SEARCH

SEARCH BY CITATION

Keywords:

  • leptin;
  • FGF-23;
  • 1,25(OH)2D3;
  • bone;
  • kidney

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Leptin is the LEP (ob) gene product secreted by adipocytes. We previously reported that leptin decreases renal expression of the 25-hydroxyvitamin D3 1α-hydroxylase (CYP27B1) gene through the leptin receptor (ObRb) by indirectly acting on the proximal tubules. This study focused on bone-derived fibroblast growth factor 23 (FGF-23) as a mediator of the influence of leptin on renal 1α-hydroxylase mRNA expression in leptin-deficient ob/ob mice. Exposure to leptin (200 ng/mL) for 24 hours stimulated FGF-23 expression by primary cultured rat osteoblasts. Administration of leptin (4 mg/kg i.p. at 12-hour intervals for 2 days) to ob/ob mice markedly increased the serum FGF-23 concentration while significantly reducing the serum levels of calcium, phosphate, and 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3]. Administration of FGF-23 (5 µg i.p. at 12-hour intervals for 2 days) to ob/ob mice suppressed renal 1α-hydroxylase mRNA expression. The main site of FGF-23 mRNA expression was the bone, and leptin markedly increased the FGF-23 mRNA level in ob/ob mice. In addition, leptin significantly reduced 1α-hydroxylase and sodium-phosphate cotransporters (NaPi-IIa and NaPi-IIc) mRNA levels but did not affect Klotho mRNA expression in the kidneys of ob/ob mice. Furthermore, the serum FGF-23 level and renal expression of 1α-hydroxylase mRNA were not influenced by administration of leptin to leptin receptor–deficient (db/db) mice. These results indicate that leptin directly stimulates FGF-23 synthesis by bone cells in ob/ob mice, suggesting that inhibition of renal 1,25(OH)2D3 synthesis in these mice is at least partly due to elevated bone production of FGF-23. © 2010 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Leptin, the product of the LEP (ob) gene, is an unglycosylated peptide of 146 amino acids with a molecular weight of 16,000 that was discovered in 1994 by Friedman and colleagues.1 This cytokine-like hormone has multiple functions, including the regulation of food intake, energy expenditure, and adiposity. Leptin is produced by white adipose tissue and is secreted in response to changes in body energy stores.1, 2 Plasma leptin levels are increased in obesity and are strongly related to the body fat mass.3 Circulating leptin acts on certain hypothalamic nuclei after crossing the blood-brain barrier to regulate food intake, energy expenditure, growth, and sexual maturation.4, 5 Leptin also acts peripherally because leptin receptors have been detected in various tissues,6, 7 and it is secreted by the placenta and gastric mucosa.8, 9

In C57BL/6J mice, a single-base mutation of the ob gene (codon 105) results in the production of a truncated and inactive form of leptin. In homozygous animals (ob/ob mice), this mutation results in a naturally occurring knockout model of leptin deficiency, which has phenotypic features that include obesity, increased body fat mass, hyperglycemia, hyperinsulinemia, hypothermia, sterility, and impaired thyroid function.10 In db/db mice, a mutation of the leptin receptor gene causes abnormal splicing, and these animals develop obesity and type 2 diabetes as a consequence of the loss of leptin activity.11 The mutant leptin receptor gene in db/db mice is ObRb, which has a long cytoplasmic domain with consensus amino acid sequences involved in binding to Janus tyrosine kinases (JAK).12, 13 Since the db/db mouse displays the phenotypic features of leptin receptor deficiency, it is an invaluable animal model for studies of leptin signaling.

Osteoporosis and obesity show an inverse correlation because bone mineral density (BMD) increases along with body weight and body fat mass, so obesity is associated with an increased BMD and has a protective effect against osteoporosis.14 These findings suggest that important interactions may exist between bone metabolism and obesity (body fat). Karsenty and colleagues15, 16 reported that leptin acts on hypothalamic neurons to inhibit bone formation by stimulating sympathetic postganglionic neurons via axons that enter the bones. In contrast, leptin administration was found to significantly increase femoral length, total-body bone area, and BMD in male ob/ob mice.17 Administration of leptin to female rats is effective against trabecular bone loss associated with estrogen deficiency but cannot fully reverse the bone loss that occurs after ovariectomy.18 Thus leptin appears to have multiple central and peripheral effects on bone metabolism.19 Recently, we demonstrated that leptin suppresses the overexpression of 25-hydroxyvitamin D3 [25(OH)D3] 1α-hydroxylase (CYP27B1) and 24-hydroxylase (CYP24) in the kidneys of ob/ob mice and also corrects hypercalcemia and hyperphosphatemia in these animals.20 The low BMD of leptin-deficient mice thus seems to be related to stimulation of bone resorption by 1α,25-dihydroxyvitamin D3 [1,25(OH)2D3]. Furthermore, we previously found that leptin attenuates renal 1α-hydroxylase expression via the ObRb. However, leptin does not inhibit 1α-hydroxylase expression in primary cultures of mouse renal tubular cells, suggesting that other factors are involved in regulating the expression of the gene for this enzyme in the proximal tubules.21

Fibroblast growth factor 23 (FGF-23) is a circulating phosphaturic factor that plays a critical role in the reabsorption of inorganic phosphate (Pi) and in renal vitamin D metabolism.22 FGF-23 is secreted by bone cells such as osteocytes and osteoblasts in response to an excessively high serum Pi concentration23, 24 and acts on target tissues such as the kidneys by binding to and activating FGF receptors (FGFRs) in the presence of an obligatory coreceptor (Klotho). Klotho is a transmembrane protein that acts as a cofactor in facilitating the binding of FGF-23 to FGFR1c, -3c, and -4.25, 26 Expression of Klotho has been observed in the distal renal tubules, parathyroid gland, and pituitary gland.27 In the kidneys, FGF-23 inhibits phosphate reabsorption and reduces the synthesis of 1,25(OH)2D3.28 Membrane-bound Klotho has been detected in the distal renal tubules and has been shown to inhibit the activity of the Na-Pi cotransporter in the proximal tubules,29, 30 possibly owing to the effect of a secreted form of Klotho.31

These findings strongly suggest that leptin may act directly on bone and may modulate osteoblast and osteocyte function. Therefore, we hypothesized that administration of leptin could suppress renal 1α-hydroxylase expression via bone-derived FGF-23. To test this hypothesis, we investigated leptin-deficient (ob/ob) and leptin receptor–deficient (db/db) mice and determined whether administration of leptin enhanced the production of FGF-23 by bone.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Animals

This animal study was approved by the Institutional Animal Care and Use Committee of Ohu University and was conducted in accordance with the Guidelines for the Care and Use of Laboratory Animals. Six-week-old male leptin-deficient mice (C57BL/6J ob/ob) and male leptin receptor–deficient mice (C57BL/KsJ db/db) bred by Jackson Laboratory (Bar Harbor, ME, USA) were purchased from Charles River Japan (Tokyo, Japan) and were maintained on rodent chow (Ca 1.45%, P 1.16%, vitamin D3 1.97 IU/g of diet). Mice were housed three to a cage in a temperature-regulated animal room (22 ± 2°C) with a 12-hour light/dark cycle and free access to food and water. In experiments involving leptin or FGF-23 administration, 10-week-old male ob/ob, db/db, and lean control mice fed ad libitum received intraperitoneal (i.p.) injections of either the vehicle alone (PBS), recombinant murine leptin (4 mg/kg; PeproTech, Rocky Hill, NJ, USA), or recombinant human FGF-23 (5 µg; ProSpec, Rehovot, Israel) at 12-hour intervals for 2 days. Serum concentrations of calcium and phosphorus were measured by colorimetry using commercial kits (total calcium: Wako Calcium E, MXB method; inorganic phosphate: Wako Phosphor C, p-methylaminophenol reduction method) manufactured by Wako Chemical Co. (Osaka, Japan). Serum 1,25(OH)2D3 levels were measured with a radioimmunoassay kit (ImmunoDiagnostic System, Boldon, UK), whereas serum 25-hydroxyvitamin D3 [25(OH)D3] levels were measured by the competitive protein-binding assay method. The serum level of mouse parathyroid hormone (PTH) was determined with an ELISA kit for mouse intact PTH (Immunotopics, San Clemente, CA, USA), whereas the FGF-23 concentration was determined by using an FGF-23 ELISA kit (Kainos Laboratories, Inc., Tokyo, Japan).

Osteoblast culture

Osteoblastic cells were obtained from calvarial outgrowth cultures of Sprague-Dawley (SD) rats. Briefly, calvaria were dissected aseptically from 8-week-old SD rats. Then the loosely adherent soft tissue was removed, and the calvaria were cut into 1- × 1-mm fragments, which were placed in culture dishes containing DMEM (MP Biomedicals, Inc., Irvine, CA, USA) supplemented with 10% (vol/vol) fetal bovine serum (FBS) and incubated under a humidified atmosphere of 5% CO2 in air at 37°C for 7 to 10 days. The culture medium was changed every 2 or 3 days until the outgrowth of osteoblasts from the calvarial fragments reached subconfluence. Then the cells were passaged with 0.125% trypsin (Sigma, St. Louis, MO, USA) and 0.1% EDTA (Sigma) in PBS. After the third passage, cells were used for the present experiments after reaching 100% confluence. Confluent rat osteoblasts were incubated in differentiation medium consisting of DMEM supplemented with 50 µg/mL of ascorbic acid (Wako Chemical), 1 mM β-glycerophosphate (Wako Chemical), and 2% charcoal/dextran-treated FBS. After 1 or 3 days of culture, the osteoblasts were exposed to 200 ng/mL of leptin (PeproTech) for 24 hours.

Real-time PCR analysis

Total RNA was isolated from tissue specimens (kidney, brain, liver, lung, spleen, and intestine) or osteoblasts using an RNeasy Mini Kit (Qiagen, Chatsworth, CA, USA) according to the manufacturer's instructions. To isolate total RNA from long bone specimens (femur and tibia), we first pulverized the frozen bones in a stainless steel chamber chilled with dry ice. Then the bone powder was homogenized in guanidine thiocyanate solution, after which total RNA was extracted and purified by CsCl centrifugation, as described previously.32 To remove contaminating genomic DNA, the RNA samples were treated with RNase-Free DNase I (Qiagen) at room temperature for 1 hour. Quantification of mRNA was performed using an Applied Biosystems 7500 Real-Time PCR system (Applied Biosystems, Inc., Tokyo, Japan). RNA samples were reverse transcribed to synthesize first-strand cDNA with an Omniscript RT kit (Qiagen) according to the manufacturer's instructions using oligo (dT) primers, and then real-time PCR was done with specific primers and the power SYBR Green PCR Master Mix (Applied Biosystems) according to the protocol of the manufacturer. The specific PCR primers used are listed in Table 1. All reactions were run with preincubation for 10 minutes at 95°C, followed by cycles of 15 seconds at 95°C and 1 minute at 60°C. Results were analyzed by the standard curve method. The levels of mRNA for the target genes were normalized for the concentration of β2-microglobulin (B2m) mRNA in the same sample.

Table 1. Oligonucleotides Used for Quantitative Real-Time PCR
Target geneSpecies SequenceGene bank reference
  1. Note: Forward (F) and reverse (R) primers are listed. B2m = β2-microglobulin; FGF-23= fibroblast growth factor 23; 1α-hydroxylase = 25-hydroxyvitamin D3 1α-hydroxylase; 24-hydroxylase = 25-hydroxyvitamin D 24-hydroxylase; Slc34a1 = solute carrier family 34 (sodium phosphate), member 1; Slc34a3 = solute carrier family 34 (sodium phosphate), member 3; ObRb = leptin/obese receptor B; ALP = alkaline phosphatase.

B2mMouse(F)5'-CTGACCGGCCTGTATGCTAT-3'NM_009735
  (R)5'-CCGTTCTTCAGCATTTGGAT-3' 
FGF-23Mouse(F)5'-GTGTCAGATTTCAAACTCAG-3'NM_022657
  (R)5'-GGATAGGCTCTAGCAGTG-3' 
Klotho-αMouse(F)5'-GGCTCAACTCTCCCAGTCAG-3'NM_013823
  (R)5'-CGCAAAGTAGCCACAAAGGT-3' 
1α-hydroxylaseMouse(F)5'-CCGCGGGCTATGCTGGAAC-3'NM_010009
  (R)5'-CTCTGGGCAAAGGCAAACATCTGA-3' 
24-hydroxylaseMouse(F)5'-TGGGCTCTAGCGAAGACAAT-3'NM_009996
  (R)5'-GGTACCAGGATGCCAAGATG-3' 
Slc34a1Mouse(F)5'-GGCTCCAACATTGGCACTAC-3'NM_011392
  (R)5'-ACAGTAGGATGCCCGAGATG-3' 
Slc34a3Mouse(F)5'-TACCCCCTCTTCTTGGGTTC-3'NM_080854
  (R)5'-CAGTCTCAAGACAGGCACCA-3' 
B2mRat(F)5'-CCGTGATCTTTCTGGTGCTT-3'NM_012512
  (R)5'-ATCTGAGGTGGGTGGAACTG-3' 
FGF-23Rat(F)5'-TCACATCAGAGGATGCTG-3'NM_130754
  (R)5'-TGGAAGATGCGCTTGGAG-3' 
ObRbRat(F)5'-GATGTTCCAAACCCCAAGAA-3'NM_012596
  (R)5'-CAGGCTCCAGAAGAAGAGGA-3' 
ALPRat(F)5'-GAACGTCTCCATGGTGGATT-3'NM_013059
  (R)5'-TGGGGGATGTAGTTCTGCTC-3' 

Mearsurement of renal hydroxylase activities

Activities of renal hydroxylases were determined using kidney homogenates as described previously.20 Renal cortex from mice was minced and washed in ice-cold homogenization buffer (0.19 M sucrose, 25 mM sodium succinate, 2 mM MgCl2, 1 mM EDTA, 20 mM Tris-HEPES, pH 7.4) and then homogenized in the same solution (20 mL/g of tissue) to produce a 5% homogenate. For measurement of 1α-hydroxylase activity the substrate [5 µg of 25(OH)D3] was used, and for 24-hydroxyalse activity assay, tritiated 1,25(OH)2D3 (250 pmol) was used as the substrate.

Western blotting

Leptin signaling was detected by identifying phosphorylation of signal transducer of activation of transcription 3 (STAT3). Confluent rat osteoblasts were cultured in differentiation medium for 3 days and then treated with 200 ng/mL of leptin for the indicated periods. Next, the cells were lysed in a buffer containing 0.058 M Tris-HCl, 1.7% SDS, 6% glycerol, 0.8% mercaptoethanol, and 0.002% bromophenol blue. The protein concentration was determined, and samples (10 µg of protein) were separated by 10% SDS-polacrylamide gel electrophoresis. Then the proteins were electrophoretically transferred to a membrane (Clear Blot Membrane-P, ATTO, Tokyo, Japan), which was incubated with rabbit polyclonal antibody for phospho-STAT3 (1:1000 dilution; Cell Signaling Technology, Danvers, MA, USA) or rabbit polyclonal antibody for STAT3 (1:1000 dilution; Cell Signaling Technology) for 2 hours at room temperature. Membranes were subsequently incubated for 1 hour at room temperature with a horseradish peroxidase–conjugated secondary antibody (1:5000 dilution; Amersham, Tokyo, Japan), after which antigen-antibody complexes were visualized with an ECL-Plus Western blotting detection system (Amersham).

Statistical analysis

Results are reported as the mean ± SEM. Differences between treated and untreated groups were assessed by Student's t test. Statistical analysis was performed with the Statview 4.02 software package (Abacus Concepts, Berkeley, CA, USA), and p values of less than .05 were considered to indicate statistical significance.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Leptin increases serum FGF-23 levels in ob/ob mice

Serum markers of bone metabolism were analyzed in leptin-deficient (ob/ob) mice. Mutant mice received a total of four doses of murine leptin (4 mg/kg i.p. every 12 hours for 2 days), and biochemical analyses were performed at 12 hours after the last injection. It was found that the serum levels of calcium and phosphate were significantly higher in untreated ob/ob mice than in wild-type control mice. Leptin administration decreased the serum calcium and phosphate levels in ob/ob mice to the range seen in lean control mice (Fig. 1A, B). Next, we examined the effect of leptin administration on FGF-23 and found that treatment with leptin significantly elevated the serum FGF-23 concentration in leptin-deficient ob/ob mice (Fig. 1C). We also demonstrated that leptin administration decreased the elevated serum concentration of 1,25(OH)2D3 in ob/ob mice and returned it to the range seen in lean control mice (Fig. 1D). Serum levels of 25(OH)D3 did not differ significantly among these three groups (Fig. 1E). The serum level of immunoreactive PTH was markedly elevated by leptin administration to ob/ob mice, which otherwise showed PTH concentrations similar to those of lean control mice (Fig. 1F). These results suggested that the decrease of 1,25(OH)2D3 after administration of leptin to ob/ob mice with defective leptin signaling was at least partly due to an increase in the serum FGF-23 level. Therefore, we focused on the regulation of FGF-23 synthesis by leptin.

thumbnail image

Figure 1. Effects of leptin on serum parameters in ob/ob mice. Male ob/ob mice were injected with either vehicle or murine leptin (4 mg/kg, i.p.) every 12 hours for a total of four doses. Age-matched control mice were injected with vehicle alone by the same regimen. Mice were killed at 12 hours after the last injection, and blood was collected to measure the following serum parameters: (A) serum calcium concentration, (B) serum inorganic phosphate (phosphorus) concentration, (C) serum FGF-23 concentration, (D) serum 1,25(OH)2D3 concentration, (E) serum 25(OH)D3 concentration, and (F) serum PTH concentration. Data are the mean ± SEM of five measurements. *p < .05, **p < .01, and ***p < .001 between the indicated groups.

Download figure to PowerPoint

FGF-23 mRNA is mainly expressed in bone

We assessed the expression of FGF-23 mRNA in various organs and tissues of lean control mice by quantitative real-time PCR and found that it was very high in the bones and was moderate in the spleen, whereas the mRNA level was extremely low in the kidneys and was undetectable in the brain, lungs, liver, and intestines (Fig. 2). The housekeeping gene murine B2m was expressed at an almost constant level in the organs and tissues examined, validating the present real-time PCR method for measurement of FGF-23 mRNA. These findings are also consistent with previous reports24, 33 that circulating FGF-23 is mainly derived from bone.

thumbnail image

Figure 2. Tissue distribution of FGF-23 mRNA expression in lean control mice. Total RNA was extracted from the indicated tissues, reverse transcribed, and assessed by quantitative real-time PCR. FGF-23 mRNA expression was calculated relative to the expression of B2m mRNA, and values are expressed as a relative ratio to the kidney level. Data are shown as the mean ± SEM for three measurements.

Download figure to PowerPoint

Leptin increases FGF-23 mRNA expression in bone

To investigate whether leptin regulates FGF-23 mRNA expression in the bones of leptin-deficient (ob/ob) mice, these mice were injected intraperitoneally with the vehicle or murine leptin (4 mg/kg i.p.) every 12 hours for 2 days. FGF-23 mRNA expression in bone tissue was significantly lower in ob/ob mice than in lean control mice, whereas administration of leptin to mutant mice markedly augmented the bone level of FGF-23 mRNA (Fig. 3A). In contrast, the splenic FGF-23 mRNA level was not altered by treating ob/ob mice with leptin (Fig. 3B). Next, to examine whether leptin directly regulates FGF-23 mRNA expression in bone cells, osteoblasts were isolated from rat calvaria and cultured in DMEM containing FBS. After reaching confluence, the osteoblasts were cultured for 1 day (day 1) or 3 days (day 3) and then were exposed to the vehicle or murine leptin (200 ng/mL) for 24 hours (Fig. 4). Expression of the genes for the active leptin receptor (ObRb) and an osteoblast marker enzyme (ALP) was clearly detected in the cultured osteoblasts, and the levels of ObRb and ALP expression were not altered throughout the culture period with leptin (Fig. 4A, B). We found that leptin treatment significantly enhanced FGF-23 mRNA expression on day 3 but not on day 1 (Fig. 4C). Significant elevation of FGF-23 protein secretion for 24 hours also was observed in osteoblasts on day 3 of culture with leptin (Fig. 4D). Subsequently, the leptin signaling pathway in osteoblasts was investigated. Activation of a downstream signaling protein of ObRb was detected by STAT3 phosphorylation, which was markedly increased after 10 minutes of leptin treatment at a concentration of 200 ng/mL. These results indicated that leptin directly stimulates FGF-23 mRNA expression by bone cells.

thumbnail image

Figure 3. Effect of leptin administration on FGF-23 mRNA expression in the bones and spleen of ob/ob mice. The experimental protocol was the same as for Fig. 1. Total RNA was extracted from bone and for real-time PCR. (A) Bone (femur and tibia) and (B) spleen. Data are presented as the mean ± SEM for five measurements. ***p < .001 between the indicated groups.

Download figure to PowerPoint

thumbnail image

Figure 4. Effect of leptin on FGF-23 expression and leptin signaling in osteoblasts. Primary cultured osteoblasts were obtained from 8-week-old rat calvaria. After reaching confluence, the cells were cultured for 1 day (day 1) or three days (day 3) and then were treated with vehicle (Veh) or murine leptin (Lep; 200 ng/mL) for 24 hours. Levels of (A) ObRb mRNA, (B) alkaline phosphatase (ALP) mRNA, and (C) FGF-23 mRNA. (D) FGF-23 level. Cells (day 3) were treated with vehicle (Veh) or murine leptin (Lep; 200 ng/mL) for 24 hours. Conditioned medium was collected, and the FGF-23 concentration was determined by ELISA. Data are presented as the mean ± SEM of three measurements. ***p < .001 between the indicated groups. (E) STAT3 posphorylation. Cells were incubated with leptin (200 ng/mL) for the indicated periods, and cell lysates were subjected to immunoblotting with anti-phospho-STAT3 and STAT3 antibodies.

Download figure to PowerPoint

Leptin regulates renal mRNA expression in ob/ob mice

FGF-23 principally functions as a phosphaturic factor by activating a receptor in the kidneys that is a complex of FGFR, Klotho, and heparin. We examined expression of the gene for Klotho, a cofactor of FGFR, in lean control mice and ob/ob mice injected with vehicle or leptin (4 mg/kg i.p.) every 12 hours for 2 days. As a result, Klotho mRNA expression did not differ among the three groups of mice (Fig. 5A). Because FGF-23 acts on the type IIa (NaPi-IIa) and type 2c (NaPi-IIc) sodium-phosphate cotransporters that mediate phosphate reabsorption by proximal tubular cells,34 renal expression of the genes for NaPi-IIa (Slc34a1) and NaPi-IIc (Slc34a3) also was assessed in lean control mice and vehicle- or leptin-treated ob/ob mice. Although Slc34a1 mRNA expression did not differ between lean control and ob/ob mice, the mRNA expression was significantly suppressed by leptin administration (Fig. 5B). Renal expression of Slc34a3 mRNA was significantly higher in ob/ob mice than in lean control mice, and leptin administration decreased the level of this mRNA in ob/ob mice (Fig. 5C). The high serum concentration of 1,25(OH)2D3 in ob/ob mice prompted us to also examine the expression of 1α-hydroxylase and 24-hydroxylase, enzymes involved in the final rate-limiting steps of vitamin D metabolism. We found that renal expression of 1α-hydroxylase mRNA was markedly enhanced in ob/ob mice, whereas administration of leptin to these mice reduced 1α-hydroxylase mRNA expression (Fig. 5D). Although ob/ob mice also showed an increase in the renal expression of 24-hydroxylase mRNA, its level was not altered by the administration of leptin (Fig. 5E). Renal 1α-hydroxylase activity was significantly higher in ob/ob mice than in lean control mice (lean versus ob/ob: 60 ± 5 versus 330 ± 123 fmol/g of protein per minute, p < .01), whereas administration of leptin to ob/ob mice decreased the level of 1α-hydroxylase activity (vehicle versus leptin: 330 ± 125 versus 87 ± 3 fmol/g of protein per minute, p < .01). Renal 24-hydroxylase activity also was increased in ob/ob mice (lean versus ob/ob: 205 ± 20 versus 520 ± 50 fmol/g of protein per minute, p < .01). These results suggested that leptin suppresses renal phosphate reabsorption and 1,25(OH)2D3 synthesis in ob/ob mice at least partly by increasing FGF-23 expression.

thumbnail image

Figure 5. Effect of leptin on renal mRNA expression in ob/ob mice. The experimental protocol was the same as for Fig. 1. Total RNA was extracted from kidneys for real-time PCR. (A) Klotho mRNA, (B) Slc34a1 (NaPi-IIa) mRNA, (C) Slc34a3 (NaPi-IIc) mRNA, (D) 1α-hydroxylase mRNA, and (E) 24-hydroxylase mRNA concentrations. Data are presented as the mean ± SEM for five measurements. *p < .05, **p < .01, and ***p < .001 between the indicated groups.

Download figure to PowerPoint

Leptin is ineffective in leptin receptor–deficient db/db mice

We also tested whether leptin acted via the long form of the leptin receptor (ObRb) by using ObRb-deficient db/db mice. As a result, we found that leptin treatment (4 mg/kg i.p. every 12 hours for 2 days) did not alter the serum FGF-23 level of these mice (Fig. 6A). Leptin administration to db/db mice moderately increased the serum concentration of 1,25(OH)2D3 (Fig. 6B), but it did not alter renal 1α-hydroxylase mRNA expression in these leptin receptor–deficient animals (Fig. 5C). These results demonstrated that leptin stimulates FGF-23 production by bone cells via ObRb in mice.

thumbnail image

Figure 6. Effect of leptin on FGF-23 and 1,25(OH)2D3 expression in db/db mice. Male db/db mice were injected with either vehicle alone or murine leptin (4 mg/kg, i.p.) every 12 hours for a total of four doses. Mice were killed at 12 hours after the last injection, and blood was collected to measure serum parameters. Total RNA was extracted from the kidneys for real-time PCR. (A) Serum FGF-23 level, (B) serum 1,25(OH)2D3 level, and (C) renal 1α-hydroxylase mRNA expression. Data are presented as the mean ± SEM for five measurements. **p < .01 between the indicated groups.

Download figure to PowerPoint

FGF-23 mimics the effect of leptin on ob/ob mice

Finally, we examined whether FGF-23 mimics the effects of leptin in ob/ob mice. Mutant mice received a total of four doses of FGF-23 (5 µg i.p. every 12 hourd for 2 days), and serum biochemical parameters and renal mRNA expression were analyzed at 12 hours after the last injection (Fig. 7). FGF-23 administration did not alter the serum concentration of FGF-23 (Fig. 7A), 1,25(OH)2D3 (Fig. 7B), or PTH (Fig. 7C). Renal expression of 1α-hydroxylase mRNA was markedly augmented in ob/ob mice compared with lean control mice, whereas administration of FGF-23 to mutant mice significantly attenuated the expression of this mRNA (Fig. 7D). Expression of Slc34a1 mRNA in the kidneys did not differ between lean control and ob/ob mice, but FGF-23 administration significantly suppressed the mRNA expression in mutant mice (Fig. 7E). Slc34a3 mRNA expression was higher in ob/ob mice than in lean control mice, and FGF-23 administration to mutant mice decreased the level of this mRNA (Fig. 7F). These results demonstrated that an increase in FGF-23, as well as an increase in leptin, can suppress renal expression of 1α-hydroxylase mRNA in ob/ob mice.

thumbnail image

Figure 7. Effects of FGF-23 on serum parameters and renal mRNA expression in ob/ob mice. Male ob/ob mice were injected with either vehicle or murine leptin (5 µg i.p.) every 12 hours for a total of four doses. Age-matched control mice were injected with vehicle alone by the same regimen. Mice were killed at 12 hours after the last injection, and blood was collected to measure serum parameters, whereas kidneys were harvested for determination of mRNA expression: (A) serum FGF-23 concentration, (B) serum 1,25(OH)2D3 concentration, (C) serum PTH concentration, (D) 1α-hydroxylase mRNA expression, (E) Slc34a1 (NaPi-IIa) mRNA expression, and (F) Slc34a3 (NaPi-IIc) mRNA expression. Data are the mean ± SEM of five measurements. *p < .05, **p < .01, and ***p < .001 between the indicated groups.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This study of leptin-deficient (ob/ob) mice provided the first evidence that leptin acts directly on bone, where it enhances FGF-23 expression. Leptin is a major regulator of bone remodeling, but its influence on bone is complex because both positive and negative relations between the serum leptin concentration and BMD have been reported in humans.35, 36 Leptin-deficient mice also have a complex phenotype, displaying an increase in trabecular bone volume in the spine but having short femora with a thin cortex and decreased trabecular bone volume.37 The increased serum calcium level in ob/ob mice may contribute to the phenotypic changes seen in these animals, such as a reduced femoral BMD. We demonstrated that ob/ob mice had a marked increase in serum 1,25(OH)2D3 that caused elevations of both serum calcium and phosphorus, reflecting the actions of 1,25(OH)2D3 on the bones and intestines. Administration of leptin to ob/ob mice strongly suppresses the upregulation of the gene for 25(OH)D3 1α-hydroxylase expression and enzyme activity in the kidneys and thus corrects the hypercalcemia.20

Our recent study demonstrated that administration of leptin to db/db mice with mutation of the long form (ObRb) of the leptin receptor did not correct the increase in renal 1α-hydroxylase mRNA expression, supporting the involvement of signaling via ObRb in suppression of this mRNA. We used primary cultures of mouse renal tubular cells to examine whether leptin directly regulates 1α-hydroxylase mRNA expression in the proximal tubules. We found that mouse renal tubule cells did not show suppression of 1α-hydroxylase mRNA expression after exposure to leptin, suggesting that leptin regulates this mRNA via some other mechanism.21

A number of studies have shown a direct influence of leptin on the regulation of bone metabolism,17, 38, 39 although leptin also acts on the hypothalamus to inhibit bone formation by stimulating postganglionic neurons.15, 16 Clinical studies have demonstrated that the circulating leptin concentration is positively associated with the bone mass at various sites, particularly in postmenopausal women.40 In addition, leptin administration increases femoral length, total bone area, and BMD in ob/ob mice,17 and it partly counteracts the reduction in bone volume induced by ovariectomy in rats.18 We previously reported that leptin deficiency reduces the BMD and inhibits the growth of mouse femurs.20 Because we and other workers have demonstrated expression of ObRb in the bones,21, 41 bone is recognized as a target tissue for leptin. Therefore, it is likely that peripheral administration of leptin would lead to the release of osteogenic factors by bone cells. Leptin is produced and secreted by cultured osteoblasts, promoting mineralization by these cells,42, 43 and it also acts as a growth factor for the chondrocytes of skeletal growth plates.44 Furthermore, leptin inhibits osteoclast generation from peripheral blood mononuclear cells and splenic cells,45 whereas it promotes the proliferation and differentiation of osteoblasts, leading to increased mineralization in human osteoblast culture.43 These findings indicate a direct action of leptin on bone. We hypothesized that leptin may regulate bone production of systemic factors such as FGF-23, a hypophosphatemic factor that is mainly synthesized in bone. This study confirmed that FGF-23 mRNA expression was higher in the bone than in other tissues examined, including the spleen, kidneys, brain, lungs, liver, and intestines, because a low level of this mRNA was found in the spleen, and it was undetectable elsewhere. These results are consistent with other reports.24, 33

FGF-23 is a bone-derived circulating peptide that has an important role in regulating phosphate and 1,25(OH)2D3 synthesis, making it essential for the maintenance of normal bone and mineral homeostasis. An excess of FGF-23 is linked to the pathogenesis of hypophosphatemic disorders, including tumor-induced osteomalacia, X-linked hypophosphatemia (XLH), and autosomal dominant and autosomal recessive hypophosphatemic rickets.34, 35 Hypophosphatemic (Hyp) mice, an animal model of XLH, show marked elevation of the serum FGF-23 level and increased FGF-23 mRNA expression in bone.23, 47 Familial tumoral calcinosis is caused by an inactivating mutation of FGF-23 and is characterized by ectopic calcification combined with hyperphosphatemia. Galnt3-deficient mice represent an animal model of familial tumoral calcinosis, and these animals exhibit hyperphosphatemia with a low circulating level of intact FGF-23 despite increased FGF-23 expression.48 We hypothesized that the mineral ion disorders of ob/ob mice may have been caused by abnormal expression of FGF-23 mRNA in bone. This study demonstrated that administration of leptin to ob/ob mice increased the serum FGF-23 concentration and stimulated FGF-23 mRNA expression in bone.

To determine whether leptin acts directly or indirectly on bone cells (including osteoblasts) to regulate FGF-23 expression, we investigated primary cultures of calvarial mouse or rat osteoblasts. We found that mouse osteoblasts had low FGF-23 mRNA expression and were unresponsive to hormone treatment. Since several earlier studies have documented the regulation of FGF-23 expression by hormones in rat osteoblasts,24, 33 we then used osteoblasts from rat calvaria. Rat osteoblasts expressing the active form of the leptin receptor (ObRb) showed marked elevation of FGF-23 mRNA expression and protein secretion after leptin treatment. Because ObRb-deficient db/db mice did not show any change of the serum FGF-23 level after injection of leptin, the effects of this hormone on FGF-23 synthesis in the bones are likely to be mediated via ObRb expressed on bone cells. Leptin activates the Janus tyrosine kinase (JAK)–STAT and PI-3 kinase pathways via ObRb in the hypothalamus and other tissues.5 In this study we showed that leptin rapidly stimulates phosphorylation of STAT3 in osteoblasts, strongly suggesting that leptin acts directly on bone cells and transduces signaling via the JAK-STAT3 pathway to stimulate FGF-23 gene expression.

In ob/ob mice, administration of leptin markedly inhibited the renal expression of 1α-hydroxylase and was effective at normalizing hypercalcemia and hyperphosphatemia. ObRb-deficient db/db mice showed an increase of renal 1α-hydroxylase gene expression, but the administration of recombinant leptin did not alter 1α-hydroxylase mRNA expression in these mice. These observations indicate that leptin signaling via ObRb inhibits renal expression of 1α-hydroxylase mRNA. Proximal tubular 1α-hydroxylase has a major role in maintaining calcium homeostasis, whereas the epithelium of the distal convoluted tubules and glomeruli shows minimal 1α-hydroxylase expression.49 According to an immunohistochemical study,50 leptin receptors are localized in the proximal straight tubule, the loop of Henle, the distal tubule, and the collecting duct but not the proximal convoluted tubule. Thus leptin appears to act indirectly on the proximal tubules to suppress 1α-hydroxylase gene expression.

FGF-23 plays a critical role in several functions of the kidneys, including regulation of the reabsorption of phosphate and the production of 1,25(OH)2D3. The target organs for FGF-23 are defined by coexpression of the membrane form of Klotho and FGFR1c, -3c, and -4.25, 26 Klotho is expressed at high levels in the kidneys, parathyroid glands, and brain but not in the bones, spleen, liver, or intestines. The kidney is the principal physiologically defined target of FGF-23, where this circulating factor inhibits phosphate reabsoption and production of 1,25(OH)2D3.22, 51 Mutation of Klotho impairs FGFR signaling via FGFRs and results in a similar phenotype to that of FGF-23-null mice.52, 53Klotho-null mice. with ablation of the Klotho gene, have high serum levels of phosphate and 1,25(OH)2D3.54, 55 Recently, it was reported that a complex consisting of FGFRs (mainly FGFR1c), Klotho, and heparin creates a specific binding site for FGF-23 on kidney cells.25 Therefore, FGF-23 seems to act on the renal tubules via this specific receptor complex composed of Klotho, FGFR1c, and heparin. This study demonstrated that leptin does not influence Klotho gene expression in mutant mouse kidneys, suggesting that leptin does not regulate renal function through changes of Klotho gene expression.

The highest level of Klotho-FGFR1-heparin complex expression is found in the distal tubules, whereas the biologic actions of FGF-23 are related to the proximal tubules.29, 34 Variable effects of FGF-23 treatment on NaPi cotransporters have been observed in ex vivo studies of proximal tubule segments or cell lines.22, 28 It is likely that FGF23 acts indirectly on the proximal tubules, possibly through stimulation of the distal tubules to release paracrine factors that regulate proximal tubule function. Such a distal proximal tubular mechanism may exist in mice. This study showed that the leptin-induced increase of bone FGF-23 production did not alter Klotho gene expression in the kidneys of ob/ob mice. A possible explanation is that FGF-23 acts on the distal tubules to stimulate Klotho protein secretion, which, in turn, acts on the proximal tubules to regulate the expression of NaPi cotransporter and the genes for 25(OH)D3 hydroxylase.31 In fact, we clearly demonstrated that FGF-23 administration to ob/ob mice suppressed renal mRNA expression of 1α-hydroxylase, NaPi-IIa, and NaPi-IIc. Recent studies56–58 have shown that FGF-23 decreases the serum 1,25(OH)2D3 level and inhibits renal expression of 1α-hydroxylase mRNA and its activity in mice. Since the dose of FGF-23 treatment (5 µg every 12 hours for 2 days) that we used was lower and the duration was shorter than in other studies,56, 58 parameters such as serum FGF-23 and 1,25(OH)2D3 were not affected by our administration of FGF-23. Leptin suppresses renal 1,25(OH)2D3 synthesis at least partly by increasing the production of FGF-23 in the bones. Thus our results suggest that a bone-kidney axis is important for suppression of 1α-hydroxylase gene expression by leptin.

We also investigated whether a short-term increase in the serum leptin level in ob/ob mice influenced the expression of target genes for FGF-23, such as NaPi-IIa, NaPi-IIc, 1α-hydroxylase, and 24-hydroxylase, and we found that leptin clearly suppressed the mRNA expression of 1α-hydroxylase, NaPi-IIa, and NaPi-IIc. An increase in the serum FGF-23 level had a similar renal effect to that of leptin in ob/ob mice. In contrast, leptin did not affect 24-hydroxylase gene expression in the kidneys of mutant mice. Mice with ablation of the Galnt3 gene have low circulating concentrations of intact FGF-23 and exhibit different changes in gene expression in the proximal tubules.48Galnt3-deficient mice have elevated expression of NaPi-IIa but not NaPi-IIc, whereas ablation of the Galnt3 gene does not alter renal levels of 1α-hydroxylase or 24-hydroxylase mRNA. These findings suggest that the target genes have different responses to FGF-23 in mice with moderate changes in the serum FGF-23 level.

It is known that 1,25(OH)2D3 is the main in vivo stimulator of 24-hydroxylase mRNA and protein expression, which catalyzes 24,25(OH)2D3 synthesis.32 Increased expression of 24-hydroxylase thus would be expected in the kidneys of ob/ob mice as a response to high serum concentrations of 1,25(OH)2D3. Leptin receptor–deficient db/db mice exhibited a moderate increase of renal 24-hydroxylase gene expression compared with the marked increase seen in ob/ob mice,21 and administration of leptin did not affect the expression of 24-hydroxylase mRNA in ob/ob mice. These observations suggest that other signaling pathways independent of ObRb are involved in the regulation of 24-hydroxylase gene expression in these mice, so further investigation is warranted.

Leptin-deficient ob/ob mice showed reduced expression of FGF-23 mRNA in the bones compared with lean control mice, whereas the serum FGF-23 level of mutant mice was similar to that of control mice. The serum 1,25(OH)2D3 concentration was consistently elevated in ob/ob mice, and 1,25(OH)2D3 has been shown to markedly stimulate FGF-23 synthesis by bone cells.24, 33 The serum level of FGF-23 is affected by several factors besides the serum 1,25(OH)2D3 concentration. For example, increased bone synthesis of FGF-23 and the rate of its degradation and excretion play a role in regulating the serum level of this hormone in ob/ob mice. Sustained elevation of the 1,25(OH)2D3 levels would compensate for suppressed bone production of FGF-23 induced by leptin deficiency, resulting in a small but significant reduction in FGF-23 mRNA expression in the bone. Therefore, the normal serum level of FGF-23 in ob/ob mice may have been due to reduced degradation or excretion of this hormone, but further investigation is required to clarify this point. In contrast, acute administration of a high dose of leptin further stimulated FGF-23 synthesis by the bone and increased the serum FGF-23 concentration in ob/ob mice.

Leptin also affects the PTH level in ob/ob mice. Injection of leptin every 12 hours for 2 days increased the serum PTH level in mutant mice, consistent with our previous findings.20, 21 An earlier study indicated that PTH is not critical for stimulation of FGF-23 synthesis because the serum FGF-23 level actually was increased in rats after parathyroidectomy.53 Although we cannot exclude the possibility that the increase of circulating FGF-23 was due to the effect of PTH, it is unlikely that PTH stimulates FGF-23 synthesis by bone in ob/ob mice.

Because the best indicator of the vitamin D status in humans and rodents is the serum 25(OH)D3 level, we measured serum concentrations in lean control mice and ob/ob mice with or without leptin treatment. Injection of leptin into ob/ob mice did not alter the serum 1,25(OH)2D3 level similar to that of lean control mice, indicating that leptin deficiency does not influence the vitamin D status in these mice.

Hamrick and colleagues38 showed that peripheral administration of leptin increases bone formation in ob/ob mice but not in control mice. The absence of any osteogenic response in lean control mice may indicates insensitivity to leptin because these animals are already in a state of leptin sufficiency. Flier59 suggested that the greatest response to leptin occurs at very low levels. An effect of leptin on FGF-23 production by bone and on renal 1α-hydroxylase expression was detected in ob/ob mice but not in lean control mice. Thus leptin acts as a hormonal factor in animals that are leptin deficient, but when the leptin concentration is normal, exogenous leptin may no longer have any positive effect on target tissues, including the bones and kidneys.

In summary, we demonstrated that leptin acts directly on bone cells to increase FGF-23 expression in leptin-deficient ob/ob mice. Leptin and FGF-23 markedly suppressed renal 1α-hydroxylase mRNA expression in ob/ob mice, suggesting that suppression of renal 1α-hydroxylase by leptin is at least partly mediated through elevated bone expression of FGF-23 in these mice.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

All the authors state that they have no conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

This work was supported by grants from the Japan Society for the Promotion of Science, a Medical Research Grant from Chugai Pharmaceutical Co., and a grant from Academic and Educational Promotion Foundation of Fukushima.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References