrometry

Npt

sodium-dependent phosphate co-transporter

Pi

inorganic phosphate

PTH

parathyroid hormone

RAR

retinoic acid receptor

RARE

retinoic acid-response element

RXR

retinoid X receptor

TTNPB

4-[E-2-(5, 6, 7, 8-Tetrahydro-5, 5, 8, 8-tetra-methyl-2-naphtalenyl)-1-propenyl] benzoic acid

VAD

vitamin A-deficient

β-gal

β-galactosidase