Scholarly article on topic 'FGF23 acts directly on renal proximal tubules to induce phosphaturia through activation of the ERK1/2–SGK1 signaling pathway'

FGF23 acts directly on renal proximal tubules to induce phosphaturia through activation of the ERK1/2–SGK1 signaling pathway Academic research paper on "Biological sciences"

CC BY-NC-ND
0
0
Share paper
Academic journal
Bone
OECD Field of science
Keywords
{"Fibroblast growth factor-23" / Klotho / "Phosphate homeostasis" / Kidney / Mice}

Abstract of research paper on Biological sciences, author of scientific article — Olena Andrukhova, Ute Zeitz, Regina Goetz, Moosa Mohammadi, Beate Lanske, et al.

Abstract Fibroblast growth factor-23 (FGF23) is a bone-derived endocrine regulator of phosphate homeostasis which inhibits renal tubular phosphate reabsorption. Binding of circulating FGF23 to FGF receptors in the cell membrane requires the concurrent presence of the co-receptor αKlotho. It is still controversial whether αKlotho is expressed in the kidney proximal tubule, the principal site of phosphate reabsorption. Hence, it has remained an enigma as to how FGF23 downregulates renal phosphate reabsorption. Here, we show that renal proximal tubular cells do express the co-receptor αKlotho together with cognate FGF receptors, and that FGF23 directly downregulates membrane expression of the sodium-phosphate cotransporter NaPi-2a by serine phosphorylation of the scaffolding protein Na+/H+ exchange regulatory cofactor (NHERF)-1 through ERK1/2 and serum/glucocorticoid-regulated kinase-1 signaling.

Academic research paper on topic "FGF23 acts directly on renal proximal tubules to induce phosphaturia through activation of the ERK1/2–SGK1 signaling pathway"

ELSEVIER

Contents lists available at SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/bone

Rapid Communication

FGF23 acts directly on renal proximal tubules to induce phosphaturia through activation of the ERK1/2-SGK1 signaling pathway

Olena Andrukhova a, Ute Zeitz a, Regina Goetz b, Moosa Mohammadi b, Beate Lanske c, Reinhold G. Erben a'*

a University of Veterinary Medicine Vienna, Vienna, Austria b New York University School of Medicine, New York, USA c Harvard School ofDental Medicine, Boston, USA

ARTICLE INFO ABSTRACT

Fibroblast growth factor-23 (FGF23) is a bone-derived endocrine regulator of phosphate homeostasis which inhibits renal tubular phosphate reabsorption. Binding of circulating FGF23 to FGF receptors in the cell membrane requires the concurrent presence of the co-receptor aKlotho. It is still controversial whether aKlotho is expressed in the kidney proximal tubule, the principal site of phosphate reabsorption. Hence, it has remained an enigma as to how FGF23 downregulates renal phosphate reabsorption. Here, we show that renal proximal tubular cells do express the co-receptor aKlotho together with cognate FGF receptors, and that FGF23 directly downregulates membrane expression of the sodium-phosphate cotransporter NaPi-2a by serine phosphorylation of the scaffolding protein Na+/H+ exchange regulatory cofactor (NHERF)-l through ERK1/2 and serum/glucocorticoid-regulated kinase-1 signaling.

© 2012 Elsevier Inc. All rights reserved.

Article history: Received 12 December 2011 Revised 9 April 2012 Accepted 22 May 2012 Available online 27 May 2012

Edited by: R Baron

Keywords:

Fibroblast growth factor-23 Klotho

Phosphate homeostasis

Kidney

Introduction

Fibroblast growth factor-23 (FGF23) was discovered as a phosphaturic hormone through genetic studies in patients suffering from autosomal dominant hypophosphatemic rickets, a renal phosphate wasting disease

[1]. FGF23 is secreted by osteocytes and osteoblasts in bone in response to elevated phosphate and vitamin D, and acts mainly on the kidney to down-regulate renal tubular phosphate reuptake and 1a-hydroxylase expression as part of a negative feedback loop between bone and kidney

[2]. The down-regulation of 1a-hydroxylase expression suppresses production of the biologically active vitamin D hormone, 1a,25-dihydroxyvitamin D3. Although the renal effects of FGF23 are well characterized at the whole organ level, the molecular mechanism underlying the phosphaturic action of FGF23 has remained elusive.

Cellular signaling of FGF23 requires the concurrent presence of FGF receptors (FGFRs) and the transmembrane protein aKlotho, which functions as a co-receptor [3]. While FGFRs are ubiquitously expressed, aKlotho expression is restricted to few tissues and hence

* Corresponding author at: Institute of Physiology, Pathophysiology and Biophysics, Dept. of Biomedical Sciences, University of Veterinary Medicine, Veterinaerplatz 1, 1210 Vienna, Austria. Fax: +43 1 250 77 4599.

E-mail address: Reinhold.Erben@vetmeduni.ac.at (R.G. Erben).

8756-3282/$ - see front matter © 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2012.05.015

targets the endocrine actions of circulating FGF23 to specific tissues. In the kidney, Klotho is expressed mainly in the distal tubule [4], but the major site of regulation of phosphate excretion is the proximal tubule. Although earlier in vitro microperfusion experiments with isolated rabbit proximal tubules suggested a possible direct effect of FGF23 on the proximal tubule [5], the current dogma is that FGF23 acts on the distal tubule, generating an unknown endocrine or paracrine secondary signal that in turn signals back to the proximal tubule to lower apical membrane expression of the sodium-phosphate cotransporters type 2a (NaPi-2a) and 2c (NaPi-2c) [6,7] that primarily mediate renal tubular phosphate reabsorption. A recent study, however, suggested that aKlotho may be expressed at low levels also in the proximal tubule, and that aKlotho may itself be a phosphaturic hormone [8]. The extracellular domain of aKlotho can be shed from the cell surface and released into the blood circulation, and it is thought that this secreted form of aKlotho may have the ability to alter the function and abundance of membrane glycopro-teins such as NaPi-2a by removing sialic acid or other terminal sugars from sugar chains through a putative glycosidase activity [8-10].

It was the aim of the current study to elucidate further the molecular mechanism underlying the phosphaturic action of FGF23. Here, we show that murine proximal tubular epithelium expresses aKlotho, and that FGF23 acts directly on proximal tubules to downregulate membrane expression of NaPi-2a via activation of ERK1/2 and serum/ glucocorticoid-regulated kinase-1 (SGK1).

Material and methods

Animals

All animal studies were approved by the Ethical Committee of the University of Veterinary Medicine, Vienna, and by the Austrian Federal Ministry of Science and Research. Wild-type C57BL/6 mice were bred in our in-house animal facility, and were kept at 24 °C with a 12 hour/ 12 hour light/dark cycle with free access to a normal mouse chow (Ssniff, Soest, Germany) and tap water.

For some additional experiments, wild-type mice, mice with a nonfunctioning vitamin D receptor (VDRA/A), and compound mutants deficient in VDR and Klotho (Kl-/-/VDRA/A) were generated by intercrossing VDR+/A/Kl+/- double heterozygous animals, and were genotyped as described earlier [11]. To normalize mineral homeostasis in VDRa/a and Kl-/-/VDRA/A mice [11], all genotypes were fed a rescue diet (Ssniff, Soest, Germany) which contained 2.0% calcium, 1.25% phosphorus, 20% lactose and 6001U vitamin D/kg, starting from 16 days of age.

Laser capture microdissection (LCM)

Kidney cryosections were prepared from snap-frozen tissues. Proximal and distal tubules were harvested using a Veritas (Arcturus) LCM system. RNA was extracted using the PicoPure RNA isolation kit (Qiagen).

qRT-PCR

RNA purity and integrity were determined with a bioanalyzer (Agilent). Reverse transcription was performed using iScript™ cDNA Synthesis Kit (Bio-Rad). Real-time PCR was performed in duplicate on a Rotor-Gene™ 6000 cycler (Corbett Life Science) using QuantiFast™ SYBR® Green PCR Kit (Qiagen). RT-minus samples served as a control to exclude amplification from genomic DNA, and amplicon melting analysis was performed to exclude primer dimerization and unspecific amplification. (32-Microglobulin was used as house-keeping gene for normalization of the mRNA expression data. N-fold change in gene expression was calculated using the Pfaffl model [12].

Isolation and culture of renal proximal tubular cells

Primary mouse proximal tubular epithelial cells were isolated from C57BL/6 mice according to previously established protocols by collagenase digestion and density gradient centrifugation, and cultured in serum-free, hormonally defined culture medium [13,14]. Purity of proximal tubular cells was examined by qRT-PCR analysis of calbindin D28k, transient receptor potential vanilloid-5 (TRPV5), and NaPi2a mRNA.

Isolation of renal proximal tubular segments

Renal proximal tubules were isolated as reported previously [15-17]. 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 containing 1% antibiotic/antimycotic and placed on ice. Individual proximal tubule segments were identified based on morphology in a dissection microscope at x 25-40 magnification by their appearance and dimensions.

In vitro experiments with proximal tubular cells and segments

In vitro experiments with cultured proximal tubular cells and dissected tubular segments were performed in serum-free, hormonally defined culture medium at 37 °C in 5% CO2 [13,14]. Proximal tubular cells were incubated with 1-100 ng/ml of recombinant human FGF23 R176/179Q (rFGF23) [18] for 0.5,1, 2, and 4 h. Proximal tubular segments were incubated with rFGF23 (100 ng/ml), 10 ng/ml of the SGK1 kinase inhibitor GSK 650394 (Axon Medchem) or 10 ng/ml of the ERK1/2 inhibitor PD184352 (Sigma) alone or in combination with rFGF23, or 10-8M hPTH(1-34) (Bachem) for 1, 2 and 4 h. For co-immunoprecipitation experiments, proximal tubular segments were incubated with rFGF23 (100 ng/ml) or 10-8M hPTH(1-34), alone or in combination with 10 ng/ml of GSK 650394 for 2 h. To assess the Klotho dependency of the effects of FGF23, proximal tubular segments from 3-month-old wild-type, VDRA/A, and Kl-/-/VDRA/A mice were incubated with 1-100 ng/ml rFGF23 for 2 h. Protein samples for Western blotting analysis or co-immunoprecipitation were collected in lysis buffer.

In vivo experiments

Four-month-old male C57BL/6 mice received a single intraperitone-al injection of vehicle (phosphate-buffered saline with 2% DMSO) or rFGF23 (10 |ag per mouse). Spontaneous urine was collected before and 8 h after injection of rFGF23. Eight hours post-injection, the mice were killed by exsanguination from the abdominal V. cava under anesthesia with ketamine/xylazine (67/7 mg/kg i.p.). Serum phosphorus was analyzed on a Hitachi 912 Autoanalyzer (Boehringer Mannheim), urinary phosphorus and urinary creatinine were measured on a Cobas c111 analyzer (Roche). Kidney cortices were immediately dissected in ice-cold isolation buffer after being removed from animals and then homogenized using a Potter-Elvehjem homogenizer at 4 °C. Brush border membrane vesicles (BBMV) were prepared using three consecutive magnesium precipitations (15 mM), and solubilized in Laemmli sample buffer for Western blotting. To verify BBM purity, the activity of the BBM enzyme alkaline phosphatase and leucine aminopeptidase was regularly monitored in BBM fractions.

Western blotting

Protein samples were fractionated on SDS-PAGE (50 ^g/well) and transferred to a nitrocellulose membrane (Thermo Scientific). Immuno-blots were incubated overnight at 4 °C with primary antibodies including anti-NaPi-2a (generous gift of Drs. Jurg Biber and Heini Murer, University of Zurich), anti-total-ERK1/2 (BD Biosciences), anti-phospho-ERK1/2 (Cell Signaling), anti-total-SGK1 (Applied Biosystems), anti-phospho-SGK1 (Santa Cruz Biotechnology), anti-aKlotho (Alpha Diagnostics, 1:1000), or anti-p-actin (Sigma) antibody 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 peroxidase-conjugated 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). The expression levels were normalized to Ponceau S stain. Expression levels of phospho-SGK1 and phospho-ERK1/2 were normalized to total SGK1 and total ERK1/2 protein expression.

Co-immunoprecipitation

Homogenate protein samples of kidney cortex (1 mg) or dissected proximal tubular segments (40 |ag) were incubated with 2 |ag of anti-NHERF-1 (Abcam), anti-phosphoserine (Alpha Diagnostics), or anti-NaPi-2a (generous gift of Drs. Jurg Biber and Heini Murer, University of Zurich) antibody at 4 °C overnight. The immune complexes were

captured by adding 50 Protein A or G agarose/sepharose beads (Santa Cruz Biotechnology), followed by overnight incubation at 4 °C with gentle rocking. The immunoprecipitates were collected by centrifugation at 1000xg for 5 min at 4 °C and washed for 4 times in PBS, each time repeating the centrifugation step. After the final wash, the pellets were suspended in 40 of electrophoresis sample buffer and boiled for 2-3 min. Western blot analysis was performed using primary anti-NHERF-1 or anti-NaPi-2a antibody.

¡mmunohistochemistry

For immunohistochemistry, 5-^m-thick paraffin sections of paraformaldehyde-fixed kidneys from untreated wild-type mice (for anti-aKlotho staining), or from wild-type mice injected with rFGF23 (n = 4) or vehicle (n = 3) (for anti-NHERF-1 and anti-phosphoserine staining) were prepared. Before immunofluorescence staining, 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. All following steps were performed in PBS containing 0.3% Triton X-100 and 5% normal goat serum. Without rinsing, sections were incubated with polyclonal rabbit anti-aKlotho (Alpha Diagnostics, 1:1000), or anti-NHERF-1 (Abcam, 1:300) and mouse monoclonal anti-phosphoserine (Alpha Diagnostics, 1:1,000) antibodies at 4 °C overnight. After washing, sections were incubated for 1.5 h with goat anti-rabbit Alexa 548 (for aKlotho and for NHERF-1 detection) and goat anti-mouse Alexa 488 (for P-Ser detection) secondary antibodies (both from Invitrogen, diluted 1:400). Controls were performed by omitting either one or both secondary antibodies. The slides were analyzed on a Zeiss LSM 510 Axioplan 2 confocal microscope equipped with a 63 x oil immersion lens (NA 1.3). By use of the multitrack function, individual fluorochromes were scanned with laser excitation at 488 and 543 nm separately with appropriate filter sets to avoid cross talk. Controls were scanned with identical laser excitation and filter settings. Pictures were processed using Adobe Photoshop (overlays). Some mouse

Fig. 1. aKlotho is expressed in renal proximal tubules. (A) mRNA expression of aKlotho, SGK1, calbindin D28k (Cal28k), TRPV5, and FGFR1 -4 normalized to (^-microglobulin expression in renal proximal tubules (PT) and distal tubules (DT) harvested from 4 mice by laser capture microdissection (LCM). For the purpose of comparison, mRNA expression levels in PT were plotted relative to the expression levels in DT. Data are mean ± SD. 'Denotes P< 0.05 vs. DT by t-test. (B) Co-staining of paraffin sections from murine kidneys with anti-aKlotho (red) antibody and DAPI (blue), showing basolateral aKlotho staining pattern in distal and proximal tubules. Proximal tubule from inset in the merged image is shown at higher magnification in the lower right panel. Lower left and middle panels show H&E-stained paraffin sections for comparison of subcellular localization. Lower middle panel (inset in lower left panel) shows H&E-stained proximal tubule at higher magnification. Original magnification x630. (C) Western blot analysis of Klotho protein expression normalized to (J-actin expression in isolated proximal and distal tubular segments from wild-type mice. PT, proximal tubule; DT, distal tubule.

kidney paraffin sections were stained with hematoxylin and eosin (H&E) by routine methods.

Statistical analyses

Statistics were computed using SPSS for Windows 17.0. The data were analyzed by t-test for comparison of 2 groups, or analysis of variance (ANOVA) followed by Student-Newman-Keuls (SNK) multiple comparison test for comparison of more than 2 groups. P values of less than 0.05 were considered significant. The data are presented as the mean ± SD.

Results

aKlotho is expressed in renal proximal tubules

addition, LCM-harvested proximal tubules expressed FGFR1,3, and 4, but not 2 (Fig. 1A), in accordance with earlier reports [19]. Molecules typically found in the distal tubule such as calbindin D28k or the transient receptor potential vanilloid-5 (TRPV5) channel were expressed at negligible levels in proximal tubules (Fig. 1A), thereby confirming that our results did not relate to contamination of the proximal tubules by distal tubules. Immu-nohistochemical staining of paraffin sections from murine kidneys showed comparable expression of aKlotho in proximal and distal tubules (Fig. 1B, upper left panel). Moreover, the major subcellular site of aKlotho protein expression appeared to be the basolateral membrane in both distal and proximal tubules (Fig. 1B, right panels). Western blot analysis of proximal and distal tubular segments isolated from wild-type C57BL/6 mice showed similar Klotho protein expression in proximal and distal tubules (Fig. 1C), confirming the immunohistochemical results.

To address the question whether FGF23 has a direct effect on the renal proximal tubule, we first measured mRNA expression of aKlotho in proximal renal tubules harvested from mice by laser capture microdissection (LCM), and compared the expression level to that found in distal tubules. As shown in Fig. 1A, aKlotho is expressed in proximal tubules, albeit at an approximately 2-fold lower level than that measured in distal tubules. In

FGF23 directly activates ERK1/2 and SGIO in proximal tubular epithelial cells

It is known that FGFR activation by FGF23 leads to phosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2) [3]. To examine whether FGF23 directly activates FGFR in proximal tubular epithelium,

Fig. 2. FGF23 directly activates ERK1/2-SGK1 signaling in proximal tubule epithelial cells. Western blot analysis of phosphorylated ERK1/2 (pERK1/2) and SGK1 (pSGKl) and total (phosphorylated and non-phosphorylated) ERK1/2 (tERK1/2) and SGK1 (tSGKl) in samples of total protein isolated from cultured primary proximal tubular cells that had been treated for 0.5,1, 2 and 4 h either with vehicle (Co) or 100 ng/ml recombinant FGF23 (rFGF23) (A), or for 2 h either with vehicle (Co) or with 1, 2.5,10 or 100 ng/ml of rFGF23 (B). (C) Western blot analysis of pSGK1/tSGK1 ratio in proximal tubular segments treated for 2 h with vehicle (Co) or 100 ng/ml rFGF23 alone or in combination with an ERK1/ 2 inhibitor (iERK1/2). The quantitative data are from 4 independent experiments each. Data are mean±SD. 'Denotes P<0.05 vs. control by ANOVA followed by SNK test.

we stimulated cultured proximal tubular epithelial cells with recombinant FGF23 (rFGF23), and analyzed ERK1/2 phosphorylation after cell stimulation. rFGF23 time and dose dependently increased phosphorylation of ERK1/2 (Figs. 2A and B). In renal mouse cortical collecting duct cells, it was shown that activation of ERK1/2 leads to downstream activation of SGK1 [20]. Since SGK1 is also expressed in proximal tubules (Fig. 1A), we tested whether FGF23 can also activate SGK1 in proximal tubular epithelium. Indeed, addition of rFGF23 to cultured proximal tubular epithelial cells led to augmented phosphorylation of SGK1 in a time and dose dependent fashion (Figs. 2A and B). Doses as low as 1 ng/ml rFGF23 clearly increased phospho-ERK1/2 and phospho-SGK1 in cultured proximal tubular epithelial cells after 2 h of incubation (Fig. 2B). To test whether SGK1 is downstream of ERK1/2 activation, we incubated isolated proximal tubular segments from wild-type mice for 2 h with rFGF23 alone or in combination with an ERK1/2 inhibitor. In the presence of an ERK1/2 inhibitor, rFGF23 did not increase phosphorylation of SGK1, showing that activation of ERK1/2 by rFGF23 leads to downstream activation of SGK1 (Fig. 2C).

FGF23 regulates NaPi-2a expression in proximal tubular segments through SGK1

To examine whether FGF23 directly affects the membrane expression of NaPi-2a in the proximal tubule and whether SGK1 is a downstream mediator of this effect, we treated isolated proximal tubular segments with rFGF23 alone or in combination with a SGK1 inhibitor. Similar to parathyroid hormone (PTH), the other major phosphaturic hormone, rFGF23 time dependently down-regulated NaPi2a protein expression in the proximal tubular segments (Fig. 3). Notably, this effect of FGF23 was completely blocked by co-treatment of the tubular segments with a SGK1 inhibitor.

FGF23 signaling reduces membrane abundance of the NaPi-2a/NHERF-1 complex in vivo and phosphorylates NHERF-1 via SGK1

PTH decreases membrane expression of NaPi-2a by phosphoryla-tion of serine-77 (S77) in the Na+/H+ exchange regulatory cofactor

Fig. 3. FGF23 regulates NaPi-2a membrane expression in the proximal tubular segments through SGK1 activation. Western blot analysis of NaPi-2a expression in isolated proximal tubular segments treated with recombinant FGF23 (rFGF23), an SGK1 inhibitor (iSGK1), or a mixture of rFGF23 and iSGK1. PTH was used as a positive control. The quantitative data are from at least 4 independent experiments each. Data are mean ± SD. 'Denotes P<0.05 vs. control for the same time point by ANOVA followed by SNK test.

(NHERF)-1, a scaffolding protein, leading to internalization and degradation of NaPi-2a [21,22]. A recent study suggested that the phosphaturic action of FGF23 may also involve phosphorylation of NHERF-1 at S77 [23]. To test this possibility in vivo, we injected mice with rFGF23, and performed reciprocal co-immunoprecipitation of the NaPi-2a/NHERF-1 complex on renal brush border membrane vesicle (BBMV) preparations. rFGF23 caused an almost 4-fold increase in urinary phosphate excretion compared to vehicle injection (46.9 ±10.4 vs. 12.3 ± 0.6 mmol/mmol creatinine in vehicle-treated controls, P<0.05), and this effect was accompanied by hypophosphatemia (2.2 ±0.2 vs. 3.4 ±0.3 mmol/l in vehicle-treated controls, P < 0.05). The rFGF23-induced phosphaturia was associated with a 50% reduction in membrane abundance of the NaPi-2a/NHERF-1 complex (Fig. 4A). Immunofluorescence staining of paraffin sections with anti-NHERF-1 and anti-phosphoserine antibodies showed a clear reduction in the apical membrane abundance of NHERF-1, and a general increase in serine phosphorylation in proximal tubular epithelium of rFGF23-treated mice (Fig. 4B). Moreover, the co-staining also suggested increased serine phosphorylation of NHERF-1 in the apical membrane, 8 h after rFGF23 injection (Fig. 4B). Collectively, these data show that FGF23 signaling targets the NaPi-2a/NHERF-1 complex in the apical cell membrane of renal tubular epithelium in vivo.

Next, we examined whether FGF23 induces phosphorylation of NHERF-1 and whether SGK1 is involved in this process in isolated proximal tubular segments. Similar to PTH, a 2-hour treatment of proximal tubules with rFGF23 led to a marked increase in serine phosphorylation of NHERF-1 (Fig. 4C). Importantly, FGF23- but not PTH-induced phosphorylation of NHERF-1 could be completely blocked by an SGK1 inhibitor (Fig. 4C), suggesting that activated SGK1 mediates phosphorylation of NHERF-1.

FGF23 signaling in proximal tubules is Klotho dependent

To confirm the functional role of Klotho in FGF23 signaling, we isolated proximal tubular segments from 3-month-old wild-type, VDRA/A, and K/-/-/VDRa/a mice, and treated them for 2 h with 1,10, and 100 ng/ml rFGF23 in vitro. We chose Kl-/-/VDRA/A mice as aKlotho deficient animal model, because Kl-/- mice are characterized by severe alterations in mineral homeostasis [11] which might affect the results of subsequent ex vivo experiments. Kl-/-/VDRA/A mice on rescue diet are nor-mocalcemic and normophosphatemic, and have unchanged PTH serum levels compared with VDRA/A mice [11]. rFGF23 treatment resulted in a similar dose dependent down-regulation of NaPi-2a expression in proximal tubular segments from wild-type and VDRA/A mice (Fig. 5), showing that this effect is vitamin D independent. However, proximal tubular segments from K(-/-/VDRA/A mice did not respond to 1 and 10 ng/ml rFGF23, and showed a distinctly diminished response to the suppressive effect of rFGF23 on NaPi-2a expression at 100 ng/ml, relative to wildtype and VDRA/A mice. These results indicate that the direct effect of FGF23 on NaPi-2a protein expression in proximal tubules is dependent on the presence of Klotho at lower FGF23 concentrations, whereas rFGF23 concentrations of greater than 10 ng/ml elicit Klotho independent effects, at least to some extent.

Discussion

FGF23 is one of the most important endocrine regulators of phosphate metabolism. However, the molecular mechanism underlying the phosphaturic action of FGF23 is still unknown. The current study has shown 1) that aKlotho is expressed at the mRNA and protein level in proximal tubules together with FGFR1, 3, and 4, 2) that FGF23 induces phosphorylation of ERK1/2 and SGK1 in cultured proximal tubule cells, 3) that the FGF23-induced downregulation of NaPi-2a expression in renal proximal tubular segments is SGK1 dependent, 4) that FGF23 signaling leads to serine phosphorylation of NHERF-1 in proximal tubular segments, 5) that FGF23 reduces membrane abundance of the NaPi-2a/NHERF-1 complex in vivo, and 6) that the

C IP: P-Ser—NHERF-1)

Co rFGF23 PTH rFGF23 rFGF23 PTH ¡SGK1 iSGK1 iSGK1

Fig. 4. FGF23 regulates renal proximal tubular membrane abundance of the NaPi-2a/NHERF-1 complex in vivo and leads to phosphorylation of NHERF-1 in vitro. (A) Reciprocal co-immunoprecipitation (co-IP) of the NaPi-2a/NHERF-1 complex in BBMV preparations from recombinant FGF23- (rFGF23) and vehicle-treated mice (Veh, n=4 animals each). BBMVs were isolated 8h post-injection. WB: Western blot. Data are mean±SD. 'Denotes P<0.05 vs. vehicle control by t-test. (B) Co-staining of kidney paraffin sections with anti-phosphoserine (green) and anti-NHERF-1 antibodies (red) from rFGF23- and vehicle-treated mice, 8 h post-injection. Original magnification x630. (C) Immunoprecipitation (IP) of serine-phosphorylated (P-Ser) proteins, followed by Western blot analysis of NHERF-1, from isolated proximal tubular segments treated for 2 h with rFGF23, PTH, or iSGK1, or mixtures of either rFGF23 or PTH plus iSGK1. Co, control; Veh, vehicle.

Fig. 5. FGF23-induced regulation of NaPi-2a membrane expression in proximal tubular segments is Klotho dependent at lower concentrations. Western blot analysis of NaPi-2a expression in proximal tubular segments isolated from 3-month-old wild-type, VDRA/A, and Kl-/-/VDRA/A mice. Segments were treated for 2 h with 1-100 ng/ml recombinant FGF23 (rFGF23) in vitro. The quantitative data are from at least 4 independent experiments each. Data are mean ± SD. 'Denotes P < 0.05 vs. control within the same genotype by ANOVA followed by SNK test.

presence of the co-receptor Klotho is essential for the FGF23-induced downregulation of NaPi-2a expression in renal proximal tubular segments at near physiological concentrations.

Our data are in very good agreement with the study by Weinman and coworkers [23]. The latter authors recently reported that FGF23 increases phosphorylation of ERK1/2 in cultured proximal tubular cells. In addition, they showed that proximal tubular cells from NHERF-1 null mice are resistant to the inhibitory action of FGF23 on phosphate transport, and that proximal tubular cells from NHERF-1 null mice infected with a mutated form of the NHERF-1 protein which cannot be phosphorylated at serine-77 do not respond to FGF23 treatment. By using a different approach, our study provides direct evidence that FGF23 induces phosphorylation of NHERF-1 in isolated proximal tubular segments, which is probably the best in vitro model of proximal tubular physiology because the cells stay in their natural environment and do not lose polarity. In addition, we show that FGF23 downregulates membrane abundance of NHERF-1 in vivo. Our finding that apical membrane abundance of NHERF-1 was profoundly reduced in mice 8 h after rFGF23 injection is in accordance with the known fact that PTH-induced phosphorylation of NHERF-1 leads to internalization and degradation of the NaPi-2a/ NHERF-1 complex [21,22]. Taken together, our study and the study by Weinman et al. [23] provide compelling evidence that NHERF-1 is a downstream target of FGF23 signaling.

The current study suggests that FGF23 acts directly on proximal tubular epithelium through its canonical, aKlotho and FGFR dependent, signaling pathway. Moreover, we show for the first time that SGK1 is an essential component of the intracellular signaling cascade leading to the FGF23-induced downregulation of NaPi-2a membrane

expression. It is currently unknown whether NHERF-1 is directly phosphorylated by activated SGK1. Since SGK1 can directly interact with NHERF family proteins in the distal tubule [24], it is conceivable that NHERF-1 is directly phosphorylated by SGK1 also in proximal tubules.

It was previously thought that aKlotho is mainly expressed in the distal tubule [4]. Earlier immunohistochemical studies using a rat monoclonal anti-Klotho antibody on cryosections failed to detect aKlotho in proximal tubules of mice [25]. In addition, Farrow and coworkers [26] showed in a time course study that the earliest changes in activation of ERK1/2 after injection of FGF23 in vivo in mice occur in the distal tubules. Therefore, the current dogma is that FGF23 acts on the distal tubule where it generates an unknown endocrine or para-crine secondary signal that in turn signals back to the proximal tubule to downregulate transcellular phosphate transport [6,7]. Hu et al. [8] proposed an alternative hypothesis, namely that aKlotho itself may be a phosphaturic hormone by altering the glycosylation pattern of NaPi-2a integrated in the apical membrane through its putative enzymatic activity. The latter hypothesis requires the presence of aKlotho at the apical cell membrane where NaPi-2a is expressed. However, our study using a polyclonal rabbit antibody clearly showed that aKlotho is expressed in proximal tubular epithelium, but mainly at the basolateral membrane, suggesting that the major function of aKlotho may be its function as a co-receptor for blood-borne FGF23. The discrepant findings regarding aKlotho expression in the kidney in our compared with earlier studies [25] may be explained by differences in the anti-Klotho antibodies used. If the phosphaturic action of FGF23 is a direct effect on proximal tubules, how can it then be explained that the earliest signaling events after injection of FGF23 in vivo occur in distal tubules [26]? Unpublished data (Andrukhova et al.) from our laboratory have shown that FGF23 is also an important regulator of the TRPV5 epithelial calcium channel in distal tubules, suggesting that FGF23 may have parallel and independent effects in proximal and distal renal tubules. The FGF23-induced signaling events in distal renal tubules may occur faster than in proximal tubules, explaining the findings by Farrow et al. [26].

A caveat of the current study is that we used a concentration of 100ng/ml in most of our in vitro experiments. Faul and coworkers [27] recently showed that FGF23 can signal in an aKlotho independent fashion at concentration of 10 ng/ml and higher. In agreement with the data of Faul et al. [27], we also found some Klotho independent activity of 100 ng/ml rFGF23 to suppress NaPi-2a expression in proximal tubular segments in vitro. Therefore, although our in vivo data fully confirm our in vitro findings, we cannot totally rule out the possibility of aKlotho independent effects induced by FGF23 signaling in most of our in vitro studies. However, our experiments with proximal tubular segments isolated from Kl-/-/VDRA/A mice clearly showed that lower, near physiological concentrations of FGF23 directly suppress NaPi-2a protein expression in proximal tubular epithelium in a Klotho dependent manner. Nevertheless, it is clear that additional experiments are necessary to confirm the FGF23-induced signaling pathways at physiological concentrations in renal proximal tubules.

We propose a model (Fig. 6) wherein FGF23 and PTH signaling converge at the NaPi-2a/NHERF-1 complex, providing a molecular explanation for the observed interaction between both signaling pathways in the regulation of proximal tubular phosphate reabsorption in vitro [23] and in vivo (Andrukhova et al., unpublished). Taken together, our data show that FGF23 directly acts on proximal tubular cells to down-regulate membrane abundance of NaPi-2a through the ERK1/2-SGK1-NHERF-1 signaling axis. Hence, our data uncover the long sought molecular mechanism of the phosphaturic action of FGF23. Improved knowledge of the cellular mechanisms involved in the phosphaturic action of FGF23 may open up new possibilities for therapeutic intervention in phosphate-wasting disorders and other diseases in which modulation of renal phosphate excretion is a therapeutic goal.

Fig. 6. Proposed model of FGF23 signaling in proximal tubular cells. FGF23 binds to the FGFR1c-aKlotho complex and activates ERK1/2 kinase leading to SGK1 phosphoryla-tion. SGK1 in turn phosphorylates NHERF-1, leading to internalization and degradation of NaPi-2a. PTH signals through activation of PKA and PKC, which also leads to phos-phorylation of NHERF-1. KL, aKlotho; PKA, protein kinase A; PKC, protein kinase C.

Acknowledgments

We thank Claudia Bergow for help with the biochemical analyses, Sonja Sabitzer for help with the LCM, Carsten Wagner, Nati Hernando, and Nicole Kampik for help with the isolation of proximal tubular segments, Martin Glösmann for help with the confocal microscopy, and Graham Tebb for critically reading and editing the manuscript. The polyclonal rabbit anti-NaPi-2a antibody was a generous gift of Drs. Jürg Biber and Heini Murer, University of Zurich. Some of the rFGF23 used in this study was a gift of Amgen Inc., Thousand Oaks, CA, USA. This work was supported by grants from the University of Veterinary Medicine Vienna and from the Austrian Science Fund (FWF P24186-B21) to R.G.E, U.S. NIH/NIDDK DK072944 to B.L., and U.S. NIH/NIDCR DE13686 to M.M. O.A. was supported by a postdoctoral fellowship of the University of Veterinary Medicine Vienna.

References

[1] The ADHR Consortium. Autosomal dominant hypophosphataemic rickets is associated with mutations in FGF23. Nat Genet 2000;26:345-8.

[2] Juppner H, Wolf M, Salusky IB. FGF-23: more than a regulator of renal phosphate handling? J Bone Miner Res 2010;25:2091-7.

[3] Urakawa I, Yamazaki Y, Shimada T, Iijima K, Hasegawa H, Okawa K, et al. Klotho converts canonical FGF receptor into a specific receptor for FGF23. Nature 2006;444:770-4.

[4] Kuro-o M, Matsumura Y, Aizawa H, Kawaguchi H, Suga T, Utsugi T, et al. Mutation of the mouse Klotho gene leads to a syndrome resembling ageing. Nature 1997;390: 45-51.

[5] Baum M, Schiavi S, Dwarakanath V, Quigley R Effect of fibroblast growth factor-23 on phosphate transport in proximal tubules. Kidney Int 2005;68:1148-53.

[6] Farrow EG, Summers lJ, Schiavi SC, McCormick JA, Ellison DH, White KE. Altered renal FGF23-mediated activity involving MAPK and Wnt: effects of the Hyp mutation. J Endocrinol 2010;207:67-75.

[7] White KE, Econs MJ. Fibroblast growth factor-23. In: Rosen CJ, editor. Primer on the metabolic bone diseases and disorders of mineral metabolism. Washington, DC: American Society of Bone and Mineral Research; 2008. p. 112-6.

[8] Hu MC, Shi M, Zhang J, Pastor J, Nakatani T, Lanske B, et al. Klotho: a novel phos-phaturic substance acting as an autocrine enzyme in the renal proximal tubule. FASEB J 2010;24:3438-50.

[9] Chang Q, Hoefs S, Van Der Kemp AW, Topala CN, Bindels RJ, Hoenderop JG. The beta-glucuronidase klotho hydrolyzes and activates the TRPV5 channel. Science 2005;310:490-3.

[10] Kurosu H, Yamamoto M, Clark JD, Pastor JV, Nandi A, Gurnani P, et al. Suppression of aging in mice by the hormone Klotho. Science 2005;309:1829-33.

[11] Anour R, Andrukhova O, Ritter E, Zeitz U, Erben RG. Klotho lacks a vitamin D independent physiological role in glucose homeostasis, bone turnover, and steady-state PTH secretion in vivo. PLoS One 2012;7:e31376.

[12] Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR Nucleic Acids Res 2001;29:e45.

[13] Qi W, Johnson DW, Vesey DA, Pollock CA, Chen X. Isolation, propagation and characterization of primary tubule cell culture from human kidney. Nephrology (Carlton) 2007;12:155-9.

[14] Cunningham REX, Steplock D, Shenolikar S, Weinman EJ. Defective PTH regulation of sodium-dependent phosphate transport in NHeRf-1 renal proximal tubule cells and wild-type cells adapted to low-phosphate media. Am J Physiol Renal Physiol 2005;289:F933-8.

[15] Schafer JA, Watkins ML, Li L, Herter P, Haxelmans S, Schlatter E. A simplified method for isolation of large numbers of defined nephron segments. Am J Physiol 1997;273:F650-7.

[16] Burg MB, Grantham J, Abramow M, Orloff J, Schafer JA. Preparation and study of fragments of single rabbit nephrons. J Am Soc Nephrol 1997;8:675-83.

[17] Gonzalez-Mariscal L, Namorado MC, Martin D, Luna J, Alarcon L, Islas S, et al. Tight junction proteins ZO-1, ZO-2, and occludin along isolated renal tubules. Kidney Int 2000;57:2386-402.

[18] Goetz R Nakada Y, Hu MC, Kurosu H, Wang L, Nakatani T, et al. Isolated C-terminal tail of FGF23 alleviates hypophosphatemia by inhibiting FGF23-FGFR-Klotho complex formation. Proc Natl Acad Sci U S A 2010;107:407-12.

[19] Gattineni J, Bates C, Twombley K, Dwarakanath V, Robinson ML, Goetz R, et al. FGF23 decreases renal NaPi-2a and NaPi-2c expression and induces hypophosphatemia in

vivo predominantly via FGF receptor 1. Am J Physiol Renal Physiol 2009;297: F282-91.

[20] Michlig S, Mercier A, Doucet A, Schild L, Horisberger JD, Rossier BC, et al. ERK1/2 controls Na, K-ATPase activity and transepithelial sodium transport in the principal cell of the cortical collecting duct of the mouse kidney. J BiolChem 2004;279: 51002-12.

[21] Deliot N, Hernando N, Horst-Liu Z, Gisler SM, Capuano P, Wagner CA, et al. Parathyroid hormone treatment induces dissociation of type IIa Na+ P(i) cotransporter-Na+/H+ exchanger regulatory factor-1 complexes. Am J Physiol Cell Physiol 2005;289:C159-67.

[22] Weinman EJ, Biswas RS, Peng G, Shen L, Turner CL, E X, et al. Parathyroid hormone inhibits renal phosphate transport by phosphorylation of serine 77 of sodium-hydrogen exchanger regulatory factor-1. J Clin Invest 2007;117:3412-20.

[23] Weinman EJ, Steplock D, Shenolikar S, Biswas R FGF-23-mediated inhibition of renal phosphate transport in mice requires NHERF-1 and synergizes with PTH. J Biol Chem 2011;286:37216-21.

[24] Yun Cc, Chen Y, Lang F. Glucocorticoid activation of Na( + )/H(+) exchanger isoform 3 revisited. The roles of SGK1 and NHERF2. J Biol Chem 2002;277:7676-83.

[25] Li SA, Watanabe M, Yamada H, Nagai A, Kinuta M, Takei K. Immunohistochemical localization of Klotho protein in brain, kidney, and reproductive organs of mice. Cell Struct Funct 2004;29:91-9.

[26] Farrow EG, Davis SI, Summers LJ, White KE. Initial FGF23-mediated signaling occurs in the distal convoluted tubule. J Am Soc Nephrol 2009;20:955-60.

[27] Faul C, Amaral AP, Oskouei B, Hu MC, Sloan A, Isakova T, et al. FGF23 induces left ventricular hypertrophy. J Clin Invest 2011;121:4393-408.