Fgf23 and parathyroid hormone signaling interact in kidney and bone

Molecular and Cellular Endocrinology 436 (2016) 224e239

Contents lists available at ScienceDirect

Molecular and Cellular Endocrinology journal homepage: www.elsevier.com/locate/mce

Fgf23 and parathyroid hormone signaling interact in kidney and bone Olena Andrukhova, Carmen Streicher, Ute Zeitz, Reinhold G. Erben* Department of Biomedical Sciences, University of Veterinary Medicine, 1210, Vienna, Austria

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 March 2016 Received in revised form 26 July 2016 Accepted 26 July 2016 Available online 4 August 2016

Fibroblast growth factor-23 (FGF23) is a bone-derived hormone, suppressing renal phosphate reabsorption and vitamin D hormone synthesis in proximal tubules, and stimulating calcium reabsorption in distal tubules of the kidney. Here, we analyzed the long term sequelae of deficient Fgf23 signaling on bone and mineral metabolism in 9-month-old mice lacking both Fgf23 or Klotho and a functioning vitamin D receptor (VDR). To prevent hypocalcemia in VDR deficient mice, all mice were kept on a rescue diet enriched with calcium, phosphate, and lactose. VDR mutants were normocalcemic and normophosphatemic, and had normal tibial bone mineral density. Relative to VDR mutants, Fgf23/VDR and Klotho/VDR compound mutants were characterized by hypocalcemia, hyperphosphatemia, and very high serum parathyroid hormone (PTH). Despite ~10-fold higher serum PTH levels in compound mutants, urinary excretion of phosphate and calcium as well as osteoclast numbers in bone remained unchanged relative to VDR mutants. The increase in plasma cAMP after hPTH(1e34) injection was similar in all genotypes. However, a 5-day infusion of hPTH(1e34) via osmotic minipumps resulted in reduced phosphorylation of extracellular signal-regulated kinase 1 and 2 (ERK1/2) in bone and kidney of Fgf23/ VDR and Klotho/VDR compound mutants, relative to VDR and WT controls. Similarly, the PTH-mediated ERK1/2 phosphorylation was reduced in primary osteoblasts isolated from Fgf23 and Klotho deficient mice, but was restored by concomitant treatment with recombinant FGF23. Collectively, our data indicate that the phosphaturic, calcium-conserving, and bone resorption-stimulating actions of PTH are blunted by Fgf23 or Klotho deficiency. Hence, FGF23 may be an important modulator of PTH signaling in bone and kidney. © 2016 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NCND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Fibroblast growth factor-23 Vitamin D Parathyroid hormone Secondary hyperparathyroidism Bone

1. Introduction FGF23 is a bone-derived hormone secreted by osteoblasts and osteocytes in response to vitamin D and increased phosphate (Saito et al., 2005). FGF23 is primarily a phosphaturic hormone, suppressing apical membrane abundance of type IIa sodiumphosphate cotransporters (NaPi-2a) in proximal renal tubules (Shimada et al., 2001, 2004a, 2004c; Larsson et al., 2004) by signaling through FGF receptors and the co-receptor Klotho (Urakawa et al., 2006). Excessive circulating concentrations of intact FGF23 lead to phosphate-wasting disorders such as tumorinduced osteomalacia (TIO), autosomal dominant hypophosphatemic rickets (ADHR), autosomal recessive hypophosphatemic rickets (ARHR), or X-linked hypophosphatemic

* Corresponding author. Institute of Physiology, Pathophysiology, and Biophysics, Dept. of Biomedical Sciences, University of Veterinary Medicine, Veterinaerplatz 1, 1210, Vienna, Austria. E-mail address: [email protected] (R.G. Erben).

rickets (XLH) (The ADHR Consortium, 2000; Baroncelli et al., 2012). In addition, FGF23 suppresses renal 1a-hydroxylase expression, and, thus, production of the vitamin D hormone 1a,25dihydroxyvitamin D3 [1,25(OH)2D3] in renal proximal tubules (Shimada et al., 2001; Shimada et al., 2004a, 2004c; Larsson et al., 2004). The FGF23-mediated suppression of renal 1a-hydroxylase is a physiologically essential process, because loss of FGF23 or Klotho function in humans and mice leads to unleashed production of 1,25(OH)2D3 with subsequent hypercalcemia, hyperphosphatemia, and soft tissue calcifications (Shimada et al., 2004b; Sitara et al., 2004; Ichikawa et al., 2007). It is likely that the inadequately high circulating 1,25(OH)2D3 levels caused by loss of function of FGF23 or Klotho are not detrimental per se, but that the actual toxicity is rather primarily mediated by the vitamin D hormone-induced hyperphosphatemia and hypercalcemia. As a consequence of hypercalcemia, hyperphosphatemia, and ectopic calcifications, Fgf23
/
and Klotho
/
mice develop a premature aging-like syndrome and die early, a phenotype which can be rescued by genetically disrupting vitamin D signaling (Hesse et al.,

http://dx.doi.org/10.1016/j.mce.2016.07.035 0303-7207/© 2016 The Authors. Published by Elsevier Ireland Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-ncnd/4.0/).

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

2007; Anour et al., 2012; Streicher et al., 2012). The molecular mechanism of the phosphaturic effect of FGF23 involves the following signaling cascade (Andrukhova et al., 2012): by signaling through the FGFR1c/Klotho receptor complex, FGF23 leads to activation of extracellular signal-regulated kinase 1 and 2 (ERK1/2) and serum/glucocorticoid-regulated kinase 1 (SGK1) in proximal renal tubules which in turn leads to internalization and degradation of the NaPi-2a/NHERF-1 (Naþ/Hþ Exchanger Regulatory Factor) complex by phosphorylation of the anchoring protein NHERF-1 (Weinman et al., 2011). Decreased apical membrane abundance of NaPi-2a reduces phosphate re-uptake in proximal tubules, forming the basis for the phosphaturic effect of FGF23. Furthermore, we recently discovered that FGF23 signaling in distal renal tubules upregulates membrane expression of the epithelial calcium channel transient receptor potential vanilloid-5 (TRPV5) and of the Naþ:Cl co-transporter (NCC) by a Klotho-dependent signaling cascade involving ERK1/2, SGK1, and with-no lysine kinase-4 (WNK4) (Andrukhova et al., 2014a, 2014b). Hence, FGF23 is not only a phosphaturic, but also a calcium- and sodiumconserving hormone. This finding was recently confirmed in conditional knockout mice: mice in which distal tubular Fgf23 signaling was blocked by a specific deletion of Fgfr1 in distal renal tubules were characterized by renal calcium wasting (Han et al., 2016). Interestingly, the other major phosphaturic hormone, parathyroid hormone (PTH), also combines phosphaturic with calciumconserving functions in the kidney, and, similar to FGF23, targets NHERF-1 in proximal and TRPV5 in distal tubules. PTH signals through the G-protein coupled PTH receptor 1 (PTHR1), resulting in increased intracellular cAMP with subsequent activation of PKA, PKC, and ERK1/2 pathways. In proximal renal tubules, PTH increases phosphate excretion by reducing apical membrane abundance of NaPi-2a and NaPi-2c by PKA- and PKC-mediated phosphorylation of NHERF-1 (Cole, 1999; Lederer et al., 2000; Bacic et al., 2003; Weinman et al., 2007). In distal renal tubules, PTH increases TRPV5-mediated calcium reabsorption by PKA-mediated phosphorylation and activation of TRPV5, and by probably ERK1/2mediated increase in Trpv5 transcription (Andrukhova et al., 2012; de Groot et al., 2009). In bone, increased endogenous PTH secretion stimulates osteoclastic bone resorption by upregulating the expression of RANKL (receptor activator of nuclear factor kB ligand) in osteoblastic cells through a PKA-, PKC-, and ERK1/2-mediated pathway. The similarities between FGF23 and PTH signaling led us to hypothesize that both signaling pathways might interact in the kidney. To test this hypothesis, we crossed Fgf23 and Klotho deficient mice with mice expressing a nonfunctioning VDR (VDRD/D), and analyzed mineral homeostasis in 9-month-old Fgf23
/
/VDRD/ D and Klotho
/
/VDRD/D compound mutants. We found that both Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutants were characterized by severe secondary hyperparathyroidism (sHPT) together with partial renal and skeletal resistance to PTH. Further in vivo and in vitro experiments revealed that FGF23 has a permissive role for the actions of PTH in bone and kidney, i.e., the biological effects of PTH were partially blunted in the absence of Fgf23 signaling.

225

double heterozygous animals. The VDR, Fgf23 and Klotho mutant mice used for the matings had been backcrossed to C57BL/6 genetic background for 10 generations. The double heterozygous offspring from these matings were intercrossed to generate wild-type, Fgf23
/
, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D mice. Genotyping of the mice was performed by multiplex PCR using genomic DNA extracted from tail as described (Hesse et al., 2007; Anour et al., 2012). The mice were kept at 24  C with a 12 h/12 h light/dark cycle, and were allowed free access to a normal rodent chow or the rescue diet and tap water. The rescue diet (Ssniff, Soest, Germany) containing 2.0% calcium, 1.25% phosphorus, 20% lactose and 600 IU vitamin D/kg was fed starting from 16 days of age. This diet has been shown to normalize mineral homeostasis in VDRablated mice (Li et al., 1998; Erben et al., 2002; Zeitz et al., 2003). All in vivo experiments were performed on 9e or 3-month-old male offspring of double heterozygous x double heterozygous matings. All animals were subcutaneously injected with calcein (SigmaAldrich, 20 mg/kg) on days 6 and 4 prior to necropsy. Urine was collected in metabolic cages before necropsy. At necropsy, 100 ml blood was collected from the retroorbital venous plexus into heparinized capillaries under anesthesia with ketamine/xylazine (67/ 7 mg/kg i.p.) for the measurement of ionized blood calcium. Immediately thereafter, the mice were exsanguinated from the abdominal V. cava for serum collection. In addition, 4-week-old male WT, Fgf23
/
and Klotho
/
mutants were used for experiments with kidney slices and isolation of primary osteoblasts. All animal procedures were approved by the Ethical Committees of the University of Veterinary Medicine Vienna and the local government authorities. 2.2. Serum and urine biochemistry Blood ionized calcium was measured with an AVL 9140 electrolyte analyzer (Roche Diagnostics). Serum and urinary calcium, phosphorus, and creatinine were analyzed on a Cobas c111 analyzer (Roche). Renal tubular reabsorption of calcium (TRCa) was calculated according to the formula %TRCa ¼ [1-(UrCa  SeCrea)/ (SeCa  UrCrea)]*100 (Ur, urinary; Se, serum; P, phosphorus; Crea, creatinine; Ca, calcium). Serum intact PTH was determined by ELISA (Immutopics). The intra-assay coefficient of variability (CV) for the PTH ELISA was 6.8%. cAMP was measured by radioimmunoassay (Biocompare). Serum Fgf23 was determined by ELISA (Kainos). Intra-assay CV for the Fgf23 ELISA was 7.9%. Total urinary excretion of the collagen crosslink deoxypyridinoline was measured by ELISA after acid hydrolysis (MicroVue DPD EIA kit, Quidel), and expressed per urinary creatinine concentration. 2.3. PTH-induced plasma and urinary cAMP increase Evaluation of the PTH response by the measurement of plasma and urinary cAMP levels was performed as described (Turan et al., 2014). Mice were injected subcutaneously with 40 mg/kg hPTH(1e34) (Bachem). Blood was collected into EDTA tubes at 0, 10, 20, 30, 40 and 60 min, and spontaneous urine was collected at 0, 10, 30, 60, 90 and 120 min after PTH injection for cAMP measurement. 2.4. Continuous PTH treatment

2. Materials and methods 2.1. Animals Heterozygous VDRþ/D mutant mice (Hesse et al., 2007) and heterozygous Fgf23þ/
(Sitara et al., 2004) and Klothoþ/- (Lexicon Genetics, Mutant Mouse Regional Resource Centers, University of California, Davis) (Anour et al., 2012) mice were mated to generate

Continuous PTH administration was performed in male 3month-old WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice, using Alzet minipumps (DURECT Corp., Cupertino, CA, USA) filled with hPTH (1e34) (8.1 pmol/0.25 ml per h, or 40 mg/kg per day) (Bachem, Weil am Rhein, Germany) or vehicle (equivalent volume of 10 mM acetic acid in sterile PBS, pH 7.42). According to the manufacturer's guidelines, the pumps were

226

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

implanted subcutaneously over the neck of mice under general ketamine/xylazine anesthesia (100/3 mg/kg). After implantation of the pumps, animals were individually housed, given free access to water and on a 12-h light/12-h dark cycle for 5 days. Some mice were additionally once daily subcutaneously injected with 10 mg human FGF23 R176/179Q (rFGF23, kindly provided by Amgen Inc., Thousand Oaks, CA, USA) per mouse for 5 days, and were killed 8e12 h after the last injection. At necropsy, serum, bone, and kidney samples were collected and stored at
80  C until analyzed. 2.5. Histology For histological analysis, parathyroid glands, kidneys, and bones were fixed in 4% paraformaldehyde (PFA) overnight. Paraffin embedding, sectioning at 5 mm, and haematoxylin/eosin staining was carried out according to standard procedures. Processing of bone specimens was performed as described (Erben, 1997). 2.6. Immunohistochemistry For immunohistochemistry, 5-mm-thick paraffin sections of PFAfixed kidneys were prepared. Dewaxed sections were pretreated with blocking solution, containing 5% normal goat serum in PBS with 0.1% bovine serum albumin and 0.3% Triton X-100 for 60 min. Without rinsing, sections were incubated with polyclonal rabbit anti-TRPV5 (Millipore, 1:500) antibody at 4  C overnight. After washing, sections were incubated for 1.5 h with biotinylated goat anti-rabbit secondary antibodies (Invitrogen, 1:400), followed by incubation with horseradish peroxidase-labeled streptavidin (Vector), and DAB substrate staining. Controls were performed by omitting the primary antibody. The slides were analyzed on a Zeiss Axioskop 2 microscope. 2.7. Bone mineral density measurements Bone mineral density (BMD) of the left femur was measured by peripheral quantitative computed tomography (pQCT) using a XCT Research Mþ pQCT machine (Stratec Medizintechnik). One slice (0.2-mm-thick) in the mid-diaphysis of the tibia, and 3 slices in the proximal tibial metaphysis located 0.5, 1, and 1.5 mm distal to the growth plate were measured. BMD values of the tibial metaphysis were calculated as the mean over 3 slices. A voxel size of 0.070 mm and a threshold of 600 mg/cm3 were used for calculation of cortical BMD. For the discrimination between trabecular and cortical BMD in the tibial metaphysis, a threshold of 450 mg/cm3 was used. 2.8. Bone histomorphometry Three-mm-thick midsagittal sections of methylmethacrylateembedded distal femurs were prepared using a HM 360 microtome (Microm, Walldorf, Germany), and were stained with von Kossa/McNeal and for tartrate resistant acid phosphatase (TRACP) enzyme activity as described (Erben, 1997). For fluorochromebased measurements, undeplasticized and unstained sections were mounted with Fluoromount (Serva, Heidelberg, Germany). Histomorphometric measurements in the distal femur were made using a semiautomatic system (OsteoMeasure, OsteoMetrics, Decatur, GA) and a Zeiss Axioskop microscope with a drawing attachment as described (Schneider et al., 2009). The area within 0.25 mm from the growth plate was excluded from the measurements. 2.9. Microcomputed tomography (m-CT) analysis Quantitative microcomputed tomography was performed on tibias stored in 70% ethanol as described previously (Schneider

et al., 2012), using a mCT35 mCT machine (SCANCO Medical AG, Brüttisellen, Switzerland) at a spatial resolution of 7 mm. 2.10. Total cell membrane isolation Mouse kidney cortex was homogenized in a homogenizing buffer [20 mM Tris (pH 7.4/HCl), 5 mM MgCl2, 5 mM NaH2PO4, 1 mM ethylenediaminetetraacetic acid (pH 8.0/NaOH), 80 mM sucrose, 1 mM phenyl-methylsulfonyl fluoride, 10 mg/ml leupeptin, and 10 mg/ml pepstatin] and subsequently centrifuged for 15 min at 4000 g. Supernatants were transferred to a new tube and centrifuged for an additional 30 min at 16,000 g. For Western blotting analysis, samples were dissolved in loading buffer and stored at
80  C until analyzed. 2.11. Brush border membrane vesicles (BBMV) preparation For brush border membrane vesicles (BBMV) preparation, kidney cortices were dissected in ice-cold isolation buffer immediately after being removed from animals, and then homogenized using a PottereElvehjem homogenizer at 4  C. BBMV were prepared using three consecutive magnesium precipitations (15 mM), and solubilized in Laemmli sample buffer for Western blotting. To verify BBMV purity, the activity of the BBM enzyme alkaline phosphatase and leucine aminopeptidase was regularly monitored in BBMV fractions. Protein samples for NaPi-2a Western blotting analysis were collected in lysis buffer.

2.12. Isolation of distal and proximal tubular segments Renal distal and proximal tubules were isolated as reported previously (Andrukhova et al., 2014b) from 3-month-old male WT, VDRD/D, and Fgf23
/
/VDRD/D mutant mice. In brief, murine kidneys were perfused with sterile culture medium (Ham's F12; GIBCO) containing 1 mg/ml collagenase (type II; Sigma) and 1 mg/ml pronase E (type XXV, Sigma) at pH 7.4 and 37  C. The cortical tissue was dissected in small pieces and placed at 37  C in sterile Ham's F12 medium containing 0.5 mg/ml collagenase II and 0.5 mg/ml pronase E for 15 min with vigorous shaking. After centrifugation at 3000 rpm for 4 min, the enzyme-containing solution was removed, and tubules were resuspended in ice-cold medium. Individual proximal and distal tubular segments were identified based on morphology in a dissection microscope at 25e40 magnification by their appearance and dimensions. To rule out contamination of the preparations, we performed purity and quality controls, using mRNA expression of distal (Trpv5, calbindin 28k) and proximal (NaPi-2a, NaPi-2c) tubule-specific genes (Andrukhova et al., 2012).

2.13. Kidney slice preparation and ex vivo experiments For kidney slice preparation, mice were anesthetized with ketamine/xylazine (67/7 mg/kg i.p.), and gradually perfused through the left ventricle with 50 mL of warm (30  C) sucrose/phosphate buffer (140 mM sucrose, 140 mM NaH2PO4/NaH2PO4, pH 7.4) for 5 min. Kidneys were rapidly removed, and longitudinal 200-mmthick live slices were prepared using a Leica VT1000 Vibratome (Leica Microsystems). Slices were transferred into pre-warmed serum-free medium for 10 min at 37  C. Thereafter, kidney slices were incubated at 37  C and in 5% CO2/95% air humidified atmosphere for 2 h with vehicle (PBS), 100 ng/ml of rFGF23, 100 ng/ml of hPTH (1e34), or 15 nM of PKA inhibitor P6062 (iPKA; Sigma), alone or in combination.

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

2.14. In vitro experiments with proximal and distal tubular segments In vitro experiments with dissected distal and proximal tubular segments were performed in serum-free, hormonally defined culture medium at 37  C in 5% CO2 (Andrukhova et al., 2012, 2014b). Tubular segments were incubated with rFGF23 (100 ng/ml) or 10
8 M hPTH(1e34) (Bachem) alone or in combination for 2 h. Moreover, to assess the role of ERK1/2 and SGK1, 10 ng/ml of the SGK1 inhibitor GSK 650394 (Axon Medchem) or 10 ng/ml of the ERK1/2 inhibitor PD184352 (Sigma) were used in combination with rFGF23 and hPTH treatments for 2 h. Protein samples for TRPV5 and NaPi-2a Western blotting analysis were collected in lysis buffer. 2.15. Osteoblast isolation and in vitro experiments Mouse femora were harvested, minced, and incubated with digestion medium [a-MEM medium containing 2 mg/ml Type II collagenase (Invitrogen) and 2% Penicillin-Streptomycin] at 37  C in a water bath for 4 h. Bone fragments were washed with PBS and cultured in a-MEM medium supplemented with 2% PenicillinStreptomycin and 10% calf serum (PAA) until 90% confluency. After passaging, 2  105 cells per well were treated in serum-free medium with 50e200 ng/ml of hPTH (1e34) and/or 100 ng/ml of rFGF23 for various time periods. After treatment, total cell homogenates were collected for cAMP measurements, RNA analysis, as well as for phosphorylated PKA and ERK1/2 protein analysis, and stored at
80  C.

227

repeating the centrifugation step. After the final wash, the pellets were suspended in 40 ml of electrophoresis sample buffer and boiled for 2e3 min. Western blot analysis was performed as described above by using a primary anti-TRPV5 antibody. 2.18. RNA isolation and quantitative RT-PCR Shock-frozen tissues were homogenized in TRI Reagent (Molecular Research Center) and total RNA was extracted according to the manufacturer's protocol. RNA purity and quality was determined using a 2100 Bioanalyzer (Agilent Technologies). One mg of RNA was used for first-strand cDNA synthesis (iScript cDNA Synthesis Kit, Bio-Rad). Quantitative RT-PCR was performed on a RotorGene™ 6000 (Corbett Life Science) using the QuantiFast™ EverGreen PCR Kit (Qiagen). A melting curve analysis was done for all assays. Primer sequences are available on request. Efficiencies were examined based on a standard curve. Expression of target genes was normalized to ornithine decarboxlase antizyme-1 (Oaz1) as house-keeping gene. 2.19. Statistical analyses Statistics were computed using SPSS for Windows 17.0. The data were analyzed by 1-way analysis of variance (ANOVA) followed by Student-Newman-Keuls multiple comparison test. P values of less than 0.05 were considered significant. The data are presented as the mean ± SEM. 3. Results

2.16. Western blotting Kidney cortex homogenates, distal or proximal renal tubule preparations, kidney slices, total cell membrane preparations, or BBMV samples were solubilized in Laemmli sample buffer, fractionated on SDSePAGE (30e50 mg/well) and transferred to a nitrocellulose membrane (Thermo Scientific). For SDSePAGE with tubular segment and kidney slice samples, 15 mg protein per well was used. Immunoblots were incubated overnight at 4  C with primary antibodies including rabbit anti-TRPV5 (1:500, Alpha Diagnostics, 1:1000), anti-NaPi-2a (1:1,500, generous gift of Drs. Jürg Biber and Heini Murer, University of Zurich), anti-total-ERK1/2 (1:2,000, BD Biosciences), anti-phospho-ERK1/2 (1:1,000, Cell Signaling), antitotal-PKA (1:2,000, tPKA C, Cell Signaling), anti-phospho-PKA (1:1,000, pPKA C (Thr 197), Cell Signaling), anti-PTHR1 (1:1,500, Antikoerper-online), and mouse anti-b-actin (1:5,000, Sigma) in 2% (w/v) bovine serum albumin (BSA, Sigma) in a TBS-T buffer [150 mM NaCl, 10 mM Tris (pH 7.4/HCl), 0.2% (v/v) Tween-20]. After washing, membranes were incubated with horseradish peroxidaseconjugated secondary antibodies (Amersham Life Sciences). Specific signal was visualized by ECL kit (Amersham Life Sciences). The protein bands were quantified by Image Quant 5.0 software (Molecular Dynamics). Loading of the samples was normalized to the Ponceau S stain. The expression levels were normalized to b-actin expression. Expression levels of phospho-ERK1/2 were normalized to total ERK1/2 protein expression. 2.17. Co-immunoprecipitation Kidney cortex homogenate protein samples (1 mg) were incubated with 2 mg of anti-phospho Tre/Tyr/Ser (Millipore) antibody at 4  C overnight. The immune complexes were captured by adding 50 ml Protein A or G agarose/sepharose beads (Santa Cruz Biotechnology) and overnight incubation at 4  C with gentle rocking. The immunoprecipitates were collected by centrifugation at 1000  g for 5 min at 4  C and washed 4 times in PBS, each time

3.1. Fgf23 and Klotho deficiency induce secondary hyperparathyroidism and partial PTH resistance in VDR-ablated mice To assess the long-term effects of Fgf23 and Klotho deficiency on bone and mineral metabolism in VDR-ablated mice, we analyzed mineral metabolism and the bone phenotype in 9-month-old male wild-type (WT), VDRD/D and Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice by peripheral quantitative computed tomography (pQCT), mCT, and undecalcified bone histology. The rescue diet largely protected VDR mutants against the development of sHPT. VDR mutants on rescue diet were normocalcemic, normophosphatemic, had parathyroid glands of normal size, normal serum osteocalcin, and unchanged BMD at the femoral shaft and metaphysis, relative to WT controls (Fig. 1AeC). However, compared with WT mice, VDR mutants on the rescue diet showed non-significantly elevated serum PTH, renal calcium wasting, and had increased urinary collagen crosslink excretion (Fig. 1A). Renal calcium wasting in VDR mutant mice on rescue diet is a well-known phenomenon caused by decreased expression of distal tubular calbindin D9k (Erben et al., 2002; Andrukhova et al., 2014b). In accordance with our earlier reports (Streicher et al., 2012; Andrukhova et al., 2014b), both Fgf23
/
/VDRD/D and Klotho
/-/ VDRD/D compound mutants on rescue diet were characterized by additional urinary loss of calcium relative to VDRD/D mice, hypocalcemia, and hyperphosphatemia (Fig. 1A). Circulating intact Fgf23 was elevated in 9-month-old Klotho
/-/VDRD/D, but not detectable in Fgf23
/
/VDRD/D mice (Supplementary Fig. S1A). Fgf23
/
/VDRD/ D and Klotho
/-/VDRD/D compound mutants had about 10-fold higher serum PTH than VDRD/D mice, and their parathyroid glands appeared enlarged (Fig. 1A and C). However, surprisingly, the profoundly increased serum PTH did not correct the renal calcium wasting, the hypocalcemia, or the hyperphosphatemia seen in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice (Fig. 1A). Although the very high serum PTH levels in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/

228

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

Fig. 1. Fgf23 and Klotho deficiency cause severe secondary hyperparathyroidism and partial PTH resistance in VDR-ablated mice. (A) Urinary calcium/creatinine (UrCa/Crea) excretion, percent tubular reabsorption of calcium (%TRCa), ionized blood calcium, serum intact PTH, serum phosphorus, urinary phosphate/creatinine (UrP/Crea) excretion, serum osteocalcin and urinary deoxypyridinoline/creatinine (DPD/Crea) excretion in 9-month-old male WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice on rescue diet. Each data point represents the mean ± SEM of 7e8 mice each. (B) Undecalcified sections of proximal tibiae stained with von Kossa/McNeal (top panels, original magnification 25), representative microecomputed tomography (mCT) images of the tibial midshaft (lower panels), total and cortical bone mineral density (BMD) of the tibial shaft, and total and trabecular BMD of the proximal tibial metaphysis measured by peripheral quantitative computed tomography (pQCT), as well as bone volume (BV/TV), osteoclast numbers (NOc/B.Pm), and bone formation rate (BFR/BS) measured by bone histomorphometry in the proximal tibial metaphysis in 9-month-old male WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice on rescue diet. Each data point represents the mean ± SEM of 5e6 mice each. C) Haematoxylin/eosin-stained paraffin sections of parathyroid gland of 9-month-old male WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice on rescue diet. Original magnification 100. * denotes P < 0.05 vs. WT, #P < 0.05 vs. VDRD/D mice, 1-way ANOVA followed by Student-Newman-Keuls test.

D mice reduced total and cortical BMD and increased cortical

porosity at the tibial shaft (Fig. 1B), they did neither cause cancellous bone osteopenia nor increased osteoclastic bone resorption as evidenced by unchanged or even lower urinary collagen crosslink excretion and osteoclast numbers in femoral cancellous bone, relative to VDRD/D controls (Fig. 1A and B). In addition, cancellous bone formation rate was comparable between all genotypes

(Fig. 1B). Hence, Fgf23 and Klotho deficiency appeared to convey at least partial resistance to the renal and skeletal effects of PTH. In order to better characterize the molecular basis for the renal PTH resistance in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mutants, we analyzed the renal expression of the sodium-phosphate cotransporter NaPi-2a, and the phosphorylation of the epithelial calcium channel TRPV5. The phosphaturic action of PTH is based on

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

a downregulation of the luminal membrane expression of NaPi-2a in proximal tubules (Lederer et al., 2000; Cole, 1999), whereas the calcium-conserving action of PTH is the result of increased transcription and activation of TRPV5 in renal distal tubules (de Groot et al., 2009). In agreement with the notion that lack of Fgf23 signaling blunts the phosphaturic and calcium-conserving function of PTH, we found increased NaPi-2a protein expression in brush border membrane vesicles (Fig. 2A and Supplementary Fig. S2), and reduced phosphorylation of TRPV5 in protein extracts of kidneys from Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice, relative to WT and VDR mutant controls (Fig. 2B). In contrast, mRNA abundance of TRPV5 was about 5-fold higher in kidneys from Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice than in VDRD/D controls (Fig. 2C). However, the increased transcription of TRPV5 in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice did not translate into increased TRPV5 protein expression. Rather, distal tubular TRPV5 protein expression as evidenced by immunohistochemistry and by Western blotting analysis was decreased in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice (Fig. 2D and E), in line with our earlier report that lack of Fgf23

229

signaling downregulates apical membrane abundance of TRPV5 in distal renal tubules (Andrukhova et al., 2014b). These data indicate that the increased serum PTH levels in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice did not result in a downregulation of NaPi2a or increased phosphorylation of TRPV5, consistent with a blunted renal PTH action in these mice. 3.2. Rapid PTH-induced cAMP increase in blood is unaffected by Fgf23 and Klotho deficiency in vivo We next asked the question whether the renal and skeletal PTH resistance in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice may be based on reduced PTH signaling. To address this issue, we measured the PTH-induced cAMP response in plasma and urine, and the mRNA and protein expression of PTHR1 in kidneys and bones of 3month-old WT, VDR, Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice. Subcutaneous injection of hPTH (1e34) led to an ~2-fold increase in plasma cAMP and to an ~50-fold increase in urinary cAMP/creatinine levels in all genotypes within 10 min after PTH injection

Fig. 2. Fgf23 and Klotho deficiency blunts the actions of PTH on NaPi-2a membrane expression and TRPV5 activation in kidneys of VDR-ablated mice. (A) Relative NaPi-2a protein expression in renal brush border membrane vesicles (BBMV); (B) pull-down assay performed in renal total membrane fraction samples with anti-phospho Tre/Tyr/Ser antibody followed by Western blotting analysis, using an anti-TRPV5 antibody; (C) relative fully glycosylated TRPV5 protein expression in renal total membrane fractions; (D) relative Trpv5 mRNA expression in total renal RNA; (E) immunohistochemical staining of renal TRPV5 protein expression in 9-month-old male WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice. Each data point in A-D represents the mean ± SEM of 7e8 mice each. Original magnification for immunohistochemical images 40. * denotes P < 0.05 vs. WT, #P < 0.05 vs. VDRD/D mice, 1-way ANOVA followed by Student-Newman-Keuls test.

230

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

(Fig. 3A). Renal Pthr1 mRNA abundance did not differ between the groups, and Pthr1 mRNA abundance in femora of 3-month-old Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice was actually higher than in WT and VDR control mice (Fig. 3B). However, PTHR1 protein expression in kidneys and bone was comparable in all genotypes (Fig. 3B). Moreover, a similar pattern of PTHR1 mRNA and protein expression was observed in kidneys and bones of 9-month-old Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D compound mutants (Fig. 3C). Taken together, these results suggest that a putative interaction between PTH and FGF23 signaling in kidney and bone occurs downstream of PTHR1 and adenylate cyclase. 3.3. Lack of Fgf23 and Klotho blunts the effects of continuous PTH treatment on kidney and bone To assess the biological response to continuous PTH treatment, we implanted osmotic mini-pumps continuously releasing vehicle or hPTH (1e34) in 3-month-old WT, VDR, Fgf23
/
/VDRD/D and

Klotho
/-/VDRD/D mice, and monitored plasma cAMP, calcium and phosphorus homeostasis, renal TRPV5 and NaPi-2a expression, as well as renal and skeletal ERK1/2 phosphorylation after 5 days of treatment. It is well known that continuous PTH administration leads to hypercalcemia by increased renal calcium conservation, increased 1,25(OH)2D3 production, and augmented osteoclastic bone resorption (Koh et al., 2005; Li et al., 2007a, 2007b; Tamasi et al., 2013). In WT and VDR mutant mice, the 5-day PTH treatment profoundly increased plasma cAMP concentration, induced hypercalcemia and hyperphosphaturia, increased bone resorption, increased renal and skeletal ERK1/2 phosphorylation, increased renal expression of TRPV5, and downregulated renal NaPi-2a expression (Fig. 4AeE). These results suggest that the calciumconserving and phosphaturic actions of PTH are largely VDRindependent. In contrast to acute PTH treatment, the increase in plasma cAMP was distinctly lower in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D compound mutants compared to WT and VDRD/D mutants after 5

Fig. 3. Rapid PTH-induced cAMP increase in blood is unaffected by Fgf23 and Klotho deficiency in vivo. (A) Plasma cAMP and urinary cAMP/creatinine (Crea) concentration in PTH-treated 3-month-old WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice at 0, 10, 20, 30, 40 and 60 min after s.c. injection of 40 mg/kg PTH. (B) relative Pthr1 mRNA and PTHR1 protein expression in kidneys and bones of PTH-treated WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice, 10 min after s.c. injection of 40 mg/kg PTH. Each data point in A and B represents the mean ± SEM of 4e8 mice each. (C) Relative Pthr1 mRNA and PTHR1 protein expression in kidneys and bones of 9-month-old WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutants. Each data point in C represents the mean ± SEM of 5e7 mice each. * denotes P < 0.05 vs. time point 0 min for the same genotype in A and vs. WT mice in B and C, # denotes P < 0.05 vs. VDRD/D mice, 1-way ANOVA followed by Student-Newman-Keuls test.

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

days of continuous PTH treatment (Fig. 4A). The 5-day PTH treatment induced a ~3-fold increase in serum intact Fgf23 in WT mice, whereas the PTH-induced rise in circulating Fgf23 was not significant in VDRD/D and Klotho
/-/VDRD/D mice (Supplementary Fig. S1B), suggesting that intact vitamin D signaling enhances the PTH-induced increase in circulating intact Fgf23 in vivo. In contrast to 9-month-old Klotho
/-/VDRD/D mice, circulating Fgf23 was not significantly elevated in 3-month-old Klotho
/-/VDRD/D mice, relative to VDRD/D mice (Supplementary Fig. S1A and B). The latter finding is similar to our earlier report in 4-week-old Klotho
/-/ VDRD/D mice (Murali et al., 2016), and suggests that the increase in serum intact Fgf23 in aged Klotho
/-/VDRD/D mice is mainly driven by chronically elevated PTH. Interestingly, continuous PTH treatment did not cause hypercalcemia or hyperphosphaturia in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice, and paradoxically decreased urinary collagen crosslink excretion in Fgf23
/
/VDRD/D mice (Fig. 4B and C). Furthermore, the PTH-induced upregulation of renal TRPV5 and the PTH-induced downregulation of renal NaPi-2a protein expression were almost completely blunted in these mice (Fig. 4E and Supplementary Fig. S2). Similarly, PTH did not increase ERK1/2 phosphorylation in kidneys or bones of Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice (Fig. 4D). Notably, phospho-ERK1/2 concentration was lower in bones of vehicle-treated Klotho
/-/ VDRD/D mice, relative to vehicle-treated VDRD/D mice (Fig. 4D). Collectively, continuous PTH was unable to regulate renal TRPV5 and NaPi-2a protein expression in the absence of a functioning Fgf23 signaling in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D compound mutants. Hence, Fgf23 signaling has a permissive function for the PTH-induced regulation of renal TRPV5 and NaPi-2a protein expression in vivo. 3.4. Recombinant FGF23 restores responsiveness to PTH in vivo and in renal tubular segments and osteoblasts isolated from Fgf23
/
/VDRD/D mice To test whether the responsiveness to continuous PTH can be restored in Fgf23
/
/VDRD/D mice by co-treatment with recombinant FGF23 (rFGF23), we implanted osmotic minipumps in 3month-old Fgf23
/
/VDRD/D mice, and subcutaneously injected the mice daily with either vehicle or rFGF23 (10 mg per mouse). PTH-treated VDR mutant mice were used as a positive control. As shown in Fig. 5, simultaneous treatment with rFGF23 restored the hypercalcemic and phosphaturic response to continuous PTH in Fgf23
/
/VDRD/D mice, as well as the PTH-induced increase in renal TRPV5 protein expression and renal and skeletal ERK1/2 phosphorylation. To assess whether the rFGF23-induced restoration of the response to PTH in Fgf23 deficient mice is a direct effect on renal epithelia, and to examine whether the combination of rFGF23 and PTH might have additive or over-additive effects, we isolated proximal and distal tubular segments from WT, VDRD/D, and Fgf23
/
/VDRD/D, and treated them with rFGF23 and PTH alone or in combination for 2 h. In agreement with our earlier report that FGF23 regulates distal tubular TRPV5 expression via activation of ERK1/2 and SGK1 (Andrukhova et al., 2014b), we found that both rFGF23 and PTH, alone or in combination, upregulated the protein expression of TRPV5 in distal tubular segments isolated from WT mice (Fig. 6A). Co-treatment with an SKG1 inhibitor completely blocked the rFGF23-induced, but not the PTH-induced, increase in TRPV5 expression, whereas co-treatment with an ERK1/2 inhibitor abolished both the rFGF23- and the PTH-induced upregulation of TRPV5 expression (Fig. 6A). There were no differences between distal tubular segments isolated from WT and VDR mutants, showing that the rFGF23- and PTH-induced regulation of TRPV5 is VDR independent (Fig. 6A). Similar to our in vivo findings, distal

231

tubular segments isolated from Fgf23
/
/VDRD/D mutants showed reduced expression of TRPV5 relative to those isolated from WT and VDRD/D mice, and were resistant to the PTH-induced upregulation of TRPV5 expression (Fig. 6A). This resistance could be overcome by co-treatment with rFGF23 (Fig. 6A). However, the effects of PTH and rFGF23 were not additive, suggesting that Fgf23 only has a permissive role in the PTH-induced upregulation of TRPV5 protein expression in distal tubules. In proximal tubular segments isolated from WT and VDRD/D mice, rFGF23 and PTH alone or in combination decreased NaPi-2a protein expression (Fig. 6B and Supplementary Fig. S2). Both the rFGF23- and the PTH-induced downregulation of NaPi-2a were inhibited by co-treatment with either SGK1 or ERK1/2 inhibitors. Vehicle-treated proximal tubular segments isolated from Fgf23
/
/VDRD/D mutants showed increased NaPi-2a expression, relative to those isolated from WT and VDRD/D mice (Fig. 6B). In analogy to the TRPV5 regulation in distal tubules, the PTH-induced downregulation of NaPi-2a was partially blunted in proximal tubular segments isolated from Fgf23
/
/VDRD/D mutants, and restored by co-treatment with rFGF23 (Fig. 6B). Similar to distal tubules, we did not see any additive effects of PTH and rFGF23 on the suppression of proximal tubular NaPi-2a expression. It is known that PTH induces activation of both PKA and PKC in renal epithelia (Bacic et al., 2003). To investigate the role of PKA in the PTH- and rFGF23-induced changes in renal TRPV5 and NaPi-2a protein abundance, we treated 200-mm-thick live kidney slices from WT, VDR and Fgf23
/
/VDRD/D mutants with rFGF23 and PTH alone or in combination with a PKA inhibitor. As expected, the PKA inhibitor completely blocked the PTH- but not the rFGF23-induced increase in TRPV5 protein expression (Fig. 6C). Interestingly, PKA inhibition had no effect on both the PTH- and the rFGF23-induced suppression of NaPi-2a protein expression. Therefore, it is likely that the PTH-induced PKC activation can compensate for PKA inhibition in the regulation of NaPi-2a in this experimental setting. It may be argued that tissues isolated from Fgf23
/
/VDRD/D mutants may show a diminished response to PTH ex vivo due to chronically elevated PTH in vivo. In addition, lack of vitamin D signaling may facilitate the development of PTH resistance in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice. To address these issues we treated live kidney slices isolated from 4-week-old Fgf23
/
and Klotho
/- single mutants ex vivo with PTH and rFGF23, alone or in combination. PTH is suppressed in Fgf23
/
and Klotho
/- mice due to hypercalcemia and elevated 1,25(OH)2D3 (Sitara et al., 2004; Anour et al., 2012; Murali et al., 2016). As shown in Fig. 6E and F, the kidney slices from Fgf23 and Klotho null mice did not respond to PTH treatment with an upregulation of TRPV5 and a downregulation of NaPi-2a protein expression ex vivo. Co-administration of rFGF23 rescued the PTH response in kidney slices from Fgf23
/
mice, but not in those from Klotho
/- mice (Fig. 6E and F). The latter finding is in agreement with the notion that the renal effects of FGF23 are Klotho dependent (Andrukhova et al., 2012, 2014b). Taken together, these data demonstrate that the renal PTH resistance in mice lacking Fgf23 signaling is neither caused by lack of vitamin D signaling nor by chronically increased blood levels of PTH. Our finding that tubular segments and kidney slices isolated from WT mice respond to a single treatment with PTH in the absence of additional administration of FGF23 is difficult to reconcile with the idea that Fgf23 signaling has a permissive function in the renal actions of PTH. Others also found that PTH efficiently downregulates NaPi-2a in vitro under serum-free conditions (Bacic et al., 2003). To find an explanation for these puzzling findings, we examined renal Fgf23 mRNA and protein expression. In line with earlier reports (Spichtig et al., 2014), we found expression of Fgf23 at both the mRNA and the protein level in the

232

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

233

Fig. 5. Recombinant FGF23 restores responsiveness to PTH of Fgf23¡/¡/VDRD/D mice in vivo. (A) Serum calcium, (B) serum phosphorus, (C) relative TRPV5 protein expression in renal total membrane fractions, (D) relative NaPi-2a protein expression (BBMVs), and (E) relative phospho-ERK1/2 protein abundance in kidney and bone of 3-month-old male VDRD/D and Fgf23
/
/VDRD/D compound mutant mice treated with vehicle (Veh) or recombinant FGF23 (10 mg/mouse/day, rFGF23) together with continuous infusion of vehicle or PTH via osmotic minipumps for 5 days. Each data point represents the mean ± SEM of 3e4 animals per group each. * denotes P < 0.05 vs. Veh-treated VDRD/D mice, # denotes P < 0.05 vs. Fgf23
/
/VDRD/D mutants treated with PTH infusion alone, 1-way ANOVA followed by Student-Newman-Keuls test.

kidney, albeit at low levels (Supplementary Fig. S3). Therefore, Fgf23 may have auto-/paracrine functions in the kidney, which may explain the partial PTH resistance of Fgf23 deficient tissues in vitro.

In addition, it is likely that a deficiency in Fgf23 signaling conveys partial, but not absolute resistance to the phosphaturic and calcium-conserving actions of PTH in vivo and in vitro.

Fig. 4. Lack of Fgf23 and Klotho blunts the renal and skeletal effects of continuous PTH infusion. (A) Plasma cAMP concentration, (B) urinary calcium/creatinine (UrCa/Crea) excretion, urinary phosphate/creatinine (UrP/Crea) excretion, serum calcium, serum phosphorus, (C) urinary deoxypyridinoline/creatinine (DPD/Crea) excretion, (D) relative phospho-ERK1/2 protein abundance in kidney and bone, and (E) relative renal TRPV5 (total membrane fractions) and NaPi-2a protein expression (BBMVs) in 3-month-old male WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice treated with vehicle (Veh) or continuous infusion of PTH via osmotic minipumps for 5 days. Each data point represents the mean ± SEM of 6 mice. * denotes P < 0.05 vs. Veh-treated mice of the same genotype, # denotes P < 0.05 vs. PTH-treated WT and VDRD/D mice, y denotes P < 0.05 vs. vehicle-treated VDRD/D mice, 1-way ANOVA followed by Student-Newman-Keuls test.

234

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

Fig. 6. Recombinant FGF23 restores responsiveness to PTH in isolated renal tubular segments or kidney slices of Fgf23 or Klotho deficient mice. (A) Relative TRPV5 protein expression in distal renal tubular segments, and (B) relative NaPi-2a protein expression in proximal renal tubular segments isolated from 3-month-old WT, VDRD/D, and Fgf23
/ /VDRD/D compound mutant mice, and treated with vehicle (Veh), PTH [hPTH (1e34), 100 ng/ml] and/or recombinant FGF23 (rFGF23, 100 ng/ml), alone or in combination with SGK1 (iSGK1) and ERK1/2 inhibitors (iERK1/2) for 2 h. (C) Relative TRPV5 protein expression, and (D) relative NaPi-2a protein expression in 200-mm-thick live kidney slices isolated from 3-month-old WT, VDRD/D, and Fgf23
/
/VDRD/D compound mutant mice, and treated with Veh, hPTH(1e34) (100 ng/ml), and/or rFGF23 (100 ng/ml), alone or in combination with a PKA (iPKA) inhibitor for 2 h. (E) Relative TRPV5 protein expression, and (F) relative NaPi-2a protein expression in kidney slices isolated from 4-week-old WT, Fgf23
/
and Klotho
/
mutant mice, and treated with Veh, hPTH (1e34) (100 ng/ml) and/or rFGF23 (100 ng/ml) for 2 h. Each data point represents the mean ± SEM of 2 separate experiments with 2e3 animals per group. * denotes P < 0.05 vs. Veh-treated segments or kidney slices of the same genotype by 1-way ANOVA followed by Student-Newman-Keuls test.




To examine whether rFGF23 was able to modulate the effects of PTH on primary osteoblasts in vitro and to find an explanation for the blunted skeletal response to PTH in Klotho
/-/VDRD/D mice, we analyzed cAMP concentration, phospho-PKA, and phospho-ERK1/2 abundance in primary osteoblasts isolated from femurs of WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mutant mice after treatment with PTH alone or in combination with rFGF23. Treatment with 50e200 ng/ml PTH increased intracellular cAMP concentration with a similar time course and to a similar extent in

osteoblasts of all genotypes (Fig. 7A). Ten minutes after PTH treatment, when the cAMP response was maximal, the dosedependent PTH-induced increase in phospho-PKA and phosphoERK1/2 was completely blunted in osteoblasts isolated from Fgf23
/
/VDRD/D mice, and partially blunted in osteoblasts from Klotho
/-/VDRD/D mice (Fig. 7B). Twenty-four hours after start of treatment, phospho-ERK1/2 abundance was reduced in vehicletreated osteoblasts isolated from Fgf23
/
/VDRD/D mice, relative to WT and VDR cells (Fig. 7C). A similar trend was seen in vehicle-

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

treated osteoblasts from Klotho
/-/VDRD/D mice (Fig. 7C). A 24-h PTH treatment increased phospho-ERK1/2 in WT and VDRD/D, but not in Fgf23
/
/VDRD/D and to a lesser extent in Klotho
/-/VDRD/D osteoblasts (Fig. 7C), whereas rFGF23 treatment increased phospho-ERK1/2 in all genotypes in a Klotho independent manner (Fig. 7C). Interestingly, co-treatment with PTH and rFGF23 led to an additive or even over-additive increase in phospho-ERK1/2 abundance in osteoblasts, independent of the genotype. In analogy to the changes observed in phospho-ERK1/2 abundance, PTH treatment increased mRNA expression of Rankl and suppressed mRNA expression of Opg in osteoblasts isolated from WT and VDRD/D mice, but not in those isolated from Fgf23
/
/VDRD/D and Klotho
/-/VDRD/ D mice (Fig. 7D). Co-treatment with PTH and rFGF23 led to an additive increase in Rankl mRNA expression, and to a suppression of Opg mRNA in osteoblasts of all genotypes (Fig. 7D). Similar to the results obtained in osteoblasts isolated from Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D compound mutant mice, phospho-ERK1/2 protein abundance remained unchanged in osteoblasts isolated from Fgf23
/
and Klotho
/- single mutant mice after a 24-h treatment with PTH (Fig. 7E), showing that the osteoblastic PTH resistance induced by lack of Fgf23 signaling occurs independent of the absence or presence of intact vitamin D signaling. Collectively, these data show that the effects of PTH on osteoblasts are completely or partially blunted by Fgf23 and Klotho deficiency, and that PTH and FGF23 have additive effects on the expression of bone resorption-stimulating factors in osteoblasts in vitro. 4. Discussion The current study has shown that 1) long-term Fgf23 and Klotho deficiency cause severe secondary hyperparathyroidism associated with renal and skeletal PTH resistance in VDR-ablated mice; 2) that Fgf23 signaling is dispensable for the rapid PTH-induced cAMP increase but essential for the responsiveness of kidney and bone to continuous PTH infusion in vivo; 3) that intact Fgf23 signaling is permissive to the PTH-mediated upregulation of TRPV5 expression in distal renal tubules and to the suppression of NaPi-2a membrane expression in proximal renal tubules in vitro; and 4) that FGF23 and PTH have additive effects on activation of ERK1/2 and on the regulation of RANKL and OPG in vitro in osteoblasts. It is well documented that Fgf23 suppresses PTH secretion in vivo and in vitro (Ben Dov et al., 2007; Olauson et al., 2013). Therefore, an alternative explanation for our finding of hyperparathyroidism in 9-month-old Fgf23
/
/VDRD/D mice would be that long-term lack of the suppressive effect of Fgf23 on the parathyroid may induce hyperparathyroidism. However, several sets of observations argue against this notion. First, both 9-month-old Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D compound mutants developed hyperparathyroidism in our experiments despite opposed Fgf23 blood levels, namely complete absence in Fgf23
/
/VDRD/D and elevated circulating intact Fgf23 in Klotho
/-/VDRD/D mice. Furthermore, it has been shown in mice with a parathyroid-specific deletion of Klotho that the suppressive effect of recombinant FGF23 on PTH secretion is Klotho independent (Olauson et al., 2013). Therefore, the similarities between the phenotypes of Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D mice make it unlikely that the profound upregulation in PTH secretion observed in these mice was caused by absence of Fgf23 signaling in the parathyroid gland. Second, it has recently been shown that conditional knockout mice with a specific deletion of Fgfr1 in distal tubules also develop calcium wasting and sHPT (Han et al., 2016), indicating that lack of Fgf23 signaling in the distal nephron induces chronic renal calcium wasting, which in turn drives the counter-regulatory upregulation of PTH secretion. Collectively, these data indicate that the hyperparathyroidism found in Fgf23
/
/VDRD/D and Klotho
/-/VDRD/D

235

mice is sHPT caused by renal calcium wasting and hyperphosphatemia. In contrast to the current study, we previously reported that Fgf23 is not essential for the bone anabolic and phosphaturic actions of PTH (Yuan et al., 2011). However, major differences in study design and endpoints may explain this apparent discrepancy. The study by Yuan et al. (2011) was performed in very young 8-day-old mice, and examined the bone anabolic effects of intermittent PTH, whereas the current in vivo experiments were performed in adult 3- to 9-month-old Fgf23 and Klotho deficient mice on a VDR deficient background, and focused on the catabolic effects of continuous PTH. Furthermore, Yuan et al. (2011) found a normal PTHinduced increase in phospho-ERK1/2 in calvarial osteoblasts isolated from Fgf23 and Klotho deficient mice and differentiated for 1 week, whereas in the current study the PTH-induced increase in ERK1/2 phosphorylation was blunted in undifferentiated osteoblasts isolated from femurs of adult Fgf23
/
/VDRD/D and 4-weekold Fgf23
/
and Klotho
/- mice. We hypothesize that the discrepant in vitro results may be explained by site-specific differences in PTH response, different age of the donors, or different differentiation status of the cells. The differences between the in vivo experiments can likely be explained by the different age of the mice and the different PTH administration regimens. It is very well known in this context that intermittent PTH induces bone anabolic effects by enhancing osteoblastic bone formation (Potts, 2005; Poole and Reeve, 2005; Lavi-Moshayoff et al., 2009), whereas continuous PTH is bone catabolic by stimulating bone resorption (Datta and Abou-Samra, 2009; Poole and Reeve, 2005). Therefore, in contrast to continuous PTH, the skeletal effects of intermittent PTH may be independent of intact Fgf23 signaling. One of the key questions in the current study is the molecular mechanism of the interaction between PTH and Fgf23 signaling in kidney and bone. A model of this interaction is schematically shown in Fig. 8. FGF23 induces phosphorylation of NHERF-1 by a signaling pathway involving ERK1/2 and SGK1 (Andrukhova et al., 2012), whereas PTH signaling leads to phosphorylation of NHERF1 by activation of PKA and PKC (Bacic et al., 2003). Phosphorylation of NHERF-1 leads to dissociation, internalization, and degradation of the NaPi-2a/NHERF-1 complex. It is unclear at present why the ability of PTH to downregulate renal NaPi-2a was reduced in the absence of Fgf23 signaling in vivo and in vitro in our study. It is conceivable that FGF23-mediated activation of SGK1 is essential for the PTH-induced internalization of NaPi-2a. SGK1 has been reported to be involved in internalization of membrane proteins in the kidney (Satoh et al., 2015). As a matter of fact, this notion is supported by our finding that SGK1 inhibition blocked the PTHmediated suppression of NaPi-2a in proximal tubular segments. Another possible explanation may be that Fgf23 signaling results in phosphorylation of specific sites in the NHERF-1 molecule that are necessary for dissociation of the NaPi-2a/NHERF-1 complex, but cannot be phosphorylated by PTH-mediated activation of PKA and PKC alone. In this context, it has been shown that PTH-mediated downstream activation of protein kinases induces cooperative phosphorylation of NHERF-1 at serine-77 and threonine-95 (Weinman et al., 2010). FGF23 signaling also targets serine-77 in the NHERF-1 molecule (Weinman et al., 2011; Andrukhova et al., 2012). It is currently not known whether FGF23 signaling targets other phosphorylation sites in NHERF-1. Nevertheless, it is conceivable that cooperative NHERF-1 phosphorylation forms the basis for the observed permissive effect of intact Fgf23 signaling on the phosphaturic actions of PTH. A mutual interaction between PTH and FGF23 signaling in the regulation of renal phosphate reabsorption is also supported by evidence from clinical studies, showing that the phosphaturic action of FGF23 is diminished in patients with hypoparathyroidism (Gupta et al., 2004; Geller et al.,

236

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

237

Fig. 8. Proposed model of FGF23 and PTH signaling interaction in kidney and bone. (A) In proximal renal tubules, blood-borne FGF23 binds to the FGFR1c/Klotho receptor complex, leading to activation of ERK1/2 and SGK1. SGK1 in turn phosphorylates the anchoring protein NHERF-1, which leads to internalization and degradation of the NaPi-2a/ NHERF-1 complex. PTH signals through the G-protein coupled PTH receptor-1 (PTHR), resulting in activation of PKA, PKC, and ERK1/2 pathways which also lead to NHERF-1 phosphorylation. PTH receptor-1 is also expressed in the apical cell membrane in proximal renal tubules, where it activates phospholipase C (PLC) and PKC upon ligand binding, which subsequently induce NHERF-1 phosphorylation (Capuano et al., 2007). It is likely that FGF23 and PTH signaling interact in proximal tubular epithelium because both signaling pathways target NHERF-1. Molecular details of this interaction are not known. In distal renal tubules, FGF23 activates WNK1/4 signaling. WNK signaling controls membrane trafficking of TRPV5. PTH signaling leads to phosphorylation and activation of TRPV5 by a PKA-mediated process. The permissive role of FGF23 signaling in the PTHinduced regulation of TRPV5 activity and cellular calcium entry may be based on the regulation of the membrane abundance of TRPV5 by FGF23. (B) In osteoblasts, FGF23 and PTH signaling pathways both induce ERK1/2 phosphorylation, resulting in additive effects in the regulation of RANKL and OPG expression.

2007; Bhadada et al., 2013). In distal renal tubules, PTH-activated cAMP-PKA signaling upregulates TRPV5 transcription, and increases the open probability of the TRPV5 channel by PKA-mediated phosphorylation of threonine-709 in TRPV5 (de Groot et al., 2009). In contrast, Fgf23 is an important regulator of membrane transport of TRPV5 via activation of WNK4 (Andrukhova et al., 2014b). Therefore, it is conceivable that Fgf23 deficiency interferes with the membrane transport of TRPV5 (Fig. 8), overruling the transcriptional effects of PTH. In line with this idea, it has recently been shown that knockdown of WNK4 attenuated the PTH-mediated activation of TRPV5 in murine DCT15 cells (Hoover et al., 2016). In addition, WNK4-mediated activation of NCC may also be involved in the PTH-induced upregulation of distal tubular calcium re-uptake (Hoover et al., 2016). Because FGF23 signaling stimulates WNK4 phosphorylation and increases NCC membrane expression (Andrukhova et al., 2014a), this model would explain why intact Fgf23 signaling has a permissive function for PTH-mediated upregulation of TRPV5 membrane expression and activation in distal renal tubules. Klotho expression is low in bone cells (Kuro-o et al., 1997; Miyagawa et al., 2014). In addition, we recently showed that

FGF23 suppresses transcription of alkaline phosphatase by activation of ERK1/2 via a Klotho-independent pathway in osteoblasts (Murali et al., 2016). In agreement with this finding, rFGF23 induced phosphorylation of ERK1/2 in Klotho deficient osteoblasts in the present study, similar to WT osteoblasts. Therefore, it is surprising that the bone resorption-stimulating effects of endogenous and externally administered PTH were blunted in Klotho
/-/VDRD/D mice, similar to Fgf23
/
/VDRD/D mutants. We hypothesize that the explanation for this puzzling finding is the reduced basal ERK1/2 activation found in bones (Fig. 4D) and partially in osteoblasts (Fig. 7C) isolated from Klotho
/-/VDRD/D mice. Although experimental evidence for this hypothesis is currently lacking, it is conceivable that lack of the transmembrane co-receptor Klotho reduces the basal activity of FGFR signaling and subsequently phospho-ERK1/2 concentration in osteoblasts. Transmembrane Klotho, even at low expression levels, may interact with FGFRs to regulate their basal activity in a ligand independent fashion. Because the PTH-mediated increase in ERK1/2 activation is probably a key event in the PTH-induced upregulation of RANKL secretion, basal activity of FGFR signaling may modulate the response to PTH in osteoblasts. A clinically relevant question is whether increased circulating

Fig. 7. FGF23 and PTH have additive effects on ERK1/2 activation and Rankl mRNA expression in osteoblasts in vitro. (A) cAMP concentration in total cell homogenates of cultures of primary osteoblasts isolated from femora of WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice, and treated with PTH [hPTH(1e34), 50, 100 and 200 ng/ml] for 0, 10, 20, and 30 min, and 1, 2, 6, and 24 h. (B) Relative phospho-PKA protein abundance, and relative phospho-ERK1/2 protein abundance in primary osteoblasts isolated from femora of WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice, and treated with vehicle (Veh) or 50, 100 and 200 ng/ml hPTH(1e34) for 10 min * in A and B denotes P < 0.05 vs. 0 min post-PTH treatment. (C) Phospho-ERK1/2 protein abundance, and (D) relative Rankl and Opg mRNA expression in primary osteoblasts isolated from femora of WT, VDRD/D, Fgf23
/
/VDRD/D and Klotho
/
/VDRD/D compound mutant mice, and treated with Veh, 100 ng/ml hPTH(1e34), and/or recombinant FGF23 (rFGF23, 100 ng/ml) for 24 h (E) phospho-ERK1/2 protein abundance in primary osteoblasts isolated from femora of WT, Fgf23
/
and Klotho
/
mice, and treated with Veh, 100 ng/ ml hPTH(1e34), and/or 100 ng/ml rFGF23 for 24 h. Each data point represents the mean ± SEM of 2 experiments with 3 animals per group each. * denotes P < 0.05 vs. Veh-treated cells, # denotes P < 0.05 vs. PTH-treated cells, y denotes P < 0.05 vs. rFGF23-treated cells of the same genotype, k denotes P < 0.05 vs. WT and VDRD/D cells in C e E by 1-way ANOVA followed by Student-Newman-Keuls test.

238

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239

levels of FGF23 might augment the actions of PTH in target tissues. In patients with chronic kidney disease (CKD), serum FGF23 concentrations rise in parallel with the decline in glomerular filtration rate during CKD progression (Gutierrez, 2010; Juppner et al., 2010). In addition, driven by hyperphosphatemia and reduced production of 1,25(OH)2D3, sHPT is commonly associated with CKD (LaviMoshayoff et al., 2010). Therefore, FGF23 and PTH are both elevated in CKD patients, and since both hormones are negatively associated with clinical outcome, a FGF23-mediated enhancement of PTH action would be of major pathophysiological importance. However, our experiments did not reveal any evidence for additive or synergistic effects in the kidney. However, PTH and FGF23 had additive effects on ERK1/2 phosphorylation and RANKL/OPG expression in osteoblasts isolated from femurs of WT and VDR mutant mice. Therefore, based on our in vitro data, it appears possible that PTH and FGF23 positively interact in bone. The question whether gain of FGF23 function sensitizes target tissues to the effects of PTH in vivo needs to be addressed in future experiments. In conclusion, we found that deficient Fgf23 signaling blunts the phosphaturic and calcium reabsorption-stimulating effects of PTH in the kidney and the resorption-stimulating effects of PTH in bone. It is tempting to speculate that this permissive role of FGF23 for PTH signaling in bone and kidney might also be the deeper biological reason why PTH stimulates FGF23 secretion in bone. Disclosure The authors declare no conflicts of interest. Acknowledgements The authors thank Claudia Bergow and Christiane Schüler for excellent technical assistance. This work was supported by a grant from the Austrian Science Fund (FWF 24186-B21) to R.G.E. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.mce.2016.07.035. References Andrukhova, O., Slavic, S., Smorodchenko, A., Zeitz, U., Shalhoub, V., Lanske, B., Pohl, E.E., Erben, R.G., 2014a. FGF23 regulates renal sodium handling and blood pressure. EMBO Mol. Med. 6, 744e759. Andrukhova, O., Smorodchenko, A., Egerbacher, M., Streicher, C., Zeitz, U., Goetz, R., Shalhoub, V., Mohammadi, M., Pohl, E.E., Lanske, B., Erben, R.G., 2014b. FGF23 promotes renal calcium reabsorption through the TRPV5 channel. EMBO J. 33, 229e246. Andrukhova, O., Zeitz, U., Goetz, R., Mohammadi, M., Lanske, B., Erben, R.G., 2012. FGF23 acts directly on renal proximal tubules to induce phosphaturia through activation of the ERK1/2-SGK1 signaling pathway. Bone 51, 621e628. Anour, R., Andrukhova, O., Ritter, E., Zeitz, U., Erben, R.G., 2012. Klotho lacks a vitamin D independent physiological role in glucose homeostasis, bone turnover, and steady-state PTH secretion in vivo. PLoS. One 7, e31376. Bacic, D., Schulz, N., Biber, J., Kaissling, B., Murer, H., Wagner, C.A., 2003. Involvement of the MAPK-kinase pathway in the PTH-mediated regulation of the proximal tubule type IIa Naþ/Pi cotransporter in mouse kidney. Pflugers Arch. 446, 52e60. Baroncelli, G.I., Toschi, B., Bertelloni, S., 2012. Hypophosphatemic rickets. Curr. Opin. Endocrinol. Diabetes Obes. 19, 460e467. Ben Dov, I.Z., Galitzer, H., Lavi-Moshayoff, V., Goetz, R., Kuro-o, M., Mohammadi, M., Sirkis, R., Naveh-Many, T., Silver, J., 2007. The parathyroid is a target organ for FGF23 in rats. J. Clin. Investig. 117, 4003e4008. Bhadada, S.K., Palnitkar, S., Qiu, S., Parikh, N., Talpos, G.B., Rao, S.D., 2013. Deliberate total parathyroidectomy: a potentially novel therapy for tumor-induced hypophosphatemic osteomalacia. J. Clin. Endocrinol. Metab. 98, 4273e4278. Capuano, P., Bacic, D., Roos, M., Gisler, S.M., Stange, G., Biber, J., Kaissling, B., Weinman, E.J., Shenolikar, S., Wagner, C.A., Murer, H., 2007. Defective coupling of apical PTH receptors to phospholipase C prevents internalization of the Naþ-

phosphate cotransporter NaPi-IIa in Nherf1-deficient mice. Am. J. Physiol. Cell Physiol. 292, C927eC934. Cole, J.A., 1999. Parathyroid hormone activates mitogen-activated protein kinase in opossum kidney cells. Endocrinology 140, 5771e5779. Datta, N.S., Abou-Samra, A.B., 2009. PTH and PTHrP signaling in osteoblasts. Cell Signal 21, 1245e1254. de Groot, T., Lee, K., Langeslag, M., Xi, Q., Jalink, K., Bindels, R.J., Hoenderop, J.G., 2009. Parathyroid hormone activates TRPV5 via PKA-dependent phosphorylation. J. Am. Soc. Nephrol. 20, 1693e1704. Erben, R.G., 1997. Embedding of bone samples in methylmethacrylate: an improved method suitable for bone histomorphometry, histochemistry, and immunohistochemistry. J. Histochem. Cytochem 45, 307e313. Erben, R.G., Soegiarto, D.W., Weber, K., Zeitz, U., Lieberherr, M., Gniadecki, R., € ller, G., Adamski, J., Balling, R., 2002. Deletion of deoxyribonucleic acid Mo binding domain of the vitamin D receptor abrogates genomic and nongenomic functions of vitamin D. Mol. Endocrinol. 16, 1524e1537. Geller, J.L., Khosravi, A., Kelly, M.H., Riminucci, M., Adams, J.S., Collins, M.T., 2007. Cinacalcet in the management of tumor-induced osteomalacia. J. Bone Min. Res. 22, 931e937. Gupta, A., Winer, K., Econs, M.J., Marx, S.J., Collins, M.T., 2004. FGF-23 is elevated by chronic hyperphosphatemia. J. Clin. Endocrinol. Metab. 89, 4489e4492. Gutierrez, O.M., 2010. Fibroblast growth factor 23 and disordered vitamin D metabolism in chronic kidney disease: updating the “trade-off” hypothesis. Clin. J. Am. Soc. Nephrol. 5, 1710e1716. Han, X., Yang, J., Li, L., Huang, J., King, G., Quarles, L.D., 2016. Conditional deletion of Fgfr1 in the proximal and distal tubule identifies distinct roles in phosphate and calcium transport. PLoS. One 11, e0147845. Hesse, M., Frohlich, L.F., Zeitz, U., Lanske, B., Erben, R.G., 2007. Ablation of vitamin D signaling rescues bone, mineral, and glucose homeostasis in Fgf-23 deficient mice. Matrix Biol. 26, 75e84. Hoover, R.S., Tomilin, V., Hanson, L., Pochynyuk, O., Ko, B., 2016. PTH modulation of NCC activity regulates TRPV5 Ca2þ reabsorption. Am. J. Physiol. Ren. Physiol. 310, F144eF151. Ichikawa, S., Imel, E.A., Kreiter, M.L., Yu, X., Mackenzie, D.S., Sorenson, A.H., Goetz, R., Mohammadi, M., White, K.E., Econs, M.J., 2007. A homozygous missense mutation in human KLOTHO causes severe tumoral calcinosis. J. Clin. Investig. 117, 2684e2691. Juppner, H., Wolf, M., Salusky, I.B., 2010. FGF-23: more than a regulator of renal phosphate handling? J. Bone Min. Res. 25, 2091e2097. Koh, A.J., Demiralp, B., Neiva, K.G., Hooten, J., Nohutcu, R.M., Shim, H., Datta, N.S., Taichman, R.S., McCauley, L.K., 2005. Cells of the osteoclast lineage as mediators of the anabolic actions of parathyroid hormone in bone. Endocrinology 146, 4584e4596. Kuro-o, M., Matsumura, Y., Aizawa, H., Kawaguchi, H., Suga, T., Utsugi, T., Ohyama, Y., Kurabayashi, M., Kaname, T., Kume, E., Iwasaki, H., Iida, A., Shiraki-Iida, T., Nishikawa, S., Nagai, R., Nabeshima, Y.I., 1997. Mutation of the mouse klotho gene leads to a syndrome resembling ageing. Nature 390, 45e51. Larsson, T., Marsell, R., Schipani, E., Ohlsson, C., Ljunggren, O., Tenenhouse, H.S., Juppner, H., Jonsson, K.B., 2004. Transgenic mice expressing fibroblast growth factor 23 under the control of the alpha1(I) collagen promoter exhibit growth retardation, osteomalacia, and disturbed phosphate homeostasis. Endocrinology 145, 3087e3094. Lavi-Moshayoff, V., Silver, J., Naveh-Many, T., 2009. Human PTH gene regulation in vivo using transgenic mice. Am. J. Physiol. Ren. Physiol. 297, F713eF719. Lavi-Moshayoff, V., Wasserman, G., Meir, T., Silver, J., Naveh-Many, T., 2010. PTH increases FGF23 gene expression and mediates the high-FGF23 levels of experimental kidney failure: a bone parathyroid feedback loop. Am. J. Physiol. Ren. Physiol. 299, F882eF889. Lederer, E.D., Sohi, S.S., McLeish, K.R., 2000. Parathyroid hormone stimulates extracellular signal-regulated kinase (ERK) activity through two independent signal transduction pathways: role of ERK in sodium-phosphate cotransport. J. Am. Soc. Nephrol. 11, 222e231. Li, X., Liu, H., Qin, L., Tamasi, J., Bergenstock, M., Shapses, S., Feyen, J.H., Notterman, D.A., Partridge, N.C., 2007a. Determination of dual effects of parathyroid hormone on skeletal gene expression in vivo by microarray and network analysis. J. Biol. Chem. 282, 33086e33097. Li, X., Qin, L., Bergenstock, M., Bevelock, L.M., Novack, D.V., Partridge, N.C., 2007b. Parathyroid hormone stimulates osteoblastic expression of MCP-1 to recruit and increase the fusion of pre/osteoclasts. J. Biol. Chem. 282, 33098e33106. Li, Y.C., Amling, M., Pirro, A.E., Priemel, M., Meuse, J., Baron, R., Delling, G., Demay, M.B., 1998. Normalization of mineral ion homeostasis by dietary means prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia in vitamin D receptor-ablated mice. Endocrinology 139, 4391e4396. Miyagawa, K., Yamazaki, M., Kawai, M., Nishino, J., Koshimizu, T., Ohata, Y., Tachikawa, K., Mikuni-Takagaki, Y., Kogo, M., Ozono, K., Michigami, T., 2014. Dysregulated gene expression in the primary osteoblasts and osteocytes isolated from hypophosphatemic Hyp mice. PLoS. One 9, e93840. Murali, S.K., Roschger, P., Zeitz, U., Klaushofer, K., Andrukhova, O., Erben, R.G., 2016. FGF23 regulates bone mineralization in a 1,25(OH) D and Klotho-independent manner. J. Bone Min. Res. 31, 129e142. Olauson, H., Lindberg, K., Amin, R., Sato, T., Jia, T., Goetz, R., Mohammadi, M., Andersson, G., Lanske, B., Larsson, T.E., 2013. Parathyroid-specific deletion of Klotho unravels a novel calcineurin-dependent FGF23 signaling pathway that regulates PTH secretion. PLoS Genet. 9, e1003975. Poole, K.E., Reeve, J., 2005. Parathyroid hormone - a bone anabolic and catabolic

O. Andrukhova et al. / Molecular and Cellular Endocrinology 436 (2016) 224e239 agent. Curr. Opin. Pharmacol. 5, 612e617. Potts, J.T., 2005. Parathyroid hormone: past and present. J. Endocrinol. 187, 311e325. Saito, H., Maeda, A., Ohtomo, S., Hirata, M., Kusano, K., Kato, S., Ogata, E., Segawa, H., Miyamoto, K., Fukushima, N., 2005. Circulating FGF-23 is regulated by 1alpha,25-dihydroxyvitamin D3 and phosphorus in vivo. J. Biol. Chem. 280, 2543e2549. Satoh, N., Nakamura, M., Suzuki, M., Suzuki, A., Seki, G., Horita, S., 2015. Roles of Akt and SGK1 in the regulation of renal tubular transport. Biomed. Res. Int. 2015, 971697. Schneider, M.R., Dahlhoff, M., Andrukhova, O., Grill, J., Glosmann, M., Schuler, C., Weber, K., Wolf, E., Erben, R.G., 2012. Normal epidermal growth factor receptor signaling is dispensable for bone anabolic effects of parathyroid hormone. Bone 50, 237e244. Schneider, M.R., Mayer-Roenne, B., Dahlhoff, M., Proell, V., Weber, K., Wolf, E., Erben, R.G., 2009. High cortical bone mass phenotype in betacellulin transgenic mice is EGFR-dependent. J. Bone Min. Res. 24, 455e467. Shimada, T., Hasegawa, H., Yamazaki, Y., Muto, T., Hino, R., Takeuchi, Y., Fujita, T., Nakahara, K., Fukumoto, S., Yamashita, T., 2004a. FGF-23 is a potent regulator of vitamin D metabolism and phosphate homeostasis. J. Bone Min. Res. 19, 429e435. Shimada, T., Kakitani, M., Yamazaki, Y., Hasegawa, H., Takeuchi, Y., Fujita, T., Fukumoto, S., Tomizuka, K., Yamashita, T., 2004b. Targeted ablation of Fgf23 demonstrates an essential physiological role of FGF23 in phosphate and vitamin D metabolism. J. Clin. Investig. 113, 561e568. Shimada, T., Mizutani, S., Muto, T., Yoneya, T., Hino, R., Takeda, S., Takeuchi, Y., Fujita, T., Fukumoto, S., Yamashita, T., 2001. Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc. Natl. Acad. Sci. U. S. A. 98, 6500e6505. Shimada, T., Urakawa, I., Yamazaki, Y., Hasegawa, H., Hino, R., Yoneya, T., Takeuchi, Y., Fujita, T., Fukumoto, S., Yamashita, T., 2004c. FGF-23 transgenic mice demonstrate hypophosphatemic rickets with reduced expression of sodium phosphate cotransporter type IIa. Biochem. Biophys. Res. Commun. 314, 409e414. Sitara, D., Razzaque, M.S., Hesse, M., Yoganathan, S., Taguchi, T., Erben, R.G., Juppner, H., Lanske, B., 2004. Homozygous ablation of fibroblast growth factor23 results in hyperphosphatemia and impaired skeletogenesis, and reverses hypophosphatemia in Phex-deficient mice. Matrix Biol. 23, 421e432. Spichtig, D., Zhang, H., Mohebbi, N., Pavik, I., Petzold, K., Stange, G., Saleh, L., Edenhofer, I., Segerer, S., Biber, J., Jaeger, P., Serra, A.L., Wagner, C.A., 2014. Renal

239

expression of FGF23 and peripheral resistance to elevated FGF23 in rodent models of polycystic kidney disease. Kidney Int. 85, 1340e1350. Streicher, C., Zeitz, U., Andrukhova, O., Rupprecht, A., Pohl, E., Larsson, T.E., Windisch, W., Lanske, B., Erben, R.G., 2012. Long-term Fgf23 deficiency does not influence aging, glucose homeostasis, or fat metabolism in mice with a nonfunctioning vitamin D receptor. Endocrinology 153, 1795e1805. Tamasi, J.A., Vasilov, A., Shimizu, E., Benton, N., Johnson, J., Bitel, C.L., Morrison, N., Partridge, N.C., 2013. Monocyte chemoattractant protein-1 is a mediator of the anabolic action of parathyroid hormone on bone. J. Bone Min. Res. 28, 1975e1986. The ADHR Consortium, 2000. Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat. Genet. 26, 345e348. Turan, S., Fernandez-Rebollo, E., Aydin, C., Zoto, T., Reyes, M., Bounoutas, G., Chen, M., Weinstein, L.S., Erben, R.G., Marshansky, V., Bastepe, M., 2014. Postnatal establishment of allelic Galphas silencing as a plausible explanation for delayed onset of parathyroid hormone resistance owing to heterozygous Galphas disruption. J. Bone Min. Res. 29, 749e760. Urakawa, I., Yamazaki, Y., Shimada, T., Iijima, K., Hasegawa, H., Okawa, K., Fujita, T., Fukumoto, S., Yamashita, T., 2006. Klotho converts canonical FGF receptor into a specific receptor for FGF23. Nature 444, 770e774. Weinman, E.J., Biswas, R.S., Peng, G., Shen, L., Turner, C.L., Steplock, D., Shenolikar, S., Cunningham, R., 2007. Parathyroid hormone inhibits renal phosphate transport by phosphorylation of serine 77 of sodium-hydrogen exchanger regulatory factor-1. J. Clin. Investig. 117, 3412e3420. Weinman, E.J., Steplock, D., Shenolikar, S., Biswas, R., 2011. Fibroblast growth factor23-mediated inhibition of renal phosphate transport in mice requires sodiumhydrogen exchanger regulatory factor-1 (NHERF-1) and synergizes with parathyroid hormone. J. Biol. Chem. 286, 37216e37221. Weinman, E.J., Steplock, D., Zhang, Y., Biswas, R., Bloch, R.J., Shenolikar, S., 2010. Cooperativity between the phosphorylation of Thr95 and Ser77 of NHERF-1 in the hormonal regulation of renal phosphate transport. J. Biol. Chem. 285, 25134e25138. Yuan, Q., Sato, T., Densmore, M., Saito, H., Schuler, C., Erben, R.G., Lanske, B., 2011. FGF-23/Klotho signaling is not essential for the phosphaturic and anabolic functions of PTH. J. Bone Min. Res. 26, 2026e2035. Zeitz, U., Weber, K., Soegiarto, D.W., Wolf, E., Balling, R., Erben, R.G., 2003. Impaired insulin secretory capacity in mice lacking a functional vitamin D receptor. FASEB J. 17, 509e511.