Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

FGF23 promotes renal calcium reabsorption through the TRPV5 channel

FGF23 promotes renal calcium reabsorption through the TRPV5 channel Article FGF23 promotes renal calcium reabsorption through the TRPV5 channel 1 1 1 1 1 Olena Andrukhova , Alina Smorodchenko , Monika Egerbacher , Carmen Streicher , Ute Zeitz , 2 3 2 1 4 Regina Goetz , Victoria Shalhoub , Moosa Mohammadi , Elena E Pohl , Beate Lanske & 1,* Reinhold G Erben Abstract enzymes, and released into the blood circulation (Imura et al, 2007). In addition, a soluble Klotho isoform can be produced by alternative aKlotho is thought to activate the epithelial calcium channel Tran- splicing of the Klotho mRNA (Matsumura et al, 1998; Shiraki-Iida sient Receptor Potential Vanilloid-5 (TRPV5) in distal renal tubules et al, 1998). This truncated Klotho gene product lacks exons 4 and 5 through its putative glucuronidase/sialidase activity, thereby pre- in mice (Shiraki-Iida et al, 1998). Klotho is mainly expressed in venting renal calcium loss. However, aKlotho also functions as the renal distal convoluted tubules (DCT) and in the brain choroid obligatory co-receptor for fibroblast growth factor-23 (FGF23), a plexus (Kuro-o et al, 1997). However, Klotho expression is also bone-derived phosphaturic hormone. Here, we show that renal found at other locations, for example in renal proximal tubules calcium reabsorption and renal membrane abundance of TRPV5 (Hu et al, 2010; Andrukhova et al, 2012) or parathyroid glands are reduced in Fgf23 knockout mice, similar to what is seen in (Shiraki-Iida et al, 1998; Urakawa et al, 2006; Imura et al, 2007). aKlotho knockout mice. We further demonstrate that aKlotho The molecular function of Klotho is still controversial. Klotho was neither co-localizes with TRPV5 nor is regulated by FGF23. Rather, initially thought to be an anti-aging factor (Kuro-o et al, 1997). Later apical membrane abundance of TRPV5 in renal distal tubules and studies suggested that soluble Klotho may have the ability to alter thus renal calcium reabsorption are regulated by FGF23, which the function and abundance of membrane glycoproteins by remov- binds the FGF receptor-aKlotho complex and activates a signaling ing sialic acid or other terminal sugars from sugar chains through a cascade involving ERK1/2, SGK1, and WNK4. Our data thereby putative glycosidase activity (Chang et al, 2005; Kurosu et al, 2005; Hu et al, 2010). identify FGF23, not aKlotho, as a calcium-conserving hormone in the kidney. Apart from its putative enzymatic function, membrane-bound Klotho also functions as the co-receptor for the bone-derived hor- Keywords calcium homeostasis; fibroblast growth factor-23; Klotho; renal mone fibroblast growth factor-23 (FGF23). Binding of FGF23 to ubiquitously expressed FGF receptors (FGFR) requires Klotho as an calcium reabsorption; transient receptor potential vanilloid-5 Subject Categories Membrane & Intracellular Transport; Signal Transduction obligatory co-receptor (Kurosu et al, 2006; Urakawa et al, 2006), DOI 10.1002/embj.201284188 | Received 8 December 2012 | Revised 30 restricting the hormonal action of FGF23 to Klotho-expressing October 2013 | Accepted 18 November 2013 tissues. FGF23 down-regulates renal proximal tubular phosphate EMBO Journal (2014) 33, 229–246 reuptake and 1a-hydroxylase expression. 1a-hydroxylase is the key enzyme for production of the biologically active vitamin D hor- mone, 1a,25-dihydroxyvitamin D (1,25(OH) D). Because FGF23 is 3 2 Introduction secreted by osteocytes and osteoblasts in bone in response to ele- vated phosphate and vitamin D, this hormone forms a negative aKlotho (Klotho) is a single pass transmembrane protein, which feedback loop between bone and kidney (Juppner et al, 2010). shares sequence homology with family I b-glycosidases, including It was reported by Chang and colleagues (Chang et al, 2005) that soluble Klotho is a regulator of the epithelial calcium channel tran- b-glucuronidases (Kuro-o et al, 1997). However, Klotho lacks essen- tial active site glutamic acid residues typical for this family of glyco- sient receptor potential vannilloid-5 (TRPV5), a glycoprotein essential sidases (Tohyama et al, 2004). The extracellular domain of Klotho, for entry of calcium in calcium-transporting renal epithelial cells. which consists of two type I b-glycosidase domains (KL1 and KL2), Apical membrane expression of fully glycosylated TRPV5 is the rate- limiting step in distal renal tubular transcellular calcium transport can be shed from the cell surface by membrane-anchored proteolytic 1 University of Veterinary Medicine Vienna, Vienna, Austria 2 New York University School of Medicine, New York, NY, USA 3 Amgen Inc., Thousand Oaks, CA, USA 4 Harvard School of Dental Medicine, Boston, MA, USA *Corresponding author. Tel: +43 1 250 77 4550; Fax: +43 1 250 77 4599; E-mail: [email protected] ª 2014 The Authors This is an open access article under the terms of the Creative Commons Attribution License, The EMBO Journal Vol 33 |No 3 | 2014 which permits use, distribution and reproduction in any medium, provided the original work is properly cited. The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al / / which also involves the calcium-binding proteins calbindin D9k and on renal 1a-hydroxylase activity, Kl and Fgf23 mice produce D28k (CalD28k) for intracellular transport, and the sodium-calcium excessive amounts of 1,25(OH) D, and die early from the sequelae exchanger (NaCX) and the plasma membrane calcium ATPase 1b of hypervitaminosis D, hypercalcemia and hyperphosphatemia. The (PMCA1b) for basolateral extrusion of calcium (Lambers et al, 2006). pivotal role of hypervitaminosis D for the development of the pre- / / The current model for TRPV5 regulation by Klotho is that soluble Klo- mature aging phenotype in Kl and Fgf23 mice is underscored tho promotes insertion of TRPV5 in the apical membrane of distal by the fact that ablation of vitamin D signaling completely rescues tubular cells through its putative glycosidase activity, thereby stabiliz- the premature aging phenotype in these mice (Hesse et al, 2007; ing the interaction between glycosylated TRPV5 and membrane- Anour et al, 2012; Streicher et al, 2012). bound galectin (Chang et al, 2005; Cha et al, 2008) (Fig 1A). A major In order to examine vitamin D independent effects of Klotho and explanatory gap in this model is how soluble Klotho crosses from Fgf23 deficiency on renal calcium excretion in skeletally mature the systemic circulation into the urinary space. It was reported that mice, we crossed mice with a non-functioning vitamin D receptor Δ/Δ / / the 130 kDa isoform of Klotho protein can be detected in the urine of (VDR ) with Kl and Fgf23 mice, and analyzed them at mice (Chang et al, 2005; Hu et al, 2010). The soluble forms of mur- 9 months of age. All mice were kept life-long on a so-called rescue ine Klotho are large proteins with molecular weights of about 130 diet rich in calcium, phosphorus, and lactose. The rescue diet and 65 kDa for the shed and alternatively spliced isoforms, respec- normalizes the calcium absorption defect of VDR mutant mice in Δ/Δ / Δ/Δ / tively (Shiraki-Iida et al, 1998; Imura et al, 2007). Because the glo- the gut, so that VDR mice, but also Kl /VDR and Fgf23 / Δ/Δ merular filter is impermeable to proteins larger than approximately VDR mice on this diet are normocalcemic (Erben et al, 2002; 60–65 kDa under physiological conditions, soluble Klotho in plasma Hesse et al, 2007; Anour et al, 2012). In this study, we show that / Δ/Δ / Δ/Δ cannot be filtered. Alternatively, Klotho could be secreted by distal skeletally mature Kl /VDR and Fgf23 /VDR mice are char- tubular cells into the urinary space, or shed from the apical cell acterized by an almost identical renal calcium wasting phenotype, membrane. However, experimental evidence for tubular secretion or and that FGF23 is a regulator of distal tubular TRPV5 membrane tubular apical membrane shedding of Klotho is lacking thus far. abundance and renal calcium reabsorption through an intracellular In accordance with an important role of Klotho in the regulation signaling cascade involving ERK1/2, SGK1, and WNK4. of distal renal tubular TRPV5 activity, Klotho null (Kl ) mice on a vitamin D deficient diet lose urinary calcium despite unchanged Results renal expression of fully glycosylated TRPV5 (Alexander et al, 2009). However, this animal model is complicated by the fact that Kl mice have elevated 1,25(OH) D blood levels. 1,25(OH) Dis We first examined renal calcium excretion in skeletally mature, 2 2 Δ/Δ / Δ/Δ 9-month-old wild-type (WT), VDR , and Kl /VDR compound thought to be involved in the regulation of distal tubular TRPV5 (Lambers et al, 2006). Due to the lacking suppressive effect of Fgf23 mutant mice on rescue diet. We found pronounced loss of urinary Figure 1. Models of the regulation of apical membrane TRPV5 in renal distal tubules by Klotho and FGF23. A Model based on previous studies for the regulation of apical membrane TRPV5 by secreted Klotho. TRPV5 is necessary for apical entry of calcium, which is then transported through the cell bound to calbindin D9k and D28k, and extruded at the basolateral side via PMCA1 and NCX. Secreted Klotho is thought to specifically hydrolyze sugar residues from the glycan chains on TRPV5 which in turn stabilizes TRPV5 in the membrane through interaction of the sugar residues with extracellular galectin (Chang et al, 2005; Cha et al, 2008). The cellular secretion process of Klotho in this model is unclear. Adapted from Cha et al (2008). B Our proposed model of Fgf23-aKlotho signaling in renal distal tubular cells. Fgf23 binds to the basolateral FGFR1c-Klotho complex and activates ERK1/2 leading to SGK1 phosphorylation. SGK1 in turn activates WNK4, stimulating WNK4-TRPV5 complex formation, and increasing intracellular transport of fully glycosylated TRPV5 from the Golgi apparatus to the plasma membrane. PTH signaling activates membrane-anchored TRPV5 by protein kinase A (PKA)-mediated phosphorylation. The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 230 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 2. Renal calcium reabsorption and TRPV5 plasma membrane abundance in Fgf23 and Klotho deficient mouse models. A–D Urinary excretion of calcium corrected for creatinine (UrCa/Crea) (A), Western blotting quantification of core (75 kDa) and complex (92 kDa) glycosylated TRPV5 protein expression in renal cortical total membrane fractions (B), and Western blot analysis of membrane-bound Klotho in renal total protein extracts (C) in Δ/Δ / Δ/Δ / Δ/Δ 9-month-old male WT, VDR , Kl /VDR , and Fgf23 /VDR on rescue diet. Antibody specificity and glycosylation of TRPV5 and Klotho was controlled by a glycosylation assay (pink staining) followed by Western blot analysis of TRPV5 and Klotho, respectively (D). Anti-Klotho antibodies detecting the membrane- bound (anti-cytoplasmic domain, upper panel) or the membrane-bound and ectodomain shed (anti KL2 domain, lower panel) forms of the protein were used. / Δ/Δ Protein extracts from lung and spleen, as well as kidney extracts from Kl /VDR mice served as negative controls for TRPV5 and Klotho protein expression, respectively. # D/D Data information: *P < 0.05 vs. WT, P < 0.05 vs. VDR mice. Only the 135 kDa transmembrane isoform of Klotho was quantified in (C). Data in (A–C) represent mean  s.e.m. of 4–9 animals each. Frames in Western blot images shown in (B) and (D) indicate splicing events. Source data are available online for this figure. ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 231 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al / Δ/Δ calcium in 9-month-old Kl /VDR mice (Fig 2A), which was enzymatically alter the glycosylation pattern of TRPV5 in the apical associated with a 50% downregulation of membrane expression of cell membrane (Chang et al, 2005). Co-localization of Klotho and Δ/Δ complex glycosylated TRPV5 relative to VDR mice (Fig 2B). It is TRPV5 in distal renal tubular cells has been demonstrated by Chang known from previous studies that VDR mutant mice on rescue diet et al (Chang et al, 2005). In our immunohistochemical analysis, show a defect in renal tubular reabsorption of calcium due to lower TRPV5 staining was exclusively seen in distal tubules also express- expression of calbindin D9k (Erben et al, 2002). The core and com- ing calbindin D28k, further demonstrating the specificity of the anti- plex glycosylated forms of TRPV5 protein were unchanged in VDR body used (Supplementary Fig S2A). The membrane-bound form of single mutant mice (Fig 2B), showing that vitamin D signaling is Klotho was mainly found in the cytoplasm and the basolateral not essential for the regulation of TRPV5. However, urinary loss of membrane, whereas TRPV5 was mainly localized in the apical Δ/Δ / Δ/Δ calcium was almost twice as high in compound mutant mice com- membrane in WT, VDR , and Fgf23 /VDR mice (Fig 3A). We pared with VDR single mutants, i.e., lack of Klotho aggravated the observed an identical subcellular distribution of Klotho in distal renal calcium wasting seen in VDR single mutants (Fig 2A). This tubular epithelium, employing an anti-Klotho antibody detecting finding corroborates earlier reports that Klotho has an essential role both the membrane-bound and the ectodomain shed form of the in the regulation of renal TRPV5 activity (Chang et al, 2005), and protein (Supplementary Fig S2B). Some TRPV5 staining was also further shows that this mechanism is VDR independent. Notably, seen basolaterally in all genotypes (Fig 3A). Co-localization of / Δ/Δ / Δ/Δ similar to Kl /VDR mice, 9-month-old Fgf23 /VDR mice Klotho and TRPV5, however, was almost absent, and only seen in also showed renal calcium wasting and reduced membrane expres- some cytoplasmic or basolateral areas of the distal tubular cells sion of TRPV5 (Fig 2A and B). Indeed, the absence of Fgf23 resulted (Fig 3A and Supplementary Fig S2). In analogy to the immunoblot- in a stronger downregulation of core and complex glycosylated ting data (Fig 2B), membrane expression of TRPV5 was clearly / Δ/Δ TRPV5 compared with the absence of Klotho (Fig 2B). Using anti- reduced in distal tubules of Fgf23 /VDR mice (Fig 3A). To Klotho antibodies raised against the short intracellular region of the assess the subcellular localization of Klotho in more detail, we per- membrane-bound Klotho isoform or against the extracellular KL2 formed immuno-electron microscopic analyses in renal tissue from domain, we found renal Klotho protein expression unchanged in WT mice, using anti-Klotho antibodies detecting either the trans- Δ/Δ / Δ/Δ both VDR single and Fgf23 /VDR compound mutants membrane or both the transmembrane and the ectodomain shed (Fig 2C and Supplementary Fig S1A). Although the anti-TRPV5 and forms of the protein. Both antibodies showed the presence of Klotho anti-Klotho antibodies we used for immunoblotting and immuno- protein in the membrane of the basal labyrinth, but staining was histochemistry have been successfully employed by other groups absent in the apical membrane of distal tubular cells (Fig 3B). Kid- (Sandulache et al, 2006; Segawa et al, 2007), we confirmed the neys from Kl mice were used as a negative control (Fig 3B). specificity of the antibodies, using tissue extracts from mouse lung To examine whether FGF23 was able to regulate TRPV5 in gain- Δ/Δ / Δ/Δ and spleen where TRPV5 is not expressed (Fig 2D), extracts from of-function models, we treated WT, VDR ,and Kl /VDR mice kidneys of TRPV5 mice (Supplementary Fig S1B), and extracts with recombinant FGF23 (rFGF23). Within 8 h post-injection, rFGF23 / Δ/Δ from kidneys of Kl /VDR mice lacking all isoforms of Klotho reduced urinary calcium excretion to very low levels in WT mice (Fig 2C and D) as negative controls. In addition, we confirmed the (Fig 4A). Serum calcium and serum parathyroid hormone (PTH) glycosylation of the 92 kDa TRPV5 band and of the 135 kDa trans- remained unchanged (Supplementary Fig S3A and Fig 4B). The membrane form of Klotho (Fig 2D). Both anti-Klotho antibodies rFGF23-induced reduction in urinary calcium excretion occurred in Δ/Δ / Δ/Δ detected the 135 kDa glycosylated full-length transmembrane form WT and VDR , but not in Kl /VDR mice (Fig 4C), and was as the predominant form of Klotho protein in renal extracts (Fig 2D). associated with distinctly upregulated TRPV5 expression in the distal Δ/Δ / The 130 kDa band which might represent the ectodomain shed form tubular apical membrane (Fig 4D) of WT and VDR , but not of Kl Δ/Δ of Klotho was also glycosylated (Fig 2D). The 65 kDa Klotho band /VDR mice. These data demonstrate that the FGF23-induced detected by the anti-transmembrane antibody was not glycosylated, upregulation of TRPV5 and the concomitant increase in renal tubular whereas the 55 kDa Klotho band detected by the anti-KL2 antibody calcium reabsorption are Klotho dependent. Despite the profound was glycosylated (Fig 2D). Therefore, the nature of these two puta- changes in TRPV5 expression and renal calcium reabsorption in WT Δ/Δ tive fragments of Klotho is not clear. Using the methodology and VDR mice, expression of membrane-bound and shed Klotho described by Hu and coworkers (Hu et al, 2010), we were unable to protein remained unchanged (Fig 4E and Supplementary Fig S3B), detect any isoform of Klotho protein in native, salt-precipitated, or and Klotho did not show co-localization with TRPV5 in the distal concentrated urine of WT mice by Western blotting (Fig 2D and Sup- tubular apical membrane (Fig 4D and Supplementary Fig S3C). / Δ/Δ plementary Fig S1C). The reason for this discrepancy is not clear, but Immunohistochemical Klotho staining was absent in Kl /VDR may be due to differences in the anti-Klotho antibodies used. mice (Fig 4D and Supplementary Fig S3C), confirming the specificity The fact that Klotho deficiency and Fgf23 deficiency have almost of both anti-Klotho antibodies used. Moreover, soluble Klotho in identical effects on renal TRPV5 is difficult to explain on the basis of serum remained below the detection limit of an immunoassay spe- the model shown in Fig 1A. Rather, this finding points to an essen- cific for murine Klotho (data not shown) and anti-Klotho immuno- tial role of Fgf23 in the regulation of TRPV5. We reported earlier blotting of urine was negative in rFGF23-treated WT mice that renal function and morphology of kidneys is normal in (Supplementary Fig S3D). Finally, we treated isolated distal tubular Δ/Δ / Δ/Δ 9-month-old VDR and Fgf23 /VDR mice (Streicher et al, segments in vitro with rFGF23 in the presence and absence of a FGFR 2012), ruling out renal functional impairment as a confounding inhibitor. The FGF23-induced upregulation of complex glycosylated factor in these experiments. A central element of the model shown TRPV5 expression was completely blunted in the presence of the in Fig 1A is a direct, albeit transient, protein-protein interaction FGFR inhibitor, showing that FGF23 signals through the FGFR to between TRPV5 and Klotho, because soluble Klotho is thought to increase distal tubular TRPV5 membrane expression (Fig 4F). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 232 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 3. Membrane-bound aKlotho and TRPV5 do not co-localize in renal distal tubule cells. A Immunohistochemical co-staining with anti-transmembrane aKlotho (red or green) antibody, anti-TRPV5 (green), anti-b-actin (red), and DAPI (blue) of paraffin Δ/Δ / Δ/Δ sections from kidneys of 9-month-old WT, VDR , and Fgf23 /VDR mice on rescue diet. Right panels show H&E-stained paraffin sections for comparison of subcellular localization. Scale bar, 20 lm. B Immuno-electron microscopic staining using anti-Klotho antibodies against transmembrane, and transmembrane and shed forms in kidneys of WT (left panels) and Klotho (right panels) mice. Upper panels show the apical cell area, lower panels show the basolateral area with the basal labyrinth. Arrows mark the apical cell membrane, asterisks mark mitochondria, and the symbol # marks the nucleus. Scale bar, 500 nm. To confirm the functional role of the FGF23-induced upregulation changes in fluorescence over time (30 min) after addition of ruthe- of TRPV5 in the apical membrane of distal tubular epithelium, we nium red in kidney slices from rFGF23- and vehicle-treated mice, performed intracellular calcium imaging, employing 2-photon respectively. For distal tubular calcium reabsorption, only TRPV5 microscopy of Fluo-4-loaded, 300-lm-thick renal slices prepared and 6 are thought to be relevant (Woudenberg-Vrenken et al, 2009). from vehicle- and rFGF23-treated WT mice, 8 h post-injection. TRPV5 is about 100-fold more sensitive to ruthenium red than Distal tubules of FGF23-treated mice showed a 5-fold increase in TRPV6 (Hoenderop et al, 2001). Therefore, ruthenium red can be fluorescence intensity, relative to those of vehicle-treated mice used to discriminate the channel function of TRPV5 from that of (Fig 4G). The FGF23-induced increase in fluorescence intensity was TRPV6. To further demonstrate the regulation of TRPV5 activity by largely abrogated by ex vivo addition of 10 lM of the TRPV inhibitor FGF23, we treated Fluo-4-loaded kidney slices from WT mice with ruthenium red (Fig 4G). Supplementary videos 1 and 2 show the rFGF23 or vehicle in vitro. rFGF23 gradually increased intracellular ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 233 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al fluorescence over 2 h in distal tubules (Fig 4H and Supplementary (Ring et al, 2007). WNK4-mediated regulation of these channels video 3). The latter effect was reversed by ruthenium red at concen- involves physical association between the channel protein and trations as low as 1 lM (Fig 4H and Supplementary video 4). WNK4 (Cai et al, 2006; He et al, 2007). Interestingly, mutations in At 1 lM, ruthenium red does not antagonize TRPV6 function WNK4 can lead to urinary calcium loss in humans (Mayan et al, (Hoenderop et al, 2001). 10 and 50 lM of ruthenium red did not 2002). In addition, studies in transfected Xenopus laevis oocytes further decrease the fluorescence signal, showing that TRPV6 does have suggested that WNK4 influences the membrane abundance of not play a major role in the FGF23-induced increase in calcium TRPV5 (Jiang et al, 2007, 2008). uptake in distal tubular cells in this experimental setting (Fig 4H Based on this knowledge, we hypothesized that FGF23 may regu- and Supplementary video 4). The discrepancy between the complete late TRPV5 through a signaling cascade involving ERK1/2, SGK1, and reversal of the rFGF23-induced increase in intracellular Fluo-4 fluo- WNK4. To test this, we first examined whether activation of ERK1/2 rescence by ruthenium red in Fig 4H and the partial reversal and SGK1 was regulated by FGF23 in loss- and gain-of-function mod- observed in Fig 4G can probably be explained by the fact that in the els. Phospho-ERK1/2 and phospho-SGK1 were decreased in the kid- / Δ/Δ ex vivo experiment shown in Fig 4G distal tubular cells were pre- ney of Fgf23 /VDR compound mutant mice (Supplementary Fig Δ/Δ / exposed to rFGF23 for 8 h, whereas rFGF23 was added at time 0 in S4A and B). rFGF23 administration to WT, VDR ,and Fgf23 / Δ/Δ the in vitro experiment shown in Fig 4H. The latter experimental VDR compound mutant mice led to an increase in renal complex setting therefore looks at very early aspects of the rFGF23-induced glycosylated TRPV5 (Fig 5A), phospho-ERK1/2 (Fig 5B), and phos- increase in distal tubular calcium uptake. This is also reflected in pho-SGK1 (Fig 5C). These findings suggest that FGF23 signaling leads the almost 2-fold higher rFGF23-induced increase in relative fluores- to activation of renal SGK1 and increased membrane abundance of cence in the ex vivo (Fig 4G) compared with the in vitro experiment complex glycosylated TRPV5 in a VDR independent fashion. (Fig 4H). Collectively, these data show that the FGF23-induced Next, we examined whether WNK4 is a downstream mediator of upregulation of TRPV5 results in increased calcium uptake in distal FGF23 signaling. Reciprocal immunoprecipitation experiments on tubular epithelial cells, and that the FGF23-induced stimulation of renal protein extracts, using anti-phosphoserine and anti-WNK4 apical calcium entry is mainly mediated by TRPV5. However, antibodies showed that serine phosphorylation of WNK4 was because ruthenium red at 10 lM did not completely reverse the increased 8 h after treatment of WT mice with rFGF23 (Fig 5E). Fur- rFGF23-induced increase in intracellular Fluo-4 fluorescence in dis- ther, reciprocal immunoprecipitation revealed a physical association tal tubules (Fig 4G) and TRPV5 and TRPV6 can form heteromulti- between fully glycosylated TRPV5 and WNK4 in the kidney of WT, Δ/Δ / Δ/Δ mers (Hellwig et al, 2005) with intermediate ruthenium red VDR and Fgf23 /VDR mice (Fig 5F). Whereas total WNK4 sensitivity, we cannot exclude a potential contribution of TRPV6 to protein abundance was not different between the genotypes the effects of FGF23 on renal tubular calcium reabsorption in vivo. (Fig 5D), the amount of TRPV5-WNK4 complexes in kidneys of / Δ/Δ The intracellular calcium-binding protein calbindin D28k is thought Fgf23 /VDR mutant mice was decreased by about 50% to be involved in intracellular calcium transport in renal distal epithe- (Fig 5F). As a control for the specificity of the co-immuno- lium (Lambers et al, 2006). Therefore, we examined the expression precipitation, we used an antibody against the sodium phosphate and intracellular distribution of this cytoplasmic protein in kidneys of transporter-2c (NaPi-2c), a protein exclusively expressed in vehicle- or rFGF23-treated WT mice. We found upregulated mRNA renal proximal tubules. As an additional specificity control, the abundance, a trend towards increased protein expression, but no co-immunoprecipitation experiments were also done using protein change in the intracellular distribution pattern of calbindin D28k, 8 h extracts from cultured murine smooth muscle cells, which neither after rFGF23 administration (Supplementary Fig S3E and F). express TRPV5 nor WNK4 (Fig 5F). In summary, these data demonstrate that FGF23 regulates renal As far as it is known, WNK4 mainly regulates the trafficking of TRPV5 in loss- and gain-of-function models independent of altera- proteins from the Golgi apparatus to the plasma membrane and tions in Klotho expression; Klotho is mainly expressed basolaterally their glycosylation (Jiang et al, 2007, 2008). However, WNK4 may in distal tubule epithelium and does not co-localize with TRPV5 in also be involved in regulating TRPV5 endocytosis (Cha & Huang, the apical cell membrane; soluble Klotho is undetectable in urine, 2010). Immunohistochemical analysis suggested a subcellular redis- and FGF23 signals through the FGFR-Klotho complex to regulate tribution of WNK4 from the basolateral side of the distal tubular TRPV5 in distal tubules. These findings do not support the signaling cells to the area beneath the apical cell membrane in rFGF23-treated Δ/Δ / Δ/Δ model shown in Fig 1A. Rather, the data suggest that FGF23 WT and VDR , but not Kl /VDR mice (Fig 6A). In addition, regulates membrane abundance of TRPV5 by a signaling mechanism WNK4 and TRPV5 appeared to co-localize after rFGF23 treatment in which Klotho functions as the co-receptor required for FGF23 sig- beneath, but not directly within, the apical cell membrane in WT Δ/Δ naling. Hence, we next asked how FGF23 signaling influenced distal and VDR mice (Fig 6A). Quantitative analysis of the immunohis- tubular membrane abundance of TRPV5. tochemical images confirmed the increase in TRPV5-WNK4 co-local- Δ/Δ / FGF23 signaling leads to activation of extracellular signal-regu- ization after rFGF23 treatment in WT and VDR , but not Kl / Δ/Δ lated kinase 1 and 2 (ERK1/2) (Urakawa et al, 2006) which itself VDR mice (Fig 6A). activates serum and glucocorticoid-induced kinase 1 (SGK1) in renal To assess the FGF23-induced changes in subcellular distribution mouse cortical collecting duct cells (Michlig et al, 2004), and in of TRPV5 and WNK4 in more detail, we performed immuno-electron renal proximal tubules (Andrukhova et al, 2012). Renal DCT SGK1 microscopic analyses in renal tissue taken from WT mice 8 h after a phosphorylates and thereby activates with-no-lysine kinase 4 single injection of rFGF23. TRPV5 immunostaining was found exclu- (WNK4) (Ring et al, 2007), which is critically involved in mem- sively in distal tubules and not in proximal tubules (Supplementary brane transport of other ion channels in the DCT such as Na -Cl Fig S5). In vehicle controls, TRPV5 was mainly found in the apical co-transporter (NCC) and renal outer medullar K channel (ROMK1) cell membrane and the basal labyrinth in distal tubular cells, with The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 234 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal some TRPV5 present in membrane vesicles (Fig 6B). FGF23 treat- Golgi apparatus to the apical plasma membrane (Fig 6B). Quantifi- ment induced a striking increase in TRPV5 present in the apical cell cation of the relative grey levels in the apical area of the distal membrane as well as in membrane vesicles trafficking from the tubular cells between nucleus and apical cell membrane in the ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 235 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 4. FGF23 increases urinary calcium reabsorption, TRPV5 plasma membrane abundance and activity in the kidney in gain-of-function mouse models. A, B Urinary calcium excretion (A) and serum PTH (B) in 4-month-old WT mice injected i.p. with vehicle or a single dose of 10 lg rFGF23 per mouse at time 0. Δ/Δ / Δ/Δ C Urinary calcium excretion in 4-month-old WT, VDR , and Kl /VDR mice on rescue diet injected i.p. with vehicle or a single dose of 10 lg rFGF23 per mouse, 8 h post-injection. D Immunohistochemical co-staining of kidney paraffin sections with anti-transmembraneaKlotho (red) antibody, anti-TRPV5 (green) and DAPI (blue). Original magnification ×630. Δ/Δ / Δ/Δ E Western blot analysis of membrane-bound Klotho in renal total protein extracts from 4-month-old WT, VDR , and Kl /VDR mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. F Complex glycosylated TRPV5 protein expression in isolated distal tubular segments treated for 2 h in vitro with rFGF23 alone or in combination with a specific FGFR inhibitor (iFGFR). 2+ G Quantification and original images of intracellular Ca levels in renal distal tubular cells in 300-lm-thick kidney slices of 3-month-old WT mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. Images are overlays of fluorescent with phase contrast images. Kidney slices were stained with the calcium-sensitive dye Fluo-4. Ruthenium red (10 lM) was used as TRPV inhibitor. H Time-dependent changes in distal tubular fluorescence in Fluo-4-loaded, 300-lm-thick kidney slices of 3-month-old WT mice treated at time 0 with rFGF23 (100 ng/ml) or vehicle. After 120, 135, and 150 min, 1, 10, and 50 lM of the TRPV inhibitor ruthenium red (RR) was added, respectively. Data information: Data in (A–C) represent mean  s.e.m. of 3–5 animals each. *P < 0.05 vs. vehicle-treated control in (A–C). Data in (E) represent mean  s.e.m. # # of 3–5 animals each. In (G), *P < 0.05 vs. vehicle-treated, P < 0.05 vs. rFGF23-treated WT mice. In (H), *P < 0.05 vs. vehicle, P < 0.05 vs. fluorescence at 120 min in rFGF23-treated slices. Fluorescence intensity in (G) and (H) was quantified in 7–9 regions of interest per experimental group and time point from three independent experiments. Scale bar, 20 lm in (D) and (G). Source data are available online for this figure. immuno-electron microscopic pictures showed reduced grey levels In order to explore the contribution of other pathways initiated (i.e., more black) in rFGF23-treated mice (Fig 6C), in accordance by FGF23 signaling through the FGFR-Klotho complex, we treated with increased abundance of TRPV5-containing membrane vesicles isolated distal tubular segments with rFGF23 or with specific Src in the sub-apical cellular area. kinase (SRC), phospholipase C (PLC), and phosphatidylinositol In vehicle controls, WNK4 was mainly associated with localized 3-kinase (PI3K) inhibitors, alone or in combination, for 2 h in vitro. areas in the basal labyrinth as well as with membrane vesicles traf- SRC and PLC inhibition did not influence the rFGF23-induced up- ficking to the apical plasma membrane, with most of the immuno- regulation of TRPV5 expression (Supplementary Fig S6). However, staining observed in the peri-nuclear area (Fig 6B). Eight hours after PI3K inhibition partially blocked the rFGF23-induced increase in rFGF23 treatment, WNK4 was still found at the basal labyrinth, but TRPV5 expression (Supplementary Fig S6). Taken together, these the main localization was around membrane vesicles in close prox- data suggest that ERK1/2 activation is the major pathway, but that imity to and along the apical plasma membrane (Fig 6B). These the PI3K pathway may play an additional role in the regulation of findings are consistent with the notion that activated WNK4 is TRPV5 by FGF23 in renal distal tubules. involved in the FGF23-induced regulation of the cellular trafficking In order to more rigorously establish the putative FGFR-ERK1/ of TRPV5 from the Golgi apparatus to the apical plasma membrane 2-SGK1-WNK4 signaling pathway of the TRPV5 regulation by in distal tubular epithelium, especially in fusion of membrane vesi- FGF23, we reconstituted this pathway in MDCK cells. To this end, cles with the apical cell membrane. We found no evidence of WNK4 we transfected MDCK cells with murine TRPV5, SGK1, and WNK4 associated with endocytotic pits. alone or in combination, treated the transfected cells with vehicle or To verify that the regulation of TRPV5 by FGF23 is a direct effect rFGF23 alone or in combination with soluble Klotho for 12 h, and on the distal tubule and to test the essential roles of ERK1/2 and subsequently examined complex glycosylated TRPV5 protein SGK1 in this process, we isolated distal tubular segments from WT expression by immunoblotting. It is known that MDCK cells express mice, and treated these segments with rFGF23 alone or rFGF23 ERK1/2 and FGFR1 (Kessler et al, 2008), and as shown in Fig 8A, together with specific SGK1 or ERK1/2 inhibitors for 1 to 4 h these cells do not endogenously express TRPV5. rFGF23 profoundly in vitro. rFGF23 time-dependently upregulated phosphorylation of upregulated TRPV5 expression in MDCK cells co-transfected with ERK1/2 and SGK1 (Fig 7A), as well as protein expression of TRPV5 TRPV5, SGK1, and WNK4 (Fig 8A). Recombinant soluble Klotho in distal tubular segments (Fig 7B). The latter effect was completely protein, alone or in combination with rFGF23, did not influence blocked in the presence of SGK1 or ERK1/2 inhibitors (Fig 7B), TRPV5 expression (Fig 8A). In the absence of SGK1, rFGF23 failed showing that ERK1/2 and SGK1 activation are essential for the to upregulate TRPV5 expression (Fig 8A). However, small increases FGF23-induced regulation of TRPV5 expression in the distal tubule. in TRPV5 expression in response to rFGF23 treatment were seen in In the presence of an ERK1/2 inhibitor, rFGF23 did not increase cells transfected with TRPV5 and SGK1 only (Fig 8A). A possible phosphorylation of SGK1 in isolated proximal tubular segments explanation for this finding may be that MDCK cells, although not from WT mice, indicating that SGK1 is a mediator downstream of reported in the literature, might express low amounts of WNK4 or ERK1/2 activation (Fig 7C). Furthermore, immunoprecipitation other WNK kinases. To examine the biological response to the experiments in isolated distal tubular segments from WT mice rFGF23-induced increase in TRPV5 expression, we transfected revealed that the rFGF23-induced increase in serine phosphorylation MDCK cells with TRPV5, SGK1, and WNK4, treated the transfected of WNK4 is blocked in the presence of an SGK1 inhibitor (Fig 7D). cells with vehicle, rFGF23 or Klotho for 12 h, and then monitored Thus, activation of ERK1/2 and SGK1 by FGF23 leads to down- intracellular calcium concentrations by 2-photon microscopy after stream activation of WNK4. loading the cells with Fluo-4. rFGF23 induced a striking increase in The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 236 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 5. FGF23 increases TRPV5 protein abundance in the plasma membrane through a signaling pathway involving ERK1/2, SGK1, and WNK4. A–D Western blot analysis of complex glycosylated TRPV5 (A) protein (92 kDa) expression in renal cortical total membrane fractions, as well as phosphorylated ERK1/2 (B), Δ/Δ / Δ/Δ phosphorylated SGK1 (C), and total WNK4 (D) in homogenized renal cortex samples of 9-month-old WT, VDR and Fgf23 /VDR mice on rescue diet that had been injected with rFGF23 (5 lg/mouse). Tissue samples were taken from mice 8 h post-injection. E Reciprocal immunoprecipitation (IP) of serine-phosphorylated (P-Ser) proteins, followed by Western blot (WB) analysis of WNK4 or vice versa from homogenized renal cortex protein samples of 4-month-old WT mice treated with vehicle (Veh) or rFGF23 (10 lg/mouse) 8 h before necropsy. Western blot analysis of WNK4 in renal cortex protein samples immunoprecipitated with WNK4 antibody was used as a loading control. F Co-immunoprecipitation of TRPV5/WNK4 complexes. WNK4 or TRPV5 were immunoprecipitated with specific antibodies (anti-TRPV5, anti-WNK4, anti-NaPi-2c) Δ/Δ / Δ/Δ from homogenized renal cortex protein samples of 9-month-old WT, VDR and Fgf23 /VDR mice on rescue diet. Subsequently, Western blot analysis was performed with corresponding anti-TRPV5 or anti-WNK4 antibodies to identify co-precipitated TRPV5 and WNK4 protein, respectively. Primary cultured murine smooth muscle cells (SMC) were used as negative control. Data information: Data in (A–D) represent mean  s.e.m. of 5–8 animals each, *P < 0.05 vs. vehicle-treated WT, P < 0.05 vs. rFGF23-treated WT mice. Data in (E) represent mean  s.e.m. of five animals each. Data in (F) represent mean  s.e.m. of 4–5 animals each. Frame in Western blot image indicates splicing # D/D event. In (F) *P < 0.05 vs. WT, P < 0.05 vs. VDR mice. Source data are available online for this figure. intracellular calcium in MDCK cells transfected with TRPV5, SGK1 blocked by the TRPV inhibitor ruthenium red (Fig 8B). To unequiv- and WNK4, reflecting increased TRPV5 channel activity (Fig 8B). In ocally show that FGF23 stimulates a TRPV5-mediated calcium influx contrast, recombinant soluble Klotho protein did not influence intra- pathway, we treated MDCK cells transfected with TRPV5, SGK1 and cellular calcium in TRPV5-expressing cells (Fig 8B). As expected, WNK4 with rFGF23 in the presence or absence of the calcium chela- the rFGF23-induced increase in intracellular calcium could be tor EGTA (Fig 8C). In the presence of EGTA, rFGF23 did not ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 237 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 6. FGF23 increases apical plasma membrane abundance of TRPV5 and causes re-distribution of WNK4. Δ/Δ A Immunohistochemical co-staining with anti-WNK4 (red) antibody, anti-TRPV5 (green) and DAPI (blue) of kidney paraffin sections from 4-month-old WT, VDR and / Δ/Δ Klotho /VDR mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. Scale bar, 20 lm. Right panel shows quantification of distal tubular TRPV5 and WNK4 co-localization performed in four animals per group with 3–6 images per animal and 4–6 different regions of interest per image. B Immuno-electron microscopic staining using anti-TRPV5 (left panels) and anti-WNK4 (right panels) antibodies in kidneys from 3 to 4-month-old WT mice treated with vehicle (Veh) or rFGF23 (10 lg/mouse) 8 h before necropsy. Lower panels show higher magnification of the apical cell area from representative sections. Scale bar, 500 nm. C Quantification of the relative grey levels in the apical area of the distal tubular cells between nucleus and apical cell membrane in the anti-TRPV5-stained immuno- electron microscopic pictures performed in four animals per group with 4–5 images per animal and 4–6 different regions of interest per image. Data information: *P < 0.05 vs. vehicle-treated controls in (A) and (C). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 238 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 7. FGF23-induced increase in TRPV5 protein abundance in the plasma membrane is mediated by a signaling cascade involving ERK1/2 and SGK1. A, B Phosphorylated ERK1/2 and SGK1 (A) and complex glycosylated TRPV5 (B). Protein expression in isolated distal tubular segments from WT mice treated in vitro with rFGF23 alone or in combination with specific SGK1 (iSGK1) and ERK1/2 (iERK1/2) inhibitors. Frame in Western blot image in (A) indicates splicing event. C Phosphorylated SGK1 in isolated distal tubular segments treated in vitro with rFGF23 alone or in combination with a specific ERK1/2 inhibitor (iERK1/2). D Reciprocal immunoprecipitation (IP) of serine-phosphorylated (P-Ser) proteins, followed by Western blot (WB) analysis of WNK4 or vice versa from isolated distal tubular segments treated in vitro with rFGF23 alone or in combination with a specific SGK1 inhibitor (iSGK1). Western blot analysis of WNK4 in renal cortex protein samples immunoprecipitated with WNK4 antibody was used as a loading control. Data information: Actin loading controls are shown for tSGK1 in (A), rFGF23 + iERK1/2 in (B), and tSGK1 in (C). All presented protein expression values were normalized to individual actin levels. Data in (A–C) represent mean  s.e.m. of 4–5 samples each. Data in (D) represent mean  s.e.m. of 3–4 samples each. *P < 0.05 vs. vehicle control (Co). Source data are available online for this figure. ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 239 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 8. FGF23 increases TRPV5 protein abundance and channel activity in MDCK cells transfected with murine TRPV5, SGK1 and WNK4. A Complex glycosylated TRPV5 protein expression in total protein homogenates of MDCK cells transiently transfected with murine TRPV5 (T), SGK1 (S) and WNK4 (W) constructs, alone or in combination, and treated for 12 h with vehicle, rFGF23 or recombinant Klotho (rKlotho) alone or in combination. Mock-transfected cells were used as a negative control (Co). 2+ B Quantification and original images of intracellular Ca levels in MDCK cells transiently transfected with murine TRPV5, SGK1 and WNK4 constructs after treatment with vehicle, rFGF23 or recombinant Klotho (rKlotho) for 12 h. Ruthenium red (10 lM) was used as TRPV inhibitor. 2+ C Time-dependent changes and original images of intracellular Ca levels in MDCK cells transiently transfected with murine TRPV5, SGK1 and WNK4 constructs, and treated at time 0 with rFGF23 (100 ng/ml) or vehicle in the presence or absence of 1.5 mM EGTA added to the culture medium. After 105 min, 1 lM of ruthenium red (RR) was added. MDCK cells were stained with the calcium-sensitive dye Fluo-4 in (B) and (C). Data information: Data in (A) represent mean  s.e.m. of 6–9 samples each from three independent experiments. Data in (B) and (C) represent mean  s.e.m. of 3–4 samples each from three independent experiments. Scale bar, 20 lm in (B), 7 lm in (C). *P < 0.05 vs. vehicle in (A) and (B), and vs. EGTA + rFGF23 in (C), P < 0.05 vs. fluorescence at 105 min in (C). Source data are available online for this figure. increase intracellular Fluo-4 fluorescence (Fig 8C). These data the model shown in Fig 1B which is able to explain the striking / Δ/Δ / Δ/Δ firmly establish that the SGK1-WNK4 pathway mediates the regula- similarities between Kl /VDR and Fgf23 /VDR mice in the tion of TRPV5 expression and activity by FGF23 in renal epithelial regulation of TRPV5. Klotho functions as the co-receptor for blood- cells, and further suggest that soluble Klotho does not influence borne FGF23 in this model, not as a calcium-regulating hormone TRPV5 channel activity in these cells. per se. In analogy to the other major phosphaturic hormone PTH, FGF23 is not only a phosphaturic, but also a calcium-conserving hormone in the kidney, activating, through the FGFR-Klotho complex, Discussion a signaling cascade involving ERK1/2, SGK1, and WNK4. Thus, our data may explain the hitherto unresolved question why injection of Our data have uncovered a previously unknown VDR independent FGF23 does not lower serum calcium despite its suppressive action on function of FGF23 signaling in renal calcium transport. We propose renal synthesis of bioactive vitamin D (Shimada et al, 2001, 2004). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 240 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Although several reports have suggested that soluble Klotho may FGF23 directly acts on proximal tubular cells to down-regulate regulate the membrane abundance and function of ion channels membrane abundance of NaPi-2a through the ERK1/2-SGK1 – + + such as TRPV5, ROMK and NaPi-2a in an FGF23-independent Na /H exchange regulatory cofactor (NHERF)-1 signaling axis manner, and likewise the function of receptors such as the IGF-1 (Andrukhova et al, 2012). In contrast, Farrow and coworkers (Far- receptor (Chang et al, 2005; Kurosu et al, 2005; Hu et al, 2010), row et al, 2009) found that the earliest changes in activation of there is accumulating evidence that Klotho does not have ERK1/2 after injection of FGF23 in vivo in mice occur in the distal FGF23-independent effects on mineral or glucose homeostasis in tubules. It was previously thought that FGF23 acts on the distal vivo. For example, the phenotype of Fgf23 and Klotho double knock- tubule where it generates an unknown endocrine or paracrine sec- / / out mice is indistinguishable from that of Fgf23 and Kl single ondary signal that in turn signals back to the proximal tubule to knockout mice (Nakatani et al, 2009), strongly arguing against downregulate transcellular phosphate transport as part of a “dis- FGF23-independent functions of Klotho. In addition, we previously tal-to-proximal tubular feedback mechanism” (White & Econs, showed that glucose metabolism, bone turnover, and steady-state 2008; Farrow et al, 2010; Martin et al, 2012). Our current study is / Δ/Δ Δ/Δ PTH secretion is unchanged in Kl /VDR compared with VDR able to explain this apparent discrepancy: Circulating FGF23 has mice (Anour et al, 2012), suggesting that Klotho does not have parallel and independent, direct effects on proximal and distal physiologically relevant VDR- and Fgf23-independent functions in renal tubules, suppressing phosphate re-uptake in proximal and the regulation of mineral and glucose homeostasis in vivo. Thus, the stimulating calcium reabsorption in distal tubular cells. In both major physiological role of Klotho appears to be its function as an proximal (Andrukhova et al, 2012) and distal tubular epithelium, obligatory co-receptor for FGF23. However, there is no conclusive ERK1/2 and SGK1 activation are the initial downstream signaling evidence at present to rule out an additional enzymatic function of events after binding of FGF23 to the binary FGFR-Klotho complex. Klotho. Based on our in vitro experiments in isolated distal tubular seg- To address the question of whether or not Klotho has FGF23- ments, FGF23-induced activation of the PI3K pathway may also independent functions on calcium metabolism, we attempted play an additional role in the regulation of transepithelial ion / / D/D breeding 4-week-old Fgf23 /Kl /VDR triple mutant mice on transport, at least in distal tubules. rescue diet in order to compare the phenotype of these mice to that Earlier studies showed that PTH increases transcription (Okano / D/D / D/D of Fgf23 /VDR and Kl /VDR double mutant mice. The et al, 2004), and activates membrane-anchored TRPV5 by protein / / D/D breeding yielded only one Fgf23 /Kl /VDR triple mutant kinase A-mediated phosphorylation (de Groot et al, 2009). Thus, it animal so far. This animal displayed no difference in gross phenotype is interesting to note that in both proximal and distal renal tubules, (body weight, BW 12.5 g) and unchanged blood ionized calcium FGF23 and PTH signaling converge on the same molecules. In the D/D concentration (1.30 mM) as compared to VDR single mutants proximal tubule, we and others showed that PTH and FGF23 signaling 2+ (means  s.d.; BW 13.3  2.1 g, Ca 1.24  0.05 mM, n = 3), as converge on NHERF-1 (Weinman et al, 2011; Andrukhova et al, / D/D 2+ well as Fgf23 /VDR (BW 12.5  1.3 g, Ca 1.24  0.04 mM, 2012) to downregulate NaPi-2a membrane abundance, whereas in / D/D 2+ n = 3) and Kl /VDR (BW 12.1  0.5 g, Ca 1.20  0.13 mM, the distal tubule both signaling pathways converge on TRPV5. In n = 3) double mutant littermates (O. Andrukhova et al, unpub- the model shown in Fig 1B, PTH increases the open probability of lished data). Although obviously still preliminary, these data membrane-anchored TRPV5 by protein kinase A-mediated phos- support the hypothesis that Klotho does not have VDR- and FGF23- phorylation, whereas FGF23 signaling is required for membrane independent actions on calcium metabolism. transport of the channel. Although at present in vivo data supporting In agreement with our finding that protein expression of TRPV5 this hypothesis are lacking, our model would predict that the levels Δ/Δ did not differ between kidneys of WT and VDR mice, earlier of circulating FGF23 may modulate renal sensitivity to the stimulat- studies reported unchanged renal TRPV5 mRNA expression in VDR- ing action of PTH on renal calcium re-absorption. ablated mice (Weber et al, 2001; Okano et al, 2004). However, Our study has established a novel link between FGF23 signal- administration of pharmacological 1,25(OH) D doses to mice or ing and WNK4 activation in distal tubular epithelium. FGF23 stimulation of primary cultures of murine renal tubular cells with signaling resulted in increased phosphorylation of WNK4. The pharmacological doses of 1,25(OH) D directly increases transcrip- immunoelectron microscopic analysis clearly showed that FGF23 tion of the TRPV5 gene in a VDR-dependent fashion (Okano et al, induces intracellular redistribution of WNK4 to the area beneath 2004). Taken together, these data suggest that vitamin D is a phar- the apical cell membrane, where membrane vesicles fuse with the macological, but not a physiological regulator of TRPV5 protein outer cell membrane. Thus, the most important function of WNK4 expression in the murine kidney. Along similar lines, some previous in the regulation of TRPV5 membrane abundance in the DCT in experimental evidence suggested that vitamin D signaling may vivo may be the control of membrane vesicle transport and facili- upregulate renal Klotho mRNA expression (Tsujikawa et al, 2003; tation of vesicle fusion with the apical membrane, rather than the Li et al, 2011). However, our data clearly showed that neither control of clathrin-dependent endocytosis as suggested by some in absent VDR signaling nor pharmacological stimulation with rFGF23 vitro experiments (Cha & Huang, 2010). The function of WNK4 at alters renal Klotho protein expression, suggesting only little physio- focal sites in the basal labyrinth as well as the function of TRPV5 logical regulation by vitamin D and FGF23 at the level of the expressed in the basal labyrinth of distal tubular cells remains co-receptor in vivo. unknown. Because WNK4 is involved in membrane transport of It is well known that FGF23 is primarily a phosphaturic other ion channels such as NCC and ROMK1, FGF23 signaling hormone, suppressing transcellular phosphate transport in proximal may also influence the membrane abundance of other ion chan- tubules. We previously reported that Klotho is expressed in the nels in distal tubules. In this context, it is interesting to note that basolateral membrane of proximal tubular epithelium, and that Kl hypomorphic mice have decreased sodium plasma levels and ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 241 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al increased circulating aldosterone concentrations (Fischer et al, Serum and urine biochemistry 2010), suggesting that Klotho deficiency, and thus reduced FGF23 signaling, may induce alterations in renal sodium handling. WNK Serum creatinine, calcium, and phosphorus as well as urinary crea- kinases act as a complex of at least 3 different kinases, WNK1, 3, tinine, calcium, and phosphorus were analyzed on a Hitachi 912 and 4, to control the intracellular transport of membrane proteins Autoanalyzer (Boehringer Mannheim) or on a Cobas c 111 analyzer (McCormick et al, 2008). In the current report, we focused on (Roche). Serum Klotho protein concentrations were determined WNK4, but it is possible that FGF23 signaling involves other using a mouse specific Klotho ELISA kit (Cusabio) according to the members of the WNK family as well. Clearly, more experimenta- manufacturer’s protocol. The detection limit of the latter assay is tion is required to elucidate all ramifications of the FGF23 signaling 0.8 pg/ml. pathway in renal tubular epithelium. Our finding that FGF23 stimulates renal tubular calcium Immunohistochemistry reabsorption is important for our understanding of the endocrine networks involved in calcium and phosphate homeostasis, and For immunohistochemistry, 5-lm-thick paraffin sections of parafor- suggests that in physiological situations where phosphate excretion maldehyde (PFA)-fixed kidneys were prepared. Some sections were needs to be increased extracellular calcium is still conserved. The stained with hematoxylin/eosin (H & E) by routine methods. pathophysiological downside of this regulation may occur in dis- Before immunofluorescence staining, dewaxed sections were pre- treated for 60 min with blocking solution containing 5% normal eases such as chronic kidney disease, where PTH and FGF23 are chronically elevated due to decreased renal 1,25(OH) D production goat serum in PBS with 0.1% bovine serum albumin and 0.3% and phosphate retention. In this situation, both hormones will stim- Triton X-100. Without rinsing, sections were incubated with poly- ulate renal calcium conservation, possibly contributing to calcium clonal rabbit anti-aKlotho (Alpha Diagnostics, 1:1,000; raised against the cytoplasmic region of aKlotho, detecting the mem- accumulation and vascular calcifications in these hyperphosphatemic patients. brane-bound form), polyclonal rabbit anti-aKlotho (abcam, 1:1,000; raised against a part of KL2, detecting membrane-bound and ecto- domain shed forms), anti-WNK4 (Novus Biologicals, 1:300), anti- Materials and Methods TRPV5 (Alpha Diagnostics, 1:1,000), anti-b-actin (Sigma, 1:5,000), or anti-calbindin D-28k (Swant, 1:1,000) antibodies at 4°C over- Animals night. After washing, sections were incubated for 1.5 h with goat anti-rabbit Alexa 548 and goat anti-mouse Alexa 488 secondary All animal procedures were approved by the Ethical Committee antibodies (Invitrogen, 1:400), respectively. Immunostaining of of the University of Veterinary Medicine Vienna. Heterozygous tissue sections in which either one or both secondary antibodies +/D +/ VDR (Erben et al, 2002) were mated with heterozygous Fgf23 were omitted served as a negative control. The slides were ana- +/ (Sitara et al, 2004), and heterozygous Klotho (Lexicon Genetics, lyzed on a Zeiss LSM 510 Axioplan 2 confocal microscope Mutant Mouse Regional Resource Centers, University of California, equipped with a 63 × oil immersion lens (NA 1.3). By use of the Davis, CA, USA) mutant mice to generate double heterozygous ani- multitrack function, individual fluorochromes were scanned with +/D +/ +/D +/ mals. VDR /Fgf23 and VDR /Klotho mutant mice on laser excitation at 488 and 543 nm separately with appropriate filter D/D C57BL/6 background were interbred to generate WT, VDR , sets to avoid cross talk. Pictures were processed using Adobe / / D/D / D/D Klotho , Fgf23 and compound VDR /Fgf23 and VDR / Photoshop (overlays). Quantification of distal tubular TRPV5 and Klotho mutant mice. Genotyping of the mice was performed by WNK4 co-localization was performed using ImageJ software multiplex PCR using genomic DNA extracted from tail as described (National Institutes of Health, Bethesda, MA, USA) in four animals (Hesse et al, 2007; Anour et al, 2012). The mice were kept at 24°C per group with 3–6 images per animal and 4–6 different regions of with a 12 h/12 h light/dark cycle, and were allowed free access to a interest per image. rescue diet and tap water. The rescue diet (Ssniff, Soest, Germany) containing 2.0% calcium, 1.25% phosphorus, 20% lactose and Immuno-electron microscopic analysis 600 IU vitamin D/kg was fed starting from 16 days of age. This diet has been shown to normalize mineral homeostasis in VDR-ablated For transmission electron microscopy, kidney samples of 4 animals mice (Li et al, 1998; Erben et al, 2002; Zeitz et al, 2003). All experi- per group were fixed in 4% PFA for 24 h at 4°C. Longitudinal ments were performed on male offspring of double heterozygous × 100-lm-thick sections were cut using a Leica VT1000 Vibratome double heterozygous matings. Urine was collected in metabolic (Leica Microsystems). Sections were post-fixed in 4% PFA for 1 h cages before necropsy. Some mice received a single intraperitoneal and rinsed in 0.1 M phosphate buffer 3 × 10 min. Peroxidase injection of vehicle (phosphate-buffered saline (PBS) with 2% activity was inhibited by 30-min incubation with 3% H O in PBS 2 2 DMSO) or recombinant human FGF23 R176/179Q (rFGF23) (Goetz for and nonspecific binding was minimized by incubation for et al, 2007) (10 lg per 3–4-month-old mouse; 5 lg per 9-month-old 60 min in 3% normal goat serum containing 1% BSA. Incubation mouse) or), and were killed 8 or 24 h post-injection. In the mice with anti-TRPV5 (Alpha Diagnostics, 1:150), anti-WNK4 (Novus receiving rFGF23, spontaneous urine was collected before as well as Biologicals, 1:500), anti-aKlotho (Alpha Diagnostics, 1:300), anti- 8 and 24 h post-injection. At necropsy, the mice were exsanguinated aKlotho (abcam, 1:400) or anti-calbindin D-28k (Swant, 1:500) from the abdominal V. cava under anesthesia with ketamine/xyla- was carried out at 4°C overnight. For negative controls, the zine (67/7 mg/kg i.p.) for serum collection. In all experiments 4–9 primary antibody was omitted. Peroxidase-labeled rabbit Power- TM mice were used per experimental group. Vision (ImmunoVision Technologies) secondary system was The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 242 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal employed for antibody detection with subsequent diaminobenzi- described (Hu et al, 2010). Immunoblots were incubated overnight dine (DAB, Sigma) staining. Post-fixation was performed in 1% at 4°C with primary antibodies including polyclonal mouse anti- osmium tetroxide for 2 h at RT followed by dehydration and incu- TRPV5 (1:3,000, Alpha Diagnostics), polyclonal rabbit anti-aKlotho bation in propylene oxide, propylene oxide–epon and subsequent (1:2,000, Alpha Diagnostics, membrane-bound form), polyclonal embedding in pure epon 812. Thin sections were stained with rabbit anti-aKlotho (1:2,500, abcam; membrane-bound and shed lead citrate, and were investigated under a transmission electron forms), polyclonal rabbit anti-WNK4 (1:2,000, Novus Biologicals), microscope (Zeiss EM 900). Apical expression of TRPV5 was polyclonal anti-calbindin D-28k (1:2,000, Swant), polyclonal anti- quantified using Image J software by gray value determination in uromodulin (1:1,500, Sigma), monoclonal mouse anti-b-actin 4–5 images of four animals per group with 4–6 regions of interest (1:5,000, Sigma), monoclonal anti-total-ERK1/2 (BD Biosciences), in each image. anti-phospho-ERK1/2 (Cell Signaling), anti-total-SGK1 (Alpha Diag- nostics), and anti-phospho-SGK1 (Santa Cruz Biotechnology) in Total cell membrane isolation 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]. Mouse kidney cortex tissue was homogenized in a buffer consisting After washing, membranes were incubated with horseradish of 20 mM Tris (pH 7.4/HCl), 5 mM MgCl , 5 mM NaH PO ,1mM peroxidase-conjugated secondary antibodies (Amersham Life 2 2 4 ethylenediamine tetraacetic acid (pH 8.0/NaOH), 80 mM sucrose, Sciences). Specific signal was visualized by ECL kit (Amersham 1 mM phenyl-methylsulfonyl fluoride, 10 lg/ml leupeptin and Life Sciences). The protein bands were quantified by Image Quant 10 lg/ml pepstatin, and subsequently centrifuged for 15 min at 5.0 software (Molecular Dynamics). The expression levels were 4,000 g. Supernatants were transferred to a new tube and centri- normalized to Ponceau S stain and to individual b-actin expression fuged for an additional 30 min at 16,000 g. levels. The glycosylation of the Klotho and TRPV5 proteins was detected on nitrocellulose membranes by GLYCO-PRO assay Isolation of distal tubular segments (Sigma) according to the manufacturer’s protocol. For validation of anti-TRPV5 antibody specificity, total kidney homogenates of Renal distal tubules were isolated as reported previously (Burg et al, 4-week-old TRPV5 male mice (generous gift of Rene J.M. 1997; Schafer et al, 1997; Gonzalez-Mariscal et al, 2000). In brief, Bindels, University Medical Center Nijmegen, Netherlands) were murine kidneys were perfused with sterile culture medium (Ham’s used as a negative control. Each experiment included 4–8 animals F12; GIBCO) containing 1 mg/ml collagenase (type II; Sigma) and per group. All presented Western blot images are from the same 1 mg/ml pronase E (type XXV, Sigma) at pH 7.4 and 37°C. The cor- gel each. All splicing events are indicated by frames in Western tical tissue was dissected in small pieces and placed at 37°C in ster- blot images. ile 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 centri- Co-immunoprecipitation fugation at 3000 rpm for 4 min, the enzyme-containing solution was removed, and tubules were resuspended in ice-cold medium. Kidney cortex homogenate protein samples (1 mg), dissected proxi- Individual distal tubule segments were identified based on morphol- mal tubular segments (40 lg), or homogenate protein samples from ogy in a dissection microscope at x25–40 magnification by their murine smooth muscle cells (1 mg) were incubated with 2 lgof WNK4 (Novus Biologicals), TRPV5 (Lifespan Biosciences), anti- appearance and dimensions. Distal tubular segments were incu- bated with rFGF23 (100 ng/ml) and/or 10 ng/ml of the specific phosphoserine (Alpha Diagnostics), or anti-NaPi-2c (kind gift of SGK1 kinase inhibitor GSK 650394 (Axon Medchem), or 10 ng/ml Drs. Murer and Biber) antibody at 4°C overnight. The immune of the ERK1/2 inhibitor PD184352 (Sigma) for 1, 2 and 4 h. For complexes were captured by adding 50 ll Protein A or G agarose/ sepharose beads (Santa Cruz Biotechnology) and overnight incuba- experiments with FGFR inhibition, distal tubular segments were incubated with rFGF23 (100 ng/ml) and/or 40 ng/ml of the specific tion at 4°C with gentle rocking. The immunoprecipitates were col- FGFR inhibitor PD173074 (Sigma) for 2 h. For experiments with lected by centrifugation at 1,000 g for 5 min at 4°C and washed for SRC, PLC and PI3K inhibition, distal tubular segments were incu- four times in PBS, each time repeating the centrifugation step. After bated with rFGF23 (100 ng/ml) and/or 150 nM Scr inhibitor-1, the final wash, the pellets were suspended in 40 ll of electrophore- 50 nM PI3K inhibitor wortmannin or 10 lM PLC inhibitor U-73122 sis sample buffer and boiled for 2–3 min. Western blot analysis was (all from Sigma) for 2 h. Protein samples were collected for Western performed using a primary anti-TRPV5 or anti-WNK4 antibody. blotting analysis in lysis buffer. Each experiment was performed in duplicates and included 4–8 animals per group. Western Blot RNA isolation and quantitative RT-PCR Kidney cortex homogenate, total cell membrane samples or dis- sected distal tubular segments were solubilized in Laemmli sample Shock-frozen tissues were homogenized in TRI Reagent (Molecu- buffer, fractionated on SDS-PAGE (30 lg/well) and transferred to a lar Research Center) and total RNA was extracted according to nitrocellulose membrane (Thermo Scientific). For detection of the manufacturer’s protocol. RNA purity and quality was deter- Klotho protein in urine, 40 ll of fresh bladder urine, salt precipi- mined using a 2100 Bioanalyzer (Agilent Technologies). One lg tated urine, or 1.3- and 2-fold concentrated urine using Amicon of RNA was used for first-strand cDNA synthesis (iScript cDNA Ultra-0.5 centrifugal filters (Millipore) were loaded per well as Synthesis Kit, Bio-Rad). Quantitative RT-PCR was performed on a ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 243 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al TM TM Rotor-Gene 6000 (Corbett Life Science) using QuantiFast fluorescence intensity and the cell square was calculated for each EverGreen PCR Kit (Qiagen). A melting curve analysis was done cell (16–48 cells per individual sample). for all assays. Primer sequences are available on request. Efficien- cies were examined based on a standard curve. Expression of tar- Statistical analyses get genes was normalized to the expression of the housekeeping gene glyceraldehyde-3-phosphate-dehydrogenase (GAPDH). Statistics were computed using PASW Statistics 17.0 (SPSS Inc., Chicago, IL, USA) The data were analyzed by two-sided t-test (2 In vitro transient transfection experiments groups) or 1-way analysis of variance (ANOVA) followed by Student-Newman-Keuls multiple comparison test (>2 groups). The pCMV6 plasmid construct containing mouse TRPV5 ORF P values of less than 0.05 were considered significant. The data are (MR225560, Origene), mouse SGK1 and mouse WNK4 (generous presented as the mean  s.e.m. gifts of David H. Ellison, Oregon Health and Science University, Port- land, Oregon, USA) were used for in vitro transient transfection Supplementary information for this article is available online: http://emboj.embopress.org experiments. MDCK (Sigma) cells were grown in EMEM (EBSS) with 2 mM glutamine, 1% non-essential amino acids, 10% fetal bovine serum and 100 lg/ml penicillin and 100 lg/ml streptomycin at 37°C Acknowledgements in a 5% CO /95% air humidified atmosphere. 24 h after seeding, We thank C. Bergow and S. Hirmer for excellent technical assistance, Martin cells at ~70% confluence were subjected to transient transfection Glosmann € for help with confocal microscopy, Ingrid Miller and Manfred using X-tremeGENE 9DNA transfection reagent (Roche) according to Gemeiner for help with immunoprecipitation, Magdalena Helmreich and the manufacturer’s protocol. The cells were co-transfected with Waltraud Tschulenk for help with the immuno-electron microscopy, and 0.5 lg of each construct. 24 h post-transfection, the cells were trea- Stefan Handschuh for help with the quantification of immunohistochemistry ted with either vehicle or 100 ng/ml rFGF23 and/or 10 ng/ml and immuno-electron microscopy. Recombinant FGF23 for some of the studies recombinant mouse Klotho (R&D Systems). Complex glycosylated was a generous gift of Vicki Shalhoub, Amgen Inc., Thousand Oaks, CA, USA. TRPV5 expression was monitored by Western blotting in triplicate / Kidney samples of TRPV5 mice were a generous gift of Rene J.M. Bindels, 12 h after cell stimulation. University Medical Center Nijmegen, Netherlands. Mouse SGK1 and WNK4 constructs were a generous gift of David H. Ellison, Oregon Health and Science 2+ Ca imaging and image analysis University, Portland, Oregon, USA. The polyclonal rabbit anti-NaPi-2c antibody was a generous gift of Drs. Jurg Biber and Heini Murer, University of Zurich, The MDCK cells and kidney slices (300-lm-thick) were incubated Switzerland. This work was supported by a grant from the Austrian Science for 30 min at 37°C with 2 lM of the calcium-sensitive dye Fluo-4 Fund (FWF 24186-B21) to R.G.E. O.A. was supported by a postdoctoral fellow- (Molecular Probes), diluted in DMSO (Merck Millipore Interna- ship of the University of Veterinary Medicine Vienna. B.L. was supported by tional) and 20% Pluronic (Merck Millipore International). Thereaf- NIH/NIDDK DK072944, and M.M. was supported by NIH/NIDCR DE13686. ter, the slices or cells were washed two times for 15 min each in PBS. Some kidney slices from WT mice were incubated in vitro for Author contributions 2 h with rFGF23 (100 ng/ml) or vehicle. For calcium visualization OA and RGE conceived and designed the experiments; OA, CS, AS and UZ per- in kidney slices, Fluo-4 was excited by a two-photon laser formed experiments and analyzed the data; OA, RG, MM, BL, and RGE wrote (Ti-sapphire laser, Coherent Inc.) at a wavelength of 820 nm. the manuscript; AS, ME, RG, VS, MM, EEP, and BL provided important tools or Images (512 × 512 pixels) were acquired every 30 s at a depth of techniques; OA, AS, ME, RG, VS, MM, EEP, BL, and RGE discussed and reviewed 50–80 lm. For inhibition of TRPV activity, tissue slices were incu- the manuscript. bated ex vivo with 1–50 lM of ruthenium red (Sigma) at 37°C, 5% CO /95% air humidified atmosphere for 30 min. Fluorescence Conflict of interest images were analyzed using Image J software. The whole epithelial layer of the tubule except for the lumen was selected (by manually VS was an employee of Amgen, Inc. The other authors declare that they have 2+ drawing the region of interest, ROI) to quantify Ca levels in renal no conflict of interest. distal tubules. Fluorescence intensity was quantified in 7–9 ROIs per experimental group from three independent experiments. The ratio between the fluorescence intensity and the ROI was calculated for References each tubule. Fluorescence imaging of cultured cells was performed using an Alexander RT, Woudenberg-Vrenken TE, Buurman J, Dijkman H, van der inverse confocal microscope (TCS SP5 Leica Microsystems), with a Eerden BC, van Leeuwen JP, Bindels RJ, Hoenderop JG (2009) Klotho Leica water immersion objective (63×). Fluo-4 was excited by an prevents renal calcium loss. J Am Soc Nephrol 20: 2371 – 2379 argon laser at 488 nm. Images were acquired every 30 s. Ruthenium Andrukhova O, Zeitz U, Goetz R, Mohammadi M, Lanske B, Erben RG (2012) red (1–10 lM) was used as TRPV inhibitor. In some experiments, FGF23 acts directly on renal proximal tubules to induce phosphaturia 1.5 mM EGTA (Sigma) was added to the culture medium to chelate through activation of the ERK1/2-SGK1 signaling pathway. Bone 51: extracellular calcium. For each MDCK cell, the whole-cell and 621 – 628 nucleus contours were drawn manually. The mean fluorescence Anour R, Andrukhova O, Ritter E, Zeitz U, Erben RG (2012) Klotho lacks a intensity was measured in 4–8 separate cells from each field of view vitamin D independent physiological role in glucose homeostasis, bone (2–7 fields of view per individual sample). The ratio of the turnover, and steady-state PTH secretion in vivo. PLoS ONE 7:e31376 The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 244 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Burg MB, Grantham J, Abramow M, Orloff J, Schafer JA (1997) Preparation Imura A, Tsuji Y, Murata M, Maeda R, Kubota K, Iwano A, Obuse C, Togashi K, and study of fragments of single rabbit nephrons. J Am Soc Nephrol 8: Tominaga M, Kita N, Tomiyama K, Iijima J, Nabeshima Y, Fujioka M, Asato 675 – 683 R, Tanaka S, Kojima K, Ito J, Nozaki K, Hashimoto N et al (2007) Cai H, Cebotaru V, Wang YH, Zhang XM, Cebotaru L, Guggino SE, Guggino alpha-Klotho as a regulator of calcium homeostasis. Science 316: WB (2006) WNK4 kinase regulates surface expression of the human 1615 – 1618 sodium chloride cotransporter in mammalian cells. Kidney Int 69: Jiang Y, Cong P, Williams SR, Zhang W, Na T, Ma HP, Peng JB (2008) WNK4 2162 – 2170 regulates the secretory pathway via which TRPV5 is targeted to the Cha SK, Huang CL (2010) WNK4 kinase stimulates caveola-mediated plasma membrane. Biochem Biophys Res Commun 375: 225 – 229 endocytosis of TRPV5 amplifying the dynamic range of regulation of the Jiang Y, Ferguson WB, Peng JB (2007) WNK4 enhances TRPV5-mediated channel by protein kinase C. J Biol Chem 285: 6604 – 6611 calcium transport: potential role in hypercalciuria of familial hyperkalemic Cha SK, Ortega B, Kurosu H, Rosenblatt KP, Kuro O, Huang CL (2008) hypertension caused by gene mutation of WNK4. Am J Physiol Renal Removal of sialic acid involving Klotho causes cell-surface retention of Physiol 292:F545 – F554 TRPV5 channel via binding to galectin-1. Proc Natl Acad Sci USA 105: Juppner H, Wolf M, Salusky IB (2010) FGF-23: more than a regulator of renal 9805 – 9810 phosphate handling? J Bone Miner Res 25: 2091 – 2097 Chang Q, Hoefs S, Van Der Kemp AW, Topala CN, Bindels RJ, Hoenderop JG Kessler W, Budde T, Gekle M, Fabian A, Schwab A (2008) Activation of cell (2005) The beta-glucuronidase klotho hydrolyzes and activates the TRPV5 migration with fibroblast growth factor-2 requires calcium-sensitive channel. Science 310: 490 – 493 potassium channels. Pflugers Arch 456: 813 – 823 Erben RG, Soegiarto DW, Weber K, Zeitz U, Lieberherr M, Gniadecki R, M€ oller Kuro-o M, Matsumura Y, Aizawa H, Kawaguchi H, Suga T, Utsugi T, Ohyama G, Adamski J, Balling R (2002) Deletion of deoxyribonucleic acid binding Y, Kurabayashi M, Kaname T, Kume E, Iwasaki H, Iida A, Shiraki-Iida T, domain of the vitamin D receptor abrogates genomic and nongenomic Nishikawa S, Nagai R, Nabeshima YI (1997) Mutation of the mouse klotho functions of vitamin D. Mol Endocrinol 16: 1524 – 1537 gene leads to a syndrome resembling ageing. Nature 390: 45 – 51 Farrow EG, Davis SI, Summers LJ, White KE (2009) Initial FGF23-mediated Kurosu H, Ogawa Y, Miyoshi M, Yamamoto M, Nandi A, Rosenblatt KP, Baum signaling occurs in the distal convoluted tubule. J Am Soc Nephrol 20: MG, Schiavi S, Hu MC, Moe OW, Kuro O (2006) Regulation of fibroblast 955 – 960 growth factor-23 signaling by Klotho. J Biol Chem 281: 6120 – 6123 Farrow EG, Summers LJ, Schiavi SC, McCormick JA, Ellison DH, White KE Kurosu H, Yamamoto M, Clark JD, Pastor JV, Nandi A, Gurnani P, McGuinness (2010) Altered renal FGF23-mediated activity involving MAPK and Wnt: OP, Chikuda H, Yamaguchi M, Kawaguchi H, Shimomura I, Takayama Y, effects of the Hyp mutation. J Endocrinol 207: 67 – 75 Herz J, Kahn CR, Rosenblatt KP, Kuro-o M (2005) Suppression of aging in Fischer SS, Kempe DS, Leibrock CB, Rexhepaj R, Siraskar B, Boini KM, mice by the hormone Klotho. Science 309: 1829 – 1833 Ackermann TF, Foller M, Hocher B, Rosenblatt KP, Kuro O, Lang F (2010) Lambers TT, Bindels RJ, Hoenderop JG (2006) Coordinated control of renal Hyperaldosteronism in Klotho-deficient mice. Am J Physiol Renal Physiol Ca2 + handling. Kidney Int 69: 650 – 654 299:F1171 – F1177 Li H, Martin A, David V, Quarles LD (2011) Compound deletion of Fgfr3 and Goetz R, Beenken A, Ibrahimi OA, Kalinina J, Olsen SK, Eliseenkova AV, Xu C, Fgfr4 partially rescues the Hyp mouse phenotype. Am J Physiol Endocrinol Neubert TA, Zhang F, Linhardt RJ, Yu X, White KE, Inagaki T, Kliewer SA, Metab 300:E508 – E517 Yamamoto M, Kurosu H, Ogawa Y, Kuro-o M, Lanske B, Razzaque MS et al Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB (2007) Molecular insights into the klotho-dependent, endocrine mode of (1998) Normalization of mineral ion homeostasis by dietary means action of fibroblast growth factor 19 subfamily members. Mol Cell Biol 27: prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia 3417 – 3428 in vitamin D receptor-ablated mice. Endocrinology 139: 4391 – 4396 Gonzalez-Mariscal L, Namorado MC, Martin D, Luna J, Alarcon L, Islas S, Martin A, David V, Quarles LD (2012) Regulation and function of the FGF23/ Valencia L, Muriel P, Ponce L, Reyes JL (2000) Tight junction proteins klotho endocrine pathways. Physiol Rev 92: 131 – 155 ZO-1, ZO-2, and occludin along isolated renal tubules. Kidney Int 57: Matsumura Y, Aizawa H, Shiraki-Iida T, Nagai R, Kuro-o M, Nabeshima Y 2386 – 2402 (1998) Identification of the human klotho gene and its two transcripts de Groot T, Lee K, Langeslag M, Xi Q, Jalink K, Bindels RJ, Hoenderop JG (2009) encoding membrane and secreted klotho protein. Biochem Biophys Res Parathyroid hormone activates TRPV5 via PKA-dependent phosphorylation. Commun 242: 626 – 630 J Am Soc Nephrol 20: 1693 – 1704 Mayan H, Vered I, Mouallem M, Tzadok-Witkon M, Pauzner R, Farfel Z (2002) He G, Wang HR, Huang SK, Huang CL (2007) Intersectin links WNK kinases to Pseudohypoaldosteronism type II: marked sensitivity to thiazides, endocytosis of ROMK1. J Clin Invest 117: 1078 – 1087 hypercalciuria, normomagnesemia, and low bone mineral density. J Clin Hellwig N, Albrecht N, Harteneck C, Schultz G, Schaefer M (2005) Homo- and Endocrinol Metab 87: 3248 – 3254 heteromeric assembly of TRPV channel subunits. J Cell Sci 118: 917 – 928 McCormick JA, Yang CL, Ellison DH (2008) WNK kinases and renal sodium Hesse M, Frohlich LF, Zeitz U, Lanske B, Erben RG (2007) Ablation of vitamin transport in health and disease: an integrated view. Hypertension 51: D signaling rescues bone, mineral, and glucose homeostasis in Fgf-23 588 – 596 deficient mice. Matrix Biol 26: 75 – 84 Michlig S, Mercier A, Doucet A, Schild L, Horisberger JD, Rossier BC, Firsov D Hoenderop JG, Vennekens R, Muller D, Prenen J, Droogmans G, Bindels RJ, (2004) ERK1/2 controls Na, K-ATPase activity and transepithelial sodium Nilius B (2001) Function and expression of the epithelial Ca(2+) channel transport in the principal cell of the cortical collecting duct of the mouse family: comparison of mammalian ECaC1 and 2. J Physiol 537: 747 – 761 kidney. J Biol Chem 279: 51002 – 51012 Hu MC, Shi M, Zhang J, Pastor J, Nakatani T, Lanske B, Razzaque MS, Nakatani T, Sarraj B, Ohnishi M, Densmore MJ, Taguchi T, Goetz R, Rosenblatt KP, Baum MG, Kuro-o M, Moe OW (2010) Klotho: a novel Mohammadi M, Lanske B, Razzaque MS (2009) In vivo genetic evidence phosphaturic substance acting as an autocrine enzyme in the renal for klotho-dependent, fibroblast growth factor 23 (Fgf23) -mediated proximal tubule. FASEB J 24: 3438 – 3450 regulation of systemic phosphate homeostasis. FASEB J 23: 433 – 441 ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 245 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Okano T, Tsugawa N, Morishita A, Kato S (2004) Regulation of gene factor-23 results in hyperphosphatemia and impaired skeletogenesis, and expression of epithelial calcium channels in intestine and kidney of mice reverses hypophosphatemia in Phex-deficient mice. Matrix Biol 23: by 1alpha,25-dihydroxyvitamin D3. J Steroid Biochem Mol Biol 89–90: 421 – 432 335 – 338 Streicher C, Zeitz U, Andrukhova O, Rupprecht A, Pohl E, Larsson TE, Ring AM, Leng Q, Rinehart J, Wilson FH, Kahle KT, Hebert SC, Lifton RP (2007) Windisch W, Lanske B, Erben RG (2012) Long-term Fgf23 deficiency An SGK1 site in WNK4 regulates Na+ channel and K+ channel activity does not influence aging, glucose homeostasis, or fat metabolism in and has implications for aldosterone signaling and K+ homeostasis. Proc mice with a nonfunctioning vitamin D receptor. Endocrinology 153: Natl Acad Sci USA 104: 4025 – 4029 1795 – 1805 Sandulache D, Grahammer F, Artunc F, Henke G, Hussain A, Nasir O, Mack A, Tohyama O, Imura A, Iwano A, Freund JN, Henrissat B, Fujimori T, Nabeshima Friedrich B, Vallon V, Wulff P, Kuhl D, Palmada M, Lang F (2006) Y(2004) Klotho is a novel beta-glucuronidase capable of hydrolyzing Renal Ca2 + handling in sgk1 knockout mice. Pflugers Arch 452: steroid beta-glucuronides. J Biol Chem 279: 9777 – 9784 444 – 452 Tsujikawa H, Kurotaki Y, Fujimori T, Fukuda K, Nabeshima Y (2003) Klotho, a Schafer JA, Watkins ML, Li L, Herter P, Haxelmans S, Schlatter E (1997)A gene related to a syndrome resembling human premature aging, simplified method for isolation of large numbers of defined nephron functions in a negative regulatory circuit of vitamin D endocrine system. segments. Am J Physiol 273:F650 – F657 Mol Endocrinol 17: 2393 – 2403 Segawa H, Yamanaka S, Ohno Y, Onitsuka A, Shiozawa K, Aranami F, Furutani Urakawa I, Yamazaki Y, Shimada T, Iijima K, Hasegawa H, Okawa K, Fujita T, J, Tomoe Y, Ito M, Kuwahata M, Imura A, Nabeshima Y, Miyamoto K Fukumoto S, Yamashita T (2006) Klotho converts canonical FGF receptor (2007) Correlation between hyperphosphatemia and type II Na-Pi into a specific receptor for FGF23. Nature 444: 770 – 774 cotransporter activity in klotho mice. Am J Physiol Renal Physiol 292: Weber K, Erben RG, Rump A, Adamski J (2001) Gene structure and regulation F769 – F779 of the murine epithelial calcium channels ECaC1 and 2. Biochem Biophys Shimada T, Hasegawa H, Yamazaki Y, Muto T, Hino R, Takeuchi Y, Fujita T, Res Commun 289: 1287 – 1294 Nakahara K, Fukumoto S, Yamashita T (2004) FGF-23 is a potent regulator Weinman EJ, Steplock D, Shenolikar S, Biswas R (2011) FGF-23-mediated of vitamin D metabolism and phosphate homeostasis. J Bone Miner Res inhibition of renal phosphate transport in mice requires NHERF-1 and 19: 429 – 435 synergizes with PTH. J Biol Chem 286: 37216 – 37221 Shimada T, Mizutani S, Muto T, Yoneya T, Hino R, Takeda S, Takeuchi Y, White KE, Econs MJ (2008) Fibroblast Growth Factor-23.In Primer on the Fujita T, Fukumoto S, Yamashita T (2001) Cloning and characterization of Metabolic Bone Diseases and Disorders of Mineral Metabolism, Rosen CJ FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl (ed), pp 112 – 116. Washington, DC: Society of Bone and Mineral Acad Sci USA 98: 6500 – 6505 Research Shiraki-Iida T, Aizawa H, Matsumura Y, Sekine S, Iida A, Anazawa H, Nagai R, Woudenberg-Vrenken TE, Bindels RJ, Hoenderop JG (2009) The role of Kuro-o M, Nabeshima Y (1998) Structure of the mouse klotho gene and its transient receptor potential channels in kidney disease. Nat Rev Nephrol 5: two transcripts encoding membrane and secreted protein. FEBS Lett 424: 441 – 449 6 – 10 Zeitz U, Weber K, Soegiarto DW, Wolf E, Balling R, Erben RG (2003) Impaired Sitara D, Razzaque MS, Hesse M, Yoganathan S, Taguchi T, Erben RG, insulin secretory capacity in mice lacking a functional vitamin D receptor. Juppner H, Lanske B (2004) Homozygous ablation of fibroblast growth FASEB J 17: 509 – 511 The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

FGF23 promotes renal calcium reabsorption through the TRPV5 channel

Download PDF
18 pages

Loading next page...
 
/lp/springer-journals/fgf23-promotes-renal-calcium-reabsorption-through-the-trpv5-channel-m0iH9GExA2

References (67)

Publisher
Springer Journals
Copyright
Copyright © The Authors. 2014
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/embj.201284188
Publisher site
See Article on Publisher Site

Abstract

Article FGF23 promotes renal calcium reabsorption through the TRPV5 channel 1 1 1 1 1 Olena Andrukhova , Alina Smorodchenko , Monika Egerbacher , Carmen Streicher , Ute Zeitz , 2 3 2 1 4 Regina Goetz , Victoria Shalhoub , Moosa Mohammadi , Elena E Pohl , Beate Lanske & 1,* Reinhold G Erben Abstract enzymes, and released into the blood circulation (Imura et al, 2007). In addition, a soluble Klotho isoform can be produced by alternative aKlotho is thought to activate the epithelial calcium channel Tran- splicing of the Klotho mRNA (Matsumura et al, 1998; Shiraki-Iida sient Receptor Potential Vanilloid-5 (TRPV5) in distal renal tubules et al, 1998). This truncated Klotho gene product lacks exons 4 and 5 through its putative glucuronidase/sialidase activity, thereby pre- in mice (Shiraki-Iida et al, 1998). Klotho is mainly expressed in venting renal calcium loss. However, aKlotho also functions as the renal distal convoluted tubules (DCT) and in the brain choroid obligatory co-receptor for fibroblast growth factor-23 (FGF23), a plexus (Kuro-o et al, 1997). However, Klotho expression is also bone-derived phosphaturic hormone. Here, we show that renal found at other locations, for example in renal proximal tubules calcium reabsorption and renal membrane abundance of TRPV5 (Hu et al, 2010; Andrukhova et al, 2012) or parathyroid glands are reduced in Fgf23 knockout mice, similar to what is seen in (Shiraki-Iida et al, 1998; Urakawa et al, 2006; Imura et al, 2007). aKlotho knockout mice. We further demonstrate that aKlotho The molecular function of Klotho is still controversial. Klotho was neither co-localizes with TRPV5 nor is regulated by FGF23. Rather, initially thought to be an anti-aging factor (Kuro-o et al, 1997). Later apical membrane abundance of TRPV5 in renal distal tubules and studies suggested that soluble Klotho may have the ability to alter thus renal calcium reabsorption are regulated by FGF23, which the function and abundance of membrane glycoproteins by remov- binds the FGF receptor-aKlotho complex and activates a signaling ing sialic acid or other terminal sugars from sugar chains through a cascade involving ERK1/2, SGK1, and WNK4. Our data thereby putative glycosidase activity (Chang et al, 2005; Kurosu et al, 2005; Hu et al, 2010). identify FGF23, not aKlotho, as a calcium-conserving hormone in the kidney. Apart from its putative enzymatic function, membrane-bound Klotho also functions as the co-receptor for the bone-derived hor- Keywords calcium homeostasis; fibroblast growth factor-23; Klotho; renal mone fibroblast growth factor-23 (FGF23). Binding of FGF23 to ubiquitously expressed FGF receptors (FGFR) requires Klotho as an calcium reabsorption; transient receptor potential vanilloid-5 Subject Categories Membrane & Intracellular Transport; Signal Transduction obligatory co-receptor (Kurosu et al, 2006; Urakawa et al, 2006), DOI 10.1002/embj.201284188 | Received 8 December 2012 | Revised 30 restricting the hormonal action of FGF23 to Klotho-expressing October 2013 | Accepted 18 November 2013 tissues. FGF23 down-regulates renal proximal tubular phosphate EMBO Journal (2014) 33, 229–246 reuptake and 1a-hydroxylase expression. 1a-hydroxylase is the key enzyme for production of the biologically active vitamin D hor- mone, 1a,25-dihydroxyvitamin D (1,25(OH) D). Because FGF23 is 3 2 Introduction secreted by osteocytes and osteoblasts in bone in response to ele- vated phosphate and vitamin D, this hormone forms a negative aKlotho (Klotho) is a single pass transmembrane protein, which feedback loop between bone and kidney (Juppner et al, 2010). shares sequence homology with family I b-glycosidases, including It was reported by Chang and colleagues (Chang et al, 2005) that soluble Klotho is a regulator of the epithelial calcium channel tran- b-glucuronidases (Kuro-o et al, 1997). However, Klotho lacks essen- tial active site glutamic acid residues typical for this family of glyco- sient receptor potential vannilloid-5 (TRPV5), a glycoprotein essential sidases (Tohyama et al, 2004). The extracellular domain of Klotho, for entry of calcium in calcium-transporting renal epithelial cells. which consists of two type I b-glycosidase domains (KL1 and KL2), Apical membrane expression of fully glycosylated TRPV5 is the rate- limiting step in distal renal tubular transcellular calcium transport can be shed from the cell surface by membrane-anchored proteolytic 1 University of Veterinary Medicine Vienna, Vienna, Austria 2 New York University School of Medicine, New York, NY, USA 3 Amgen Inc., Thousand Oaks, CA, USA 4 Harvard School of Dental Medicine, Boston, MA, USA *Corresponding author. Tel: +43 1 250 77 4550; Fax: +43 1 250 77 4599; E-mail: [email protected] ª 2014 The Authors This is an open access article under the terms of the Creative Commons Attribution License, The EMBO Journal Vol 33 |No 3 | 2014 which permits use, distribution and reproduction in any medium, provided the original work is properly cited. The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al / / which also involves the calcium-binding proteins calbindin D9k and on renal 1a-hydroxylase activity, Kl and Fgf23 mice produce D28k (CalD28k) for intracellular transport, and the sodium-calcium excessive amounts of 1,25(OH) D, and die early from the sequelae exchanger (NaCX) and the plasma membrane calcium ATPase 1b of hypervitaminosis D, hypercalcemia and hyperphosphatemia. The (PMCA1b) for basolateral extrusion of calcium (Lambers et al, 2006). pivotal role of hypervitaminosis D for the development of the pre- / / The current model for TRPV5 regulation by Klotho is that soluble Klo- mature aging phenotype in Kl and Fgf23 mice is underscored tho promotes insertion of TRPV5 in the apical membrane of distal by the fact that ablation of vitamin D signaling completely rescues tubular cells through its putative glycosidase activity, thereby stabiliz- the premature aging phenotype in these mice (Hesse et al, 2007; ing the interaction between glycosylated TRPV5 and membrane- Anour et al, 2012; Streicher et al, 2012). bound galectin (Chang et al, 2005; Cha et al, 2008) (Fig 1A). A major In order to examine vitamin D independent effects of Klotho and explanatory gap in this model is how soluble Klotho crosses from Fgf23 deficiency on renal calcium excretion in skeletally mature the systemic circulation into the urinary space. It was reported that mice, we crossed mice with a non-functioning vitamin D receptor Δ/Δ / / the 130 kDa isoform of Klotho protein can be detected in the urine of (VDR ) with Kl and Fgf23 mice, and analyzed them at mice (Chang et al, 2005; Hu et al, 2010). The soluble forms of mur- 9 months of age. All mice were kept life-long on a so-called rescue ine Klotho are large proteins with molecular weights of about 130 diet rich in calcium, phosphorus, and lactose. The rescue diet and 65 kDa for the shed and alternatively spliced isoforms, respec- normalizes the calcium absorption defect of VDR mutant mice in Δ/Δ / Δ/Δ / tively (Shiraki-Iida et al, 1998; Imura et al, 2007). Because the glo- the gut, so that VDR mice, but also Kl /VDR and Fgf23 / Δ/Δ merular filter is impermeable to proteins larger than approximately VDR mice on this diet are normocalcemic (Erben et al, 2002; 60–65 kDa under physiological conditions, soluble Klotho in plasma Hesse et al, 2007; Anour et al, 2012). In this study, we show that / Δ/Δ / Δ/Δ cannot be filtered. Alternatively, Klotho could be secreted by distal skeletally mature Kl /VDR and Fgf23 /VDR mice are char- tubular cells into the urinary space, or shed from the apical cell acterized by an almost identical renal calcium wasting phenotype, membrane. However, experimental evidence for tubular secretion or and that FGF23 is a regulator of distal tubular TRPV5 membrane tubular apical membrane shedding of Klotho is lacking thus far. abundance and renal calcium reabsorption through an intracellular In accordance with an important role of Klotho in the regulation signaling cascade involving ERK1/2, SGK1, and WNK4. of distal renal tubular TRPV5 activity, Klotho null (Kl ) mice on a vitamin D deficient diet lose urinary calcium despite unchanged Results renal expression of fully glycosylated TRPV5 (Alexander et al, 2009). However, this animal model is complicated by the fact that Kl mice have elevated 1,25(OH) D blood levels. 1,25(OH) Dis We first examined renal calcium excretion in skeletally mature, 2 2 Δ/Δ / Δ/Δ 9-month-old wild-type (WT), VDR , and Kl /VDR compound thought to be involved in the regulation of distal tubular TRPV5 (Lambers et al, 2006). Due to the lacking suppressive effect of Fgf23 mutant mice on rescue diet. We found pronounced loss of urinary Figure 1. Models of the regulation of apical membrane TRPV5 in renal distal tubules by Klotho and FGF23. A Model based on previous studies for the regulation of apical membrane TRPV5 by secreted Klotho. TRPV5 is necessary for apical entry of calcium, which is then transported through the cell bound to calbindin D9k and D28k, and extruded at the basolateral side via PMCA1 and NCX. Secreted Klotho is thought to specifically hydrolyze sugar residues from the glycan chains on TRPV5 which in turn stabilizes TRPV5 in the membrane through interaction of the sugar residues with extracellular galectin (Chang et al, 2005; Cha et al, 2008). The cellular secretion process of Klotho in this model is unclear. Adapted from Cha et al (2008). B Our proposed model of Fgf23-aKlotho signaling in renal distal tubular cells. Fgf23 binds to the basolateral FGFR1c-Klotho complex and activates ERK1/2 leading to SGK1 phosphorylation. SGK1 in turn activates WNK4, stimulating WNK4-TRPV5 complex formation, and increasing intracellular transport of fully glycosylated TRPV5 from the Golgi apparatus to the plasma membrane. PTH signaling activates membrane-anchored TRPV5 by protein kinase A (PKA)-mediated phosphorylation. The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 230 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 2. Renal calcium reabsorption and TRPV5 plasma membrane abundance in Fgf23 and Klotho deficient mouse models. A–D Urinary excretion of calcium corrected for creatinine (UrCa/Crea) (A), Western blotting quantification of core (75 kDa) and complex (92 kDa) glycosylated TRPV5 protein expression in renal cortical total membrane fractions (B), and Western blot analysis of membrane-bound Klotho in renal total protein extracts (C) in Δ/Δ / Δ/Δ / Δ/Δ 9-month-old male WT, VDR , Kl /VDR , and Fgf23 /VDR on rescue diet. Antibody specificity and glycosylation of TRPV5 and Klotho was controlled by a glycosylation assay (pink staining) followed by Western blot analysis of TRPV5 and Klotho, respectively (D). Anti-Klotho antibodies detecting the membrane- bound (anti-cytoplasmic domain, upper panel) or the membrane-bound and ectodomain shed (anti KL2 domain, lower panel) forms of the protein were used. / Δ/Δ Protein extracts from lung and spleen, as well as kidney extracts from Kl /VDR mice served as negative controls for TRPV5 and Klotho protein expression, respectively. # D/D Data information: *P < 0.05 vs. WT, P < 0.05 vs. VDR mice. Only the 135 kDa transmembrane isoform of Klotho was quantified in (C). Data in (A–C) represent mean  s.e.m. of 4–9 animals each. Frames in Western blot images shown in (B) and (D) indicate splicing events. Source data are available online for this figure. ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 231 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al / Δ/Δ calcium in 9-month-old Kl /VDR mice (Fig 2A), which was enzymatically alter the glycosylation pattern of TRPV5 in the apical associated with a 50% downregulation of membrane expression of cell membrane (Chang et al, 2005). Co-localization of Klotho and Δ/Δ complex glycosylated TRPV5 relative to VDR mice (Fig 2B). It is TRPV5 in distal renal tubular cells has been demonstrated by Chang known from previous studies that VDR mutant mice on rescue diet et al (Chang et al, 2005). In our immunohistochemical analysis, show a defect in renal tubular reabsorption of calcium due to lower TRPV5 staining was exclusively seen in distal tubules also express- expression of calbindin D9k (Erben et al, 2002). The core and com- ing calbindin D28k, further demonstrating the specificity of the anti- plex glycosylated forms of TRPV5 protein were unchanged in VDR body used (Supplementary Fig S2A). The membrane-bound form of single mutant mice (Fig 2B), showing that vitamin D signaling is Klotho was mainly found in the cytoplasm and the basolateral not essential for the regulation of TRPV5. However, urinary loss of membrane, whereas TRPV5 was mainly localized in the apical Δ/Δ / Δ/Δ calcium was almost twice as high in compound mutant mice com- membrane in WT, VDR , and Fgf23 /VDR mice (Fig 3A). We pared with VDR single mutants, i.e., lack of Klotho aggravated the observed an identical subcellular distribution of Klotho in distal renal calcium wasting seen in VDR single mutants (Fig 2A). This tubular epithelium, employing an anti-Klotho antibody detecting finding corroborates earlier reports that Klotho has an essential role both the membrane-bound and the ectodomain shed form of the in the regulation of renal TRPV5 activity (Chang et al, 2005), and protein (Supplementary Fig S2B). Some TRPV5 staining was also further shows that this mechanism is VDR independent. Notably, seen basolaterally in all genotypes (Fig 3A). Co-localization of / Δ/Δ / Δ/Δ similar to Kl /VDR mice, 9-month-old Fgf23 /VDR mice Klotho and TRPV5, however, was almost absent, and only seen in also showed renal calcium wasting and reduced membrane expres- some cytoplasmic or basolateral areas of the distal tubular cells sion of TRPV5 (Fig 2A and B). Indeed, the absence of Fgf23 resulted (Fig 3A and Supplementary Fig S2). In analogy to the immunoblot- in a stronger downregulation of core and complex glycosylated ting data (Fig 2B), membrane expression of TRPV5 was clearly / Δ/Δ TRPV5 compared with the absence of Klotho (Fig 2B). Using anti- reduced in distal tubules of Fgf23 /VDR mice (Fig 3A). To Klotho antibodies raised against the short intracellular region of the assess the subcellular localization of Klotho in more detail, we per- membrane-bound Klotho isoform or against the extracellular KL2 formed immuno-electron microscopic analyses in renal tissue from domain, we found renal Klotho protein expression unchanged in WT mice, using anti-Klotho antibodies detecting either the trans- Δ/Δ / Δ/Δ both VDR single and Fgf23 /VDR compound mutants membrane or both the transmembrane and the ectodomain shed (Fig 2C and Supplementary Fig S1A). Although the anti-TRPV5 and forms of the protein. Both antibodies showed the presence of Klotho anti-Klotho antibodies we used for immunoblotting and immuno- protein in the membrane of the basal labyrinth, but staining was histochemistry have been successfully employed by other groups absent in the apical membrane of distal tubular cells (Fig 3B). Kid- (Sandulache et al, 2006; Segawa et al, 2007), we confirmed the neys from Kl mice were used as a negative control (Fig 3B). specificity of the antibodies, using tissue extracts from mouse lung To examine whether FGF23 was able to regulate TRPV5 in gain- Δ/Δ / Δ/Δ and spleen where TRPV5 is not expressed (Fig 2D), extracts from of-function models, we treated WT, VDR ,and Kl /VDR mice kidneys of TRPV5 mice (Supplementary Fig S1B), and extracts with recombinant FGF23 (rFGF23). Within 8 h post-injection, rFGF23 / Δ/Δ from kidneys of Kl /VDR mice lacking all isoforms of Klotho reduced urinary calcium excretion to very low levels in WT mice (Fig 2C and D) as negative controls. In addition, we confirmed the (Fig 4A). Serum calcium and serum parathyroid hormone (PTH) glycosylation of the 92 kDa TRPV5 band and of the 135 kDa trans- remained unchanged (Supplementary Fig S3A and Fig 4B). The membrane form of Klotho (Fig 2D). Both anti-Klotho antibodies rFGF23-induced reduction in urinary calcium excretion occurred in Δ/Δ / Δ/Δ detected the 135 kDa glycosylated full-length transmembrane form WT and VDR , but not in Kl /VDR mice (Fig 4C), and was as the predominant form of Klotho protein in renal extracts (Fig 2D). associated with distinctly upregulated TRPV5 expression in the distal Δ/Δ / The 130 kDa band which might represent the ectodomain shed form tubular apical membrane (Fig 4D) of WT and VDR , but not of Kl Δ/Δ of Klotho was also glycosylated (Fig 2D). The 65 kDa Klotho band /VDR mice. These data demonstrate that the FGF23-induced detected by the anti-transmembrane antibody was not glycosylated, upregulation of TRPV5 and the concomitant increase in renal tubular whereas the 55 kDa Klotho band detected by the anti-KL2 antibody calcium reabsorption are Klotho dependent. Despite the profound was glycosylated (Fig 2D). Therefore, the nature of these two puta- changes in TRPV5 expression and renal calcium reabsorption in WT Δ/Δ tive fragments of Klotho is not clear. Using the methodology and VDR mice, expression of membrane-bound and shed Klotho described by Hu and coworkers (Hu et al, 2010), we were unable to protein remained unchanged (Fig 4E and Supplementary Fig S3B), detect any isoform of Klotho protein in native, salt-precipitated, or and Klotho did not show co-localization with TRPV5 in the distal concentrated urine of WT mice by Western blotting (Fig 2D and Sup- tubular apical membrane (Fig 4D and Supplementary Fig S3C). / Δ/Δ plementary Fig S1C). The reason for this discrepancy is not clear, but Immunohistochemical Klotho staining was absent in Kl /VDR may be due to differences in the anti-Klotho antibodies used. mice (Fig 4D and Supplementary Fig S3C), confirming the specificity The fact that Klotho deficiency and Fgf23 deficiency have almost of both anti-Klotho antibodies used. Moreover, soluble Klotho in identical effects on renal TRPV5 is difficult to explain on the basis of serum remained below the detection limit of an immunoassay spe- the model shown in Fig 1A. Rather, this finding points to an essen- cific for murine Klotho (data not shown) and anti-Klotho immuno- tial role of Fgf23 in the regulation of TRPV5. We reported earlier blotting of urine was negative in rFGF23-treated WT mice that renal function and morphology of kidneys is normal in (Supplementary Fig S3D). Finally, we treated isolated distal tubular Δ/Δ / Δ/Δ 9-month-old VDR and Fgf23 /VDR mice (Streicher et al, segments in vitro with rFGF23 in the presence and absence of a FGFR 2012), ruling out renal functional impairment as a confounding inhibitor. The FGF23-induced upregulation of complex glycosylated factor in these experiments. A central element of the model shown TRPV5 expression was completely blunted in the presence of the in Fig 1A is a direct, albeit transient, protein-protein interaction FGFR inhibitor, showing that FGF23 signals through the FGFR to between TRPV5 and Klotho, because soluble Klotho is thought to increase distal tubular TRPV5 membrane expression (Fig 4F). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 232 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 3. Membrane-bound aKlotho and TRPV5 do not co-localize in renal distal tubule cells. A Immunohistochemical co-staining with anti-transmembrane aKlotho (red or green) antibody, anti-TRPV5 (green), anti-b-actin (red), and DAPI (blue) of paraffin Δ/Δ / Δ/Δ sections from kidneys of 9-month-old WT, VDR , and Fgf23 /VDR mice on rescue diet. Right panels show H&E-stained paraffin sections for comparison of subcellular localization. Scale bar, 20 lm. B Immuno-electron microscopic staining using anti-Klotho antibodies against transmembrane, and transmembrane and shed forms in kidneys of WT (left panels) and Klotho (right panels) mice. Upper panels show the apical cell area, lower panels show the basolateral area with the basal labyrinth. Arrows mark the apical cell membrane, asterisks mark mitochondria, and the symbol # marks the nucleus. Scale bar, 500 nm. To confirm the functional role of the FGF23-induced upregulation changes in fluorescence over time (30 min) after addition of ruthe- of TRPV5 in the apical membrane of distal tubular epithelium, we nium red in kidney slices from rFGF23- and vehicle-treated mice, performed intracellular calcium imaging, employing 2-photon respectively. For distal tubular calcium reabsorption, only TRPV5 microscopy of Fluo-4-loaded, 300-lm-thick renal slices prepared and 6 are thought to be relevant (Woudenberg-Vrenken et al, 2009). from vehicle- and rFGF23-treated WT mice, 8 h post-injection. TRPV5 is about 100-fold more sensitive to ruthenium red than Distal tubules of FGF23-treated mice showed a 5-fold increase in TRPV6 (Hoenderop et al, 2001). Therefore, ruthenium red can be fluorescence intensity, relative to those of vehicle-treated mice used to discriminate the channel function of TRPV5 from that of (Fig 4G). The FGF23-induced increase in fluorescence intensity was TRPV6. To further demonstrate the regulation of TRPV5 activity by largely abrogated by ex vivo addition of 10 lM of the TRPV inhibitor FGF23, we treated Fluo-4-loaded kidney slices from WT mice with ruthenium red (Fig 4G). Supplementary videos 1 and 2 show the rFGF23 or vehicle in vitro. rFGF23 gradually increased intracellular ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 233 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al fluorescence over 2 h in distal tubules (Fig 4H and Supplementary (Ring et al, 2007). WNK4-mediated regulation of these channels video 3). The latter effect was reversed by ruthenium red at concen- involves physical association between the channel protein and trations as low as 1 lM (Fig 4H and Supplementary video 4). WNK4 (Cai et al, 2006; He et al, 2007). Interestingly, mutations in At 1 lM, ruthenium red does not antagonize TRPV6 function WNK4 can lead to urinary calcium loss in humans (Mayan et al, (Hoenderop et al, 2001). 10 and 50 lM of ruthenium red did not 2002). In addition, studies in transfected Xenopus laevis oocytes further decrease the fluorescence signal, showing that TRPV6 does have suggested that WNK4 influences the membrane abundance of not play a major role in the FGF23-induced increase in calcium TRPV5 (Jiang et al, 2007, 2008). uptake in distal tubular cells in this experimental setting (Fig 4H Based on this knowledge, we hypothesized that FGF23 may regu- and Supplementary video 4). The discrepancy between the complete late TRPV5 through a signaling cascade involving ERK1/2, SGK1, and reversal of the rFGF23-induced increase in intracellular Fluo-4 fluo- WNK4. To test this, we first examined whether activation of ERK1/2 rescence by ruthenium red in Fig 4H and the partial reversal and SGK1 was regulated by FGF23 in loss- and gain-of-function mod- observed in Fig 4G can probably be explained by the fact that in the els. Phospho-ERK1/2 and phospho-SGK1 were decreased in the kid- / Δ/Δ ex vivo experiment shown in Fig 4G distal tubular cells were pre- ney of Fgf23 /VDR compound mutant mice (Supplementary Fig Δ/Δ / exposed to rFGF23 for 8 h, whereas rFGF23 was added at time 0 in S4A and B). rFGF23 administration to WT, VDR ,and Fgf23 / Δ/Δ the in vitro experiment shown in Fig 4H. The latter experimental VDR compound mutant mice led to an increase in renal complex setting therefore looks at very early aspects of the rFGF23-induced glycosylated TRPV5 (Fig 5A), phospho-ERK1/2 (Fig 5B), and phos- increase in distal tubular calcium uptake. This is also reflected in pho-SGK1 (Fig 5C). These findings suggest that FGF23 signaling leads the almost 2-fold higher rFGF23-induced increase in relative fluores- to activation of renal SGK1 and increased membrane abundance of cence in the ex vivo (Fig 4G) compared with the in vitro experiment complex glycosylated TRPV5 in a VDR independent fashion. (Fig 4H). Collectively, these data show that the FGF23-induced Next, we examined whether WNK4 is a downstream mediator of upregulation of TRPV5 results in increased calcium uptake in distal FGF23 signaling. Reciprocal immunoprecipitation experiments on tubular epithelial cells, and that the FGF23-induced stimulation of renal protein extracts, using anti-phosphoserine and anti-WNK4 apical calcium entry is mainly mediated by TRPV5. However, antibodies showed that serine phosphorylation of WNK4 was because ruthenium red at 10 lM did not completely reverse the increased 8 h after treatment of WT mice with rFGF23 (Fig 5E). Fur- rFGF23-induced increase in intracellular Fluo-4 fluorescence in dis- ther, reciprocal immunoprecipitation revealed a physical association tal tubules (Fig 4G) and TRPV5 and TRPV6 can form heteromulti- between fully glycosylated TRPV5 and WNK4 in the kidney of WT, Δ/Δ / Δ/Δ mers (Hellwig et al, 2005) with intermediate ruthenium red VDR and Fgf23 /VDR mice (Fig 5F). Whereas total WNK4 sensitivity, we cannot exclude a potential contribution of TRPV6 to protein abundance was not different between the genotypes the effects of FGF23 on renal tubular calcium reabsorption in vivo. (Fig 5D), the amount of TRPV5-WNK4 complexes in kidneys of / Δ/Δ The intracellular calcium-binding protein calbindin D28k is thought Fgf23 /VDR mutant mice was decreased by about 50% to be involved in intracellular calcium transport in renal distal epithe- (Fig 5F). As a control for the specificity of the co-immuno- lium (Lambers et al, 2006). Therefore, we examined the expression precipitation, we used an antibody against the sodium phosphate and intracellular distribution of this cytoplasmic protein in kidneys of transporter-2c (NaPi-2c), a protein exclusively expressed in vehicle- or rFGF23-treated WT mice. We found upregulated mRNA renal proximal tubules. As an additional specificity control, the abundance, a trend towards increased protein expression, but no co-immunoprecipitation experiments were also done using protein change in the intracellular distribution pattern of calbindin D28k, 8 h extracts from cultured murine smooth muscle cells, which neither after rFGF23 administration (Supplementary Fig S3E and F). express TRPV5 nor WNK4 (Fig 5F). In summary, these data demonstrate that FGF23 regulates renal As far as it is known, WNK4 mainly regulates the trafficking of TRPV5 in loss- and gain-of-function models independent of altera- proteins from the Golgi apparatus to the plasma membrane and tions in Klotho expression; Klotho is mainly expressed basolaterally their glycosylation (Jiang et al, 2007, 2008). However, WNK4 may in distal tubule epithelium and does not co-localize with TRPV5 in also be involved in regulating TRPV5 endocytosis (Cha & Huang, the apical cell membrane; soluble Klotho is undetectable in urine, 2010). Immunohistochemical analysis suggested a subcellular redis- and FGF23 signals through the FGFR-Klotho complex to regulate tribution of WNK4 from the basolateral side of the distal tubular TRPV5 in distal tubules. These findings do not support the signaling cells to the area beneath the apical cell membrane in rFGF23-treated Δ/Δ / Δ/Δ model shown in Fig 1A. Rather, the data suggest that FGF23 WT and VDR , but not Kl /VDR mice (Fig 6A). In addition, regulates membrane abundance of TRPV5 by a signaling mechanism WNK4 and TRPV5 appeared to co-localize after rFGF23 treatment in which Klotho functions as the co-receptor required for FGF23 sig- beneath, but not directly within, the apical cell membrane in WT Δ/Δ naling. Hence, we next asked how FGF23 signaling influenced distal and VDR mice (Fig 6A). Quantitative analysis of the immunohis- tubular membrane abundance of TRPV5. tochemical images confirmed the increase in TRPV5-WNK4 co-local- Δ/Δ / FGF23 signaling leads to activation of extracellular signal-regu- ization after rFGF23 treatment in WT and VDR , but not Kl / Δ/Δ lated kinase 1 and 2 (ERK1/2) (Urakawa et al, 2006) which itself VDR mice (Fig 6A). activates serum and glucocorticoid-induced kinase 1 (SGK1) in renal To assess the FGF23-induced changes in subcellular distribution mouse cortical collecting duct cells (Michlig et al, 2004), and in of TRPV5 and WNK4 in more detail, we performed immuno-electron renal proximal tubules (Andrukhova et al, 2012). Renal DCT SGK1 microscopic analyses in renal tissue taken from WT mice 8 h after a phosphorylates and thereby activates with-no-lysine kinase 4 single injection of rFGF23. TRPV5 immunostaining was found exclu- (WNK4) (Ring et al, 2007), which is critically involved in mem- sively in distal tubules and not in proximal tubules (Supplementary brane transport of other ion channels in the DCT such as Na -Cl Fig S5). In vehicle controls, TRPV5 was mainly found in the apical co-transporter (NCC) and renal outer medullar K channel (ROMK1) cell membrane and the basal labyrinth in distal tubular cells, with The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 234 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal some TRPV5 present in membrane vesicles (Fig 6B). FGF23 treat- Golgi apparatus to the apical plasma membrane (Fig 6B). Quantifi- ment induced a striking increase in TRPV5 present in the apical cell cation of the relative grey levels in the apical area of the distal membrane as well as in membrane vesicles trafficking from the tubular cells between nucleus and apical cell membrane in the ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 235 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 4. FGF23 increases urinary calcium reabsorption, TRPV5 plasma membrane abundance and activity in the kidney in gain-of-function mouse models. A, B Urinary calcium excretion (A) and serum PTH (B) in 4-month-old WT mice injected i.p. with vehicle or a single dose of 10 lg rFGF23 per mouse at time 0. Δ/Δ / Δ/Δ C Urinary calcium excretion in 4-month-old WT, VDR , and Kl /VDR mice on rescue diet injected i.p. with vehicle or a single dose of 10 lg rFGF23 per mouse, 8 h post-injection. D Immunohistochemical co-staining of kidney paraffin sections with anti-transmembraneaKlotho (red) antibody, anti-TRPV5 (green) and DAPI (blue). Original magnification ×630. Δ/Δ / Δ/Δ E Western blot analysis of membrane-bound Klotho in renal total protein extracts from 4-month-old WT, VDR , and Kl /VDR mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. F Complex glycosylated TRPV5 protein expression in isolated distal tubular segments treated for 2 h in vitro with rFGF23 alone or in combination with a specific FGFR inhibitor (iFGFR). 2+ G Quantification and original images of intracellular Ca levels in renal distal tubular cells in 300-lm-thick kidney slices of 3-month-old WT mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. Images are overlays of fluorescent with phase contrast images. Kidney slices were stained with the calcium-sensitive dye Fluo-4. Ruthenium red (10 lM) was used as TRPV inhibitor. H Time-dependent changes in distal tubular fluorescence in Fluo-4-loaded, 300-lm-thick kidney slices of 3-month-old WT mice treated at time 0 with rFGF23 (100 ng/ml) or vehicle. After 120, 135, and 150 min, 1, 10, and 50 lM of the TRPV inhibitor ruthenium red (RR) was added, respectively. Data information: Data in (A–C) represent mean  s.e.m. of 3–5 animals each. *P < 0.05 vs. vehicle-treated control in (A–C). Data in (E) represent mean  s.e.m. # # of 3–5 animals each. In (G), *P < 0.05 vs. vehicle-treated, P < 0.05 vs. rFGF23-treated WT mice. In (H), *P < 0.05 vs. vehicle, P < 0.05 vs. fluorescence at 120 min in rFGF23-treated slices. Fluorescence intensity in (G) and (H) was quantified in 7–9 regions of interest per experimental group and time point from three independent experiments. Scale bar, 20 lm in (D) and (G). Source data are available online for this figure. immuno-electron microscopic pictures showed reduced grey levels In order to explore the contribution of other pathways initiated (i.e., more black) in rFGF23-treated mice (Fig 6C), in accordance by FGF23 signaling through the FGFR-Klotho complex, we treated with increased abundance of TRPV5-containing membrane vesicles isolated distal tubular segments with rFGF23 or with specific Src in the sub-apical cellular area. kinase (SRC), phospholipase C (PLC), and phosphatidylinositol In vehicle controls, WNK4 was mainly associated with localized 3-kinase (PI3K) inhibitors, alone or in combination, for 2 h in vitro. areas in the basal labyrinth as well as with membrane vesicles traf- SRC and PLC inhibition did not influence the rFGF23-induced up- ficking to the apical plasma membrane, with most of the immuno- regulation of TRPV5 expression (Supplementary Fig S6). However, staining observed in the peri-nuclear area (Fig 6B). Eight hours after PI3K inhibition partially blocked the rFGF23-induced increase in rFGF23 treatment, WNK4 was still found at the basal labyrinth, but TRPV5 expression (Supplementary Fig S6). Taken together, these the main localization was around membrane vesicles in close prox- data suggest that ERK1/2 activation is the major pathway, but that imity to and along the apical plasma membrane (Fig 6B). These the PI3K pathway may play an additional role in the regulation of findings are consistent with the notion that activated WNK4 is TRPV5 by FGF23 in renal distal tubules. involved in the FGF23-induced regulation of the cellular trafficking In order to more rigorously establish the putative FGFR-ERK1/ of TRPV5 from the Golgi apparatus to the apical plasma membrane 2-SGK1-WNK4 signaling pathway of the TRPV5 regulation by in distal tubular epithelium, especially in fusion of membrane vesi- FGF23, we reconstituted this pathway in MDCK cells. To this end, cles with the apical cell membrane. We found no evidence of WNK4 we transfected MDCK cells with murine TRPV5, SGK1, and WNK4 associated with endocytotic pits. alone or in combination, treated the transfected cells with vehicle or To verify that the regulation of TRPV5 by FGF23 is a direct effect rFGF23 alone or in combination with soluble Klotho for 12 h, and on the distal tubule and to test the essential roles of ERK1/2 and subsequently examined complex glycosylated TRPV5 protein SGK1 in this process, we isolated distal tubular segments from WT expression by immunoblotting. It is known that MDCK cells express mice, and treated these segments with rFGF23 alone or rFGF23 ERK1/2 and FGFR1 (Kessler et al, 2008), and as shown in Fig 8A, together with specific SGK1 or ERK1/2 inhibitors for 1 to 4 h these cells do not endogenously express TRPV5. rFGF23 profoundly in vitro. rFGF23 time-dependently upregulated phosphorylation of upregulated TRPV5 expression in MDCK cells co-transfected with ERK1/2 and SGK1 (Fig 7A), as well as protein expression of TRPV5 TRPV5, SGK1, and WNK4 (Fig 8A). Recombinant soluble Klotho in distal tubular segments (Fig 7B). The latter effect was completely protein, alone or in combination with rFGF23, did not influence blocked in the presence of SGK1 or ERK1/2 inhibitors (Fig 7B), TRPV5 expression (Fig 8A). In the absence of SGK1, rFGF23 failed showing that ERK1/2 and SGK1 activation are essential for the to upregulate TRPV5 expression (Fig 8A). However, small increases FGF23-induced regulation of TRPV5 expression in the distal tubule. in TRPV5 expression in response to rFGF23 treatment were seen in In the presence of an ERK1/2 inhibitor, rFGF23 did not increase cells transfected with TRPV5 and SGK1 only (Fig 8A). A possible phosphorylation of SGK1 in isolated proximal tubular segments explanation for this finding may be that MDCK cells, although not from WT mice, indicating that SGK1 is a mediator downstream of reported in the literature, might express low amounts of WNK4 or ERK1/2 activation (Fig 7C). Furthermore, immunoprecipitation other WNK kinases. To examine the biological response to the experiments in isolated distal tubular segments from WT mice rFGF23-induced increase in TRPV5 expression, we transfected revealed that the rFGF23-induced increase in serine phosphorylation MDCK cells with TRPV5, SGK1, and WNK4, treated the transfected of WNK4 is blocked in the presence of an SGK1 inhibitor (Fig 7D). cells with vehicle, rFGF23 or Klotho for 12 h, and then monitored Thus, activation of ERK1/2 and SGK1 by FGF23 leads to down- intracellular calcium concentrations by 2-photon microscopy after stream activation of WNK4. loading the cells with Fluo-4. rFGF23 induced a striking increase in The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 236 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 5. FGF23 increases TRPV5 protein abundance in the plasma membrane through a signaling pathway involving ERK1/2, SGK1, and WNK4. A–D Western blot analysis of complex glycosylated TRPV5 (A) protein (92 kDa) expression in renal cortical total membrane fractions, as well as phosphorylated ERK1/2 (B), Δ/Δ / Δ/Δ phosphorylated SGK1 (C), and total WNK4 (D) in homogenized renal cortex samples of 9-month-old WT, VDR and Fgf23 /VDR mice on rescue diet that had been injected with rFGF23 (5 lg/mouse). Tissue samples were taken from mice 8 h post-injection. E Reciprocal immunoprecipitation (IP) of serine-phosphorylated (P-Ser) proteins, followed by Western blot (WB) analysis of WNK4 or vice versa from homogenized renal cortex protein samples of 4-month-old WT mice treated with vehicle (Veh) or rFGF23 (10 lg/mouse) 8 h before necropsy. Western blot analysis of WNK4 in renal cortex protein samples immunoprecipitated with WNK4 antibody was used as a loading control. F Co-immunoprecipitation of TRPV5/WNK4 complexes. WNK4 or TRPV5 were immunoprecipitated with specific antibodies (anti-TRPV5, anti-WNK4, anti-NaPi-2c) Δ/Δ / Δ/Δ from homogenized renal cortex protein samples of 9-month-old WT, VDR and Fgf23 /VDR mice on rescue diet. Subsequently, Western blot analysis was performed with corresponding anti-TRPV5 or anti-WNK4 antibodies to identify co-precipitated TRPV5 and WNK4 protein, respectively. Primary cultured murine smooth muscle cells (SMC) were used as negative control. Data information: Data in (A–D) represent mean  s.e.m. of 5–8 animals each, *P < 0.05 vs. vehicle-treated WT, P < 0.05 vs. rFGF23-treated WT mice. Data in (E) represent mean  s.e.m. of five animals each. Data in (F) represent mean  s.e.m. of 4–5 animals each. Frame in Western blot image indicates splicing # D/D event. In (F) *P < 0.05 vs. WT, P < 0.05 vs. VDR mice. Source data are available online for this figure. intracellular calcium in MDCK cells transfected with TRPV5, SGK1 blocked by the TRPV inhibitor ruthenium red (Fig 8B). To unequiv- and WNK4, reflecting increased TRPV5 channel activity (Fig 8B). In ocally show that FGF23 stimulates a TRPV5-mediated calcium influx contrast, recombinant soluble Klotho protein did not influence intra- pathway, we treated MDCK cells transfected with TRPV5, SGK1 and cellular calcium in TRPV5-expressing cells (Fig 8B). As expected, WNK4 with rFGF23 in the presence or absence of the calcium chela- the rFGF23-induced increase in intracellular calcium could be tor EGTA (Fig 8C). In the presence of EGTA, rFGF23 did not ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 237 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 6. FGF23 increases apical plasma membrane abundance of TRPV5 and causes re-distribution of WNK4. Δ/Δ A Immunohistochemical co-staining with anti-WNK4 (red) antibody, anti-TRPV5 (green) and DAPI (blue) of kidney paraffin sections from 4-month-old WT, VDR and / Δ/Δ Klotho /VDR mice treated with vehicle or rFGF23 (10 lg/mouse) 8 h before necropsy. Scale bar, 20 lm. Right panel shows quantification of distal tubular TRPV5 and WNK4 co-localization performed in four animals per group with 3–6 images per animal and 4–6 different regions of interest per image. B Immuno-electron microscopic staining using anti-TRPV5 (left panels) and anti-WNK4 (right panels) antibodies in kidneys from 3 to 4-month-old WT mice treated with vehicle (Veh) or rFGF23 (10 lg/mouse) 8 h before necropsy. Lower panels show higher magnification of the apical cell area from representative sections. Scale bar, 500 nm. C Quantification of the relative grey levels in the apical area of the distal tubular cells between nucleus and apical cell membrane in the anti-TRPV5-stained immuno- electron microscopic pictures performed in four animals per group with 4–5 images per animal and 4–6 different regions of interest per image. Data information: *P < 0.05 vs. vehicle-treated controls in (A) and (C). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 238 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Figure 7. FGF23-induced increase in TRPV5 protein abundance in the plasma membrane is mediated by a signaling cascade involving ERK1/2 and SGK1. A, B Phosphorylated ERK1/2 and SGK1 (A) and complex glycosylated TRPV5 (B). Protein expression in isolated distal tubular segments from WT mice treated in vitro with rFGF23 alone or in combination with specific SGK1 (iSGK1) and ERK1/2 (iERK1/2) inhibitors. Frame in Western blot image in (A) indicates splicing event. C Phosphorylated SGK1 in isolated distal tubular segments treated in vitro with rFGF23 alone or in combination with a specific ERK1/2 inhibitor (iERK1/2). D Reciprocal immunoprecipitation (IP) of serine-phosphorylated (P-Ser) proteins, followed by Western blot (WB) analysis of WNK4 or vice versa from isolated distal tubular segments treated in vitro with rFGF23 alone or in combination with a specific SGK1 inhibitor (iSGK1). Western blot analysis of WNK4 in renal cortex protein samples immunoprecipitated with WNK4 antibody was used as a loading control. Data information: Actin loading controls are shown for tSGK1 in (A), rFGF23 + iERK1/2 in (B), and tSGK1 in (C). All presented protein expression values were normalized to individual actin levels. Data in (A–C) represent mean  s.e.m. of 4–5 samples each. Data in (D) represent mean  s.e.m. of 3–4 samples each. *P < 0.05 vs. vehicle control (Co). Source data are available online for this figure. ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 239 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Figure 8. FGF23 increases TRPV5 protein abundance and channel activity in MDCK cells transfected with murine TRPV5, SGK1 and WNK4. A Complex glycosylated TRPV5 protein expression in total protein homogenates of MDCK cells transiently transfected with murine TRPV5 (T), SGK1 (S) and WNK4 (W) constructs, alone or in combination, and treated for 12 h with vehicle, rFGF23 or recombinant Klotho (rKlotho) alone or in combination. Mock-transfected cells were used as a negative control (Co). 2+ B Quantification and original images of intracellular Ca levels in MDCK cells transiently transfected with murine TRPV5, SGK1 and WNK4 constructs after treatment with vehicle, rFGF23 or recombinant Klotho (rKlotho) for 12 h. Ruthenium red (10 lM) was used as TRPV inhibitor. 2+ C Time-dependent changes and original images of intracellular Ca levels in MDCK cells transiently transfected with murine TRPV5, SGK1 and WNK4 constructs, and treated at time 0 with rFGF23 (100 ng/ml) or vehicle in the presence or absence of 1.5 mM EGTA added to the culture medium. After 105 min, 1 lM of ruthenium red (RR) was added. MDCK cells were stained with the calcium-sensitive dye Fluo-4 in (B) and (C). Data information: Data in (A) represent mean  s.e.m. of 6–9 samples each from three independent experiments. Data in (B) and (C) represent mean  s.e.m. of 3–4 samples each from three independent experiments. Scale bar, 20 lm in (B), 7 lm in (C). *P < 0.05 vs. vehicle in (A) and (B), and vs. EGTA + rFGF23 in (C), P < 0.05 vs. fluorescence at 105 min in (C). Source data are available online for this figure. increase intracellular Fluo-4 fluorescence (Fig 8C). These data the model shown in Fig 1B which is able to explain the striking / Δ/Δ / Δ/Δ firmly establish that the SGK1-WNK4 pathway mediates the regula- similarities between Kl /VDR and Fgf23 /VDR mice in the tion of TRPV5 expression and activity by FGF23 in renal epithelial regulation of TRPV5. Klotho functions as the co-receptor for blood- cells, and further suggest that soluble Klotho does not influence borne FGF23 in this model, not as a calcium-regulating hormone TRPV5 channel activity in these cells. per se. In analogy to the other major phosphaturic hormone PTH, FGF23 is not only a phosphaturic, but also a calcium-conserving hormone in the kidney, activating, through the FGFR-Klotho complex, Discussion a signaling cascade involving ERK1/2, SGK1, and WNK4. Thus, our data may explain the hitherto unresolved question why injection of Our data have uncovered a previously unknown VDR independent FGF23 does not lower serum calcium despite its suppressive action on function of FGF23 signaling in renal calcium transport. We propose renal synthesis of bioactive vitamin D (Shimada et al, 2001, 2004). The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 240 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Although several reports have suggested that soluble Klotho may FGF23 directly acts on proximal tubular cells to down-regulate regulate the membrane abundance and function of ion channels membrane abundance of NaPi-2a through the ERK1/2-SGK1 – + + such as TRPV5, ROMK and NaPi-2a in an FGF23-independent Na /H exchange regulatory cofactor (NHERF)-1 signaling axis manner, and likewise the function of receptors such as the IGF-1 (Andrukhova et al, 2012). In contrast, Farrow and coworkers (Far- receptor (Chang et al, 2005; Kurosu et al, 2005; Hu et al, 2010), row et al, 2009) found that the earliest changes in activation of there is accumulating evidence that Klotho does not have ERK1/2 after injection of FGF23 in vivo in mice occur in the distal FGF23-independent effects on mineral or glucose homeostasis in tubules. It was previously thought that FGF23 acts on the distal vivo. For example, the phenotype of Fgf23 and Klotho double knock- tubule where it generates an unknown endocrine or paracrine sec- / / out mice is indistinguishable from that of Fgf23 and Kl single ondary signal that in turn signals back to the proximal tubule to knockout mice (Nakatani et al, 2009), strongly arguing against downregulate transcellular phosphate transport as part of a “dis- FGF23-independent functions of Klotho. In addition, we previously tal-to-proximal tubular feedback mechanism” (White & Econs, showed that glucose metabolism, bone turnover, and steady-state 2008; Farrow et al, 2010; Martin et al, 2012). Our current study is / Δ/Δ Δ/Δ PTH secretion is unchanged in Kl /VDR compared with VDR able to explain this apparent discrepancy: Circulating FGF23 has mice (Anour et al, 2012), suggesting that Klotho does not have parallel and independent, direct effects on proximal and distal physiologically relevant VDR- and Fgf23-independent functions in renal tubules, suppressing phosphate re-uptake in proximal and the regulation of mineral and glucose homeostasis in vivo. Thus, the stimulating calcium reabsorption in distal tubular cells. In both major physiological role of Klotho appears to be its function as an proximal (Andrukhova et al, 2012) and distal tubular epithelium, obligatory co-receptor for FGF23. However, there is no conclusive ERK1/2 and SGK1 activation are the initial downstream signaling evidence at present to rule out an additional enzymatic function of events after binding of FGF23 to the binary FGFR-Klotho complex. Klotho. Based on our in vitro experiments in isolated distal tubular seg- To address the question of whether or not Klotho has FGF23- ments, FGF23-induced activation of the PI3K pathway may also independent functions on calcium metabolism, we attempted play an additional role in the regulation of transepithelial ion / / D/D breeding 4-week-old Fgf23 /Kl /VDR triple mutant mice on transport, at least in distal tubules. rescue diet in order to compare the phenotype of these mice to that Earlier studies showed that PTH increases transcription (Okano / D/D / D/D of Fgf23 /VDR and Kl /VDR double mutant mice. The et al, 2004), and activates membrane-anchored TRPV5 by protein / / D/D breeding yielded only one Fgf23 /Kl /VDR triple mutant kinase A-mediated phosphorylation (de Groot et al, 2009). Thus, it animal so far. This animal displayed no difference in gross phenotype is interesting to note that in both proximal and distal renal tubules, (body weight, BW 12.5 g) and unchanged blood ionized calcium FGF23 and PTH signaling converge on the same molecules. In the D/D concentration (1.30 mM) as compared to VDR single mutants proximal tubule, we and others showed that PTH and FGF23 signaling 2+ (means  s.d.; BW 13.3  2.1 g, Ca 1.24  0.05 mM, n = 3), as converge on NHERF-1 (Weinman et al, 2011; Andrukhova et al, / D/D 2+ well as Fgf23 /VDR (BW 12.5  1.3 g, Ca 1.24  0.04 mM, 2012) to downregulate NaPi-2a membrane abundance, whereas in / D/D 2+ n = 3) and Kl /VDR (BW 12.1  0.5 g, Ca 1.20  0.13 mM, the distal tubule both signaling pathways converge on TRPV5. In n = 3) double mutant littermates (O. Andrukhova et al, unpub- the model shown in Fig 1B, PTH increases the open probability of lished data). Although obviously still preliminary, these data membrane-anchored TRPV5 by protein kinase A-mediated phos- support the hypothesis that Klotho does not have VDR- and FGF23- phorylation, whereas FGF23 signaling is required for membrane independent actions on calcium metabolism. transport of the channel. Although at present in vivo data supporting In agreement with our finding that protein expression of TRPV5 this hypothesis are lacking, our model would predict that the levels Δ/Δ did not differ between kidneys of WT and VDR mice, earlier of circulating FGF23 may modulate renal sensitivity to the stimulat- studies reported unchanged renal TRPV5 mRNA expression in VDR- ing action of PTH on renal calcium re-absorption. ablated mice (Weber et al, 2001; Okano et al, 2004). However, Our study has established a novel link between FGF23 signal- administration of pharmacological 1,25(OH) D doses to mice or ing and WNK4 activation in distal tubular epithelium. FGF23 stimulation of primary cultures of murine renal tubular cells with signaling resulted in increased phosphorylation of WNK4. The pharmacological doses of 1,25(OH) D directly increases transcrip- immunoelectron microscopic analysis clearly showed that FGF23 tion of the TRPV5 gene in a VDR-dependent fashion (Okano et al, induces intracellular redistribution of WNK4 to the area beneath 2004). Taken together, these data suggest that vitamin D is a phar- the apical cell membrane, where membrane vesicles fuse with the macological, but not a physiological regulator of TRPV5 protein outer cell membrane. Thus, the most important function of WNK4 expression in the murine kidney. Along similar lines, some previous in the regulation of TRPV5 membrane abundance in the DCT in experimental evidence suggested that vitamin D signaling may vivo may be the control of membrane vesicle transport and facili- upregulate renal Klotho mRNA expression (Tsujikawa et al, 2003; tation of vesicle fusion with the apical membrane, rather than the Li et al, 2011). However, our data clearly showed that neither control of clathrin-dependent endocytosis as suggested by some in absent VDR signaling nor pharmacological stimulation with rFGF23 vitro experiments (Cha & Huang, 2010). The function of WNK4 at alters renal Klotho protein expression, suggesting only little physio- focal sites in the basal labyrinth as well as the function of TRPV5 logical regulation by vitamin D and FGF23 at the level of the expressed in the basal labyrinth of distal tubular cells remains co-receptor in vivo. unknown. Because WNK4 is involved in membrane transport of It is well known that FGF23 is primarily a phosphaturic other ion channels such as NCC and ROMK1, FGF23 signaling hormone, suppressing transcellular phosphate transport in proximal may also influence the membrane abundance of other ion chan- tubules. We previously reported that Klotho is expressed in the nels in distal tubules. In this context, it is interesting to note that basolateral membrane of proximal tubular epithelium, and that Kl hypomorphic mice have decreased sodium plasma levels and ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 241 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al increased circulating aldosterone concentrations (Fischer et al, Serum and urine biochemistry 2010), suggesting that Klotho deficiency, and thus reduced FGF23 signaling, may induce alterations in renal sodium handling. WNK Serum creatinine, calcium, and phosphorus as well as urinary crea- kinases act as a complex of at least 3 different kinases, WNK1, 3, tinine, calcium, and phosphorus were analyzed on a Hitachi 912 and 4, to control the intracellular transport of membrane proteins Autoanalyzer (Boehringer Mannheim) or on a Cobas c 111 analyzer (McCormick et al, 2008). In the current report, we focused on (Roche). Serum Klotho protein concentrations were determined WNK4, but it is possible that FGF23 signaling involves other using a mouse specific Klotho ELISA kit (Cusabio) according to the members of the WNK family as well. Clearly, more experimenta- manufacturer’s protocol. The detection limit of the latter assay is tion is required to elucidate all ramifications of the FGF23 signaling 0.8 pg/ml. pathway in renal tubular epithelium. Our finding that FGF23 stimulates renal tubular calcium Immunohistochemistry reabsorption is important for our understanding of the endocrine networks involved in calcium and phosphate homeostasis, and For immunohistochemistry, 5-lm-thick paraffin sections of parafor- suggests that in physiological situations where phosphate excretion maldehyde (PFA)-fixed kidneys were prepared. Some sections were needs to be increased extracellular calcium is still conserved. The stained with hematoxylin/eosin (H & E) by routine methods. pathophysiological downside of this regulation may occur in dis- Before immunofluorescence staining, dewaxed sections were pre- treated for 60 min with blocking solution containing 5% normal eases such as chronic kidney disease, where PTH and FGF23 are chronically elevated due to decreased renal 1,25(OH) D production goat serum in PBS with 0.1% bovine serum albumin and 0.3% and phosphate retention. In this situation, both hormones will stim- Triton X-100. Without rinsing, sections were incubated with poly- ulate renal calcium conservation, possibly contributing to calcium clonal rabbit anti-aKlotho (Alpha Diagnostics, 1:1,000; raised against the cytoplasmic region of aKlotho, detecting the mem- accumulation and vascular calcifications in these hyperphosphatemic patients. brane-bound form), polyclonal rabbit anti-aKlotho (abcam, 1:1,000; raised against a part of KL2, detecting membrane-bound and ecto- domain shed forms), anti-WNK4 (Novus Biologicals, 1:300), anti- Materials and Methods TRPV5 (Alpha Diagnostics, 1:1,000), anti-b-actin (Sigma, 1:5,000), or anti-calbindin D-28k (Swant, 1:1,000) antibodies at 4°C over- Animals night. After washing, sections were incubated for 1.5 h with goat anti-rabbit Alexa 548 and goat anti-mouse Alexa 488 secondary All animal procedures were approved by the Ethical Committee antibodies (Invitrogen, 1:400), respectively. Immunostaining of of the University of Veterinary Medicine Vienna. Heterozygous tissue sections in which either one or both secondary antibodies +/D +/ VDR (Erben et al, 2002) were mated with heterozygous Fgf23 were omitted served as a negative control. The slides were ana- +/ (Sitara et al, 2004), and heterozygous Klotho (Lexicon Genetics, lyzed on a Zeiss LSM 510 Axioplan 2 confocal microscope Mutant Mouse Regional Resource Centers, University of California, equipped with a 63 × oil immersion lens (NA 1.3). By use of the Davis, CA, USA) mutant mice to generate double heterozygous ani- multitrack function, individual fluorochromes were scanned with +/D +/ +/D +/ mals. VDR /Fgf23 and VDR /Klotho mutant mice on laser excitation at 488 and 543 nm separately with appropriate filter D/D C57BL/6 background were interbred to generate WT, VDR , sets to avoid cross talk. Pictures were processed using Adobe / / D/D / D/D Klotho , Fgf23 and compound VDR /Fgf23 and VDR / Photoshop (overlays). Quantification of distal tubular TRPV5 and Klotho mutant mice. Genotyping of the mice was performed by WNK4 co-localization was performed using ImageJ software multiplex PCR using genomic DNA extracted from tail as described (National Institutes of Health, Bethesda, MA, USA) in four animals (Hesse et al, 2007; Anour et al, 2012). The mice were kept at 24°C per group with 3–6 images per animal and 4–6 different regions of with a 12 h/12 h light/dark cycle, and were allowed free access to a interest per image. rescue diet and tap water. The rescue diet (Ssniff, Soest, Germany) containing 2.0% calcium, 1.25% phosphorus, 20% lactose and Immuno-electron microscopic analysis 600 IU vitamin D/kg was fed starting from 16 days of age. This diet has been shown to normalize mineral homeostasis in VDR-ablated For transmission electron microscopy, kidney samples of 4 animals mice (Li et al, 1998; Erben et al, 2002; Zeitz et al, 2003). All experi- per group were fixed in 4% PFA for 24 h at 4°C. Longitudinal ments were performed on male offspring of double heterozygous × 100-lm-thick sections were cut using a Leica VT1000 Vibratome double heterozygous matings. Urine was collected in metabolic (Leica Microsystems). Sections were post-fixed in 4% PFA for 1 h cages before necropsy. Some mice received a single intraperitoneal and rinsed in 0.1 M phosphate buffer 3 × 10 min. Peroxidase injection of vehicle (phosphate-buffered saline (PBS) with 2% activity was inhibited by 30-min incubation with 3% H O in PBS 2 2 DMSO) or recombinant human FGF23 R176/179Q (rFGF23) (Goetz for and nonspecific binding was minimized by incubation for et al, 2007) (10 lg per 3–4-month-old mouse; 5 lg per 9-month-old 60 min in 3% normal goat serum containing 1% BSA. Incubation mouse) or), and were killed 8 or 24 h post-injection. In the mice with anti-TRPV5 (Alpha Diagnostics, 1:150), anti-WNK4 (Novus receiving rFGF23, spontaneous urine was collected before as well as Biologicals, 1:500), anti-aKlotho (Alpha Diagnostics, 1:300), anti- 8 and 24 h post-injection. At necropsy, the mice were exsanguinated aKlotho (abcam, 1:400) or anti-calbindin D-28k (Swant, 1:500) from the abdominal V. cava under anesthesia with ketamine/xyla- was carried out at 4°C overnight. For negative controls, the zine (67/7 mg/kg i.p.) for serum collection. In all experiments 4–9 primary antibody was omitted. Peroxidase-labeled rabbit Power- TM mice were used per experimental group. Vision (ImmunoVision Technologies) secondary system was The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 242 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal employed for antibody detection with subsequent diaminobenzi- described (Hu et al, 2010). Immunoblots were incubated overnight dine (DAB, Sigma) staining. Post-fixation was performed in 1% at 4°C with primary antibodies including polyclonal mouse anti- osmium tetroxide for 2 h at RT followed by dehydration and incu- TRPV5 (1:3,000, Alpha Diagnostics), polyclonal rabbit anti-aKlotho bation in propylene oxide, propylene oxide–epon and subsequent (1:2,000, Alpha Diagnostics, membrane-bound form), polyclonal embedding in pure epon 812. Thin sections were stained with rabbit anti-aKlotho (1:2,500, abcam; membrane-bound and shed lead citrate, and were investigated under a transmission electron forms), polyclonal rabbit anti-WNK4 (1:2,000, Novus Biologicals), microscope (Zeiss EM 900). Apical expression of TRPV5 was polyclonal anti-calbindin D-28k (1:2,000, Swant), polyclonal anti- quantified using Image J software by gray value determination in uromodulin (1:1,500, Sigma), monoclonal mouse anti-b-actin 4–5 images of four animals per group with 4–6 regions of interest (1:5,000, Sigma), monoclonal anti-total-ERK1/2 (BD Biosciences), in each image. anti-phospho-ERK1/2 (Cell Signaling), anti-total-SGK1 (Alpha Diag- nostics), and anti-phospho-SGK1 (Santa Cruz Biotechnology) in Total cell membrane isolation 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]. Mouse kidney cortex tissue was homogenized in a buffer consisting After washing, membranes were incubated with horseradish of 20 mM Tris (pH 7.4/HCl), 5 mM MgCl , 5 mM NaH PO ,1mM peroxidase-conjugated secondary antibodies (Amersham Life 2 2 4 ethylenediamine tetraacetic acid (pH 8.0/NaOH), 80 mM sucrose, Sciences). Specific signal was visualized by ECL kit (Amersham 1 mM phenyl-methylsulfonyl fluoride, 10 lg/ml leupeptin and Life Sciences). The protein bands were quantified by Image Quant 10 lg/ml pepstatin, and subsequently centrifuged for 15 min at 5.0 software (Molecular Dynamics). The expression levels were 4,000 g. Supernatants were transferred to a new tube and centri- normalized to Ponceau S stain and to individual b-actin expression fuged for an additional 30 min at 16,000 g. levels. The glycosylation of the Klotho and TRPV5 proteins was detected on nitrocellulose membranes by GLYCO-PRO assay Isolation of distal tubular segments (Sigma) according to the manufacturer’s protocol. For validation of anti-TRPV5 antibody specificity, total kidney homogenates of Renal distal tubules were isolated as reported previously (Burg et al, 4-week-old TRPV5 male mice (generous gift of Rene J.M. 1997; Schafer et al, 1997; Gonzalez-Mariscal et al, 2000). In brief, Bindels, University Medical Center Nijmegen, Netherlands) were murine kidneys were perfused with sterile culture medium (Ham’s used as a negative control. Each experiment included 4–8 animals F12; GIBCO) containing 1 mg/ml collagenase (type II; Sigma) and per group. All presented Western blot images are from the same 1 mg/ml pronase E (type XXV, Sigma) at pH 7.4 and 37°C. The cor- gel each. All splicing events are indicated by frames in Western tical tissue was dissected in small pieces and placed at 37°C in ster- blot images. ile 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 centri- Co-immunoprecipitation fugation at 3000 rpm for 4 min, the enzyme-containing solution was removed, and tubules were resuspended in ice-cold medium. Kidney cortex homogenate protein samples (1 mg), dissected proxi- Individual distal tubule segments were identified based on morphol- mal tubular segments (40 lg), or homogenate protein samples from ogy in a dissection microscope at x25–40 magnification by their murine smooth muscle cells (1 mg) were incubated with 2 lgof WNK4 (Novus Biologicals), TRPV5 (Lifespan Biosciences), anti- appearance and dimensions. Distal tubular segments were incu- bated with rFGF23 (100 ng/ml) and/or 10 ng/ml of the specific phosphoserine (Alpha Diagnostics), or anti-NaPi-2c (kind gift of SGK1 kinase inhibitor GSK 650394 (Axon Medchem), or 10 ng/ml Drs. Murer and Biber) antibody at 4°C overnight. The immune of the ERK1/2 inhibitor PD184352 (Sigma) for 1, 2 and 4 h. For complexes were captured by adding 50 ll Protein A or G agarose/ sepharose beads (Santa Cruz Biotechnology) and overnight incuba- experiments with FGFR inhibition, distal tubular segments were incubated with rFGF23 (100 ng/ml) and/or 40 ng/ml of the specific tion at 4°C with gentle rocking. The immunoprecipitates were col- FGFR inhibitor PD173074 (Sigma) for 2 h. For experiments with lected by centrifugation at 1,000 g for 5 min at 4°C and washed for SRC, PLC and PI3K inhibition, distal tubular segments were incu- four times in PBS, each time repeating the centrifugation step. After bated with rFGF23 (100 ng/ml) and/or 150 nM Scr inhibitor-1, the final wash, the pellets were suspended in 40 ll of electrophore- 50 nM PI3K inhibitor wortmannin or 10 lM PLC inhibitor U-73122 sis sample buffer and boiled for 2–3 min. Western blot analysis was (all from Sigma) for 2 h. Protein samples were collected for Western performed using a primary anti-TRPV5 or anti-WNK4 antibody. blotting analysis in lysis buffer. Each experiment was performed in duplicates and included 4–8 animals per group. Western Blot RNA isolation and quantitative RT-PCR Kidney cortex homogenate, total cell membrane samples or dis- sected distal tubular segments were solubilized in Laemmli sample Shock-frozen tissues were homogenized in TRI Reagent (Molecu- buffer, fractionated on SDS-PAGE (30 lg/well) and transferred to a lar Research Center) and total RNA was extracted according to nitrocellulose membrane (Thermo Scientific). For detection of the manufacturer’s protocol. RNA purity and quality was deter- Klotho protein in urine, 40 ll of fresh bladder urine, salt precipi- mined using a 2100 Bioanalyzer (Agilent Technologies). One lg tated urine, or 1.3- and 2-fold concentrated urine using Amicon of RNA was used for first-strand cDNA synthesis (iScript cDNA Ultra-0.5 centrifugal filters (Millipore) were loaded per well as Synthesis Kit, Bio-Rad). Quantitative RT-PCR was performed on a ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 243 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al TM TM Rotor-Gene 6000 (Corbett Life Science) using QuantiFast fluorescence intensity and the cell square was calculated for each EverGreen PCR Kit (Qiagen). A melting curve analysis was done cell (16–48 cells per individual sample). for all assays. Primer sequences are available on request. Efficien- cies were examined based on a standard curve. Expression of tar- Statistical analyses get genes was normalized to the expression of the housekeeping gene glyceraldehyde-3-phosphate-dehydrogenase (GAPDH). Statistics were computed using PASW Statistics 17.0 (SPSS Inc., Chicago, IL, USA) The data were analyzed by two-sided t-test (2 In vitro transient transfection experiments groups) or 1-way analysis of variance (ANOVA) followed by Student-Newman-Keuls multiple comparison test (>2 groups). The pCMV6 plasmid construct containing mouse TRPV5 ORF P values of less than 0.05 were considered significant. The data are (MR225560, Origene), mouse SGK1 and mouse WNK4 (generous presented as the mean  s.e.m. gifts of David H. Ellison, Oregon Health and Science University, Port- land, Oregon, USA) were used for in vitro transient transfection Supplementary information for this article is available online: http://emboj.embopress.org experiments. MDCK (Sigma) cells were grown in EMEM (EBSS) with 2 mM glutamine, 1% non-essential amino acids, 10% fetal bovine serum and 100 lg/ml penicillin and 100 lg/ml streptomycin at 37°C Acknowledgements in a 5% CO /95% air humidified atmosphere. 24 h after seeding, We thank C. Bergow and S. Hirmer for excellent technical assistance, Martin cells at ~70% confluence were subjected to transient transfection Glosmann € for help with confocal microscopy, Ingrid Miller and Manfred using X-tremeGENE 9DNA transfection reagent (Roche) according to Gemeiner for help with immunoprecipitation, Magdalena Helmreich and the manufacturer’s protocol. The cells were co-transfected with Waltraud Tschulenk for help with the immuno-electron microscopy, and 0.5 lg of each construct. 24 h post-transfection, the cells were trea- Stefan Handschuh for help with the quantification of immunohistochemistry ted with either vehicle or 100 ng/ml rFGF23 and/or 10 ng/ml and immuno-electron microscopy. Recombinant FGF23 for some of the studies recombinant mouse Klotho (R&D Systems). Complex glycosylated was a generous gift of Vicki Shalhoub, Amgen Inc., Thousand Oaks, CA, USA. TRPV5 expression was monitored by Western blotting in triplicate / Kidney samples of TRPV5 mice were a generous gift of Rene J.M. Bindels, 12 h after cell stimulation. University Medical Center Nijmegen, Netherlands. Mouse SGK1 and WNK4 constructs were a generous gift of David H. Ellison, Oregon Health and Science 2+ Ca imaging and image analysis University, Portland, Oregon, USA. The polyclonal rabbit anti-NaPi-2c antibody was a generous gift of Drs. Jurg Biber and Heini Murer, University of Zurich, The MDCK cells and kidney slices (300-lm-thick) were incubated Switzerland. This work was supported by a grant from the Austrian Science for 30 min at 37°C with 2 lM of the calcium-sensitive dye Fluo-4 Fund (FWF 24186-B21) to R.G.E. O.A. was supported by a postdoctoral fellow- (Molecular Probes), diluted in DMSO (Merck Millipore Interna- ship of the University of Veterinary Medicine Vienna. B.L. was supported by tional) and 20% Pluronic (Merck Millipore International). Thereaf- NIH/NIDDK DK072944, and M.M. was supported by NIH/NIDCR DE13686. ter, the slices or cells were washed two times for 15 min each in PBS. Some kidney slices from WT mice were incubated in vitro for Author contributions 2 h with rFGF23 (100 ng/ml) or vehicle. For calcium visualization OA and RGE conceived and designed the experiments; OA, CS, AS and UZ per- in kidney slices, Fluo-4 was excited by a two-photon laser formed experiments and analyzed the data; OA, RG, MM, BL, and RGE wrote (Ti-sapphire laser, Coherent Inc.) at a wavelength of 820 nm. the manuscript; AS, ME, RG, VS, MM, EEP, and BL provided important tools or Images (512 × 512 pixels) were acquired every 30 s at a depth of techniques; OA, AS, ME, RG, VS, MM, EEP, BL, and RGE discussed and reviewed 50–80 lm. For inhibition of TRPV activity, tissue slices were incu- the manuscript. bated ex vivo with 1–50 lM of ruthenium red (Sigma) at 37°C, 5% CO /95% air humidified atmosphere for 30 min. Fluorescence Conflict of interest images were analyzed using Image J software. The whole epithelial layer of the tubule except for the lumen was selected (by manually VS was an employee of Amgen, Inc. The other authors declare that they have 2+ drawing the region of interest, ROI) to quantify Ca levels in renal no conflict of interest. distal tubules. Fluorescence intensity was quantified in 7–9 ROIs per experimental group from three independent experiments. The ratio between the fluorescence intensity and the ROI was calculated for References each tubule. Fluorescence imaging of cultured cells was performed using an Alexander RT, Woudenberg-Vrenken TE, Buurman J, Dijkman H, van der inverse confocal microscope (TCS SP5 Leica Microsystems), with a Eerden BC, van Leeuwen JP, Bindels RJ, Hoenderop JG (2009) Klotho Leica water immersion objective (63×). Fluo-4 was excited by an prevents renal calcium loss. J Am Soc Nephrol 20: 2371 – 2379 argon laser at 488 nm. Images were acquired every 30 s. Ruthenium Andrukhova O, Zeitz U, Goetz R, Mohammadi M, Lanske B, Erben RG (2012) red (1–10 lM) was used as TRPV inhibitor. In some experiments, FGF23 acts directly on renal proximal tubules to induce phosphaturia 1.5 mM EGTA (Sigma) was added to the culture medium to chelate through activation of the ERK1/2-SGK1 signaling pathway. Bone 51: extracellular calcium. For each MDCK cell, the whole-cell and 621 – 628 nucleus contours were drawn manually. The mean fluorescence Anour R, Andrukhova O, Ritter E, Zeitz U, Erben RG (2012) Klotho lacks a intensity was measured in 4–8 separate cells from each field of view vitamin D independent physiological role in glucose homeostasis, bone (2–7 fields of view per individual sample). The ratio of the turnover, and steady-state PTH secretion in vivo. PLoS ONE 7:e31376 The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors 244 Olena Andrukhova et al FGF23 regulates renal TRPV5 The EMBO Journal Burg MB, Grantham J, Abramow M, Orloff J, Schafer JA (1997) Preparation Imura A, Tsuji Y, Murata M, Maeda R, Kubota K, Iwano A, Obuse C, Togashi K, and study of fragments of single rabbit nephrons. J Am Soc Nephrol 8: Tominaga M, Kita N, Tomiyama K, Iijima J, Nabeshima Y, Fujioka M, Asato 675 – 683 R, Tanaka S, Kojima K, Ito J, Nozaki K, Hashimoto N et al (2007) Cai H, Cebotaru V, Wang YH, Zhang XM, Cebotaru L, Guggino SE, Guggino alpha-Klotho as a regulator of calcium homeostasis. Science 316: WB (2006) WNK4 kinase regulates surface expression of the human 1615 – 1618 sodium chloride cotransporter in mammalian cells. Kidney Int 69: Jiang Y, Cong P, Williams SR, Zhang W, Na T, Ma HP, Peng JB (2008) WNK4 2162 – 2170 regulates the secretory pathway via which TRPV5 is targeted to the Cha SK, Huang CL (2010) WNK4 kinase stimulates caveola-mediated plasma membrane. Biochem Biophys Res Commun 375: 225 – 229 endocytosis of TRPV5 amplifying the dynamic range of regulation of the Jiang Y, Ferguson WB, Peng JB (2007) WNK4 enhances TRPV5-mediated channel by protein kinase C. J Biol Chem 285: 6604 – 6611 calcium transport: potential role in hypercalciuria of familial hyperkalemic Cha SK, Ortega B, Kurosu H, Rosenblatt KP, Kuro O, Huang CL (2008) hypertension caused by gene mutation of WNK4. Am J Physiol Renal Removal of sialic acid involving Klotho causes cell-surface retention of Physiol 292:F545 – F554 TRPV5 channel via binding to galectin-1. Proc Natl Acad Sci USA 105: Juppner H, Wolf M, Salusky IB (2010) FGF-23: more than a regulator of renal 9805 – 9810 phosphate handling? J Bone Miner Res 25: 2091 – 2097 Chang Q, Hoefs S, Van Der Kemp AW, Topala CN, Bindels RJ, Hoenderop JG Kessler W, Budde T, Gekle M, Fabian A, Schwab A (2008) Activation of cell (2005) The beta-glucuronidase klotho hydrolyzes and activates the TRPV5 migration with fibroblast growth factor-2 requires calcium-sensitive channel. Science 310: 490 – 493 potassium channels. Pflugers Arch 456: 813 – 823 Erben RG, Soegiarto DW, Weber K, Zeitz U, Lieberherr M, Gniadecki R, M€ oller Kuro-o M, Matsumura Y, Aizawa H, Kawaguchi H, Suga T, Utsugi T, Ohyama G, Adamski J, Balling R (2002) Deletion of deoxyribonucleic acid binding Y, Kurabayashi M, Kaname T, Kume E, Iwasaki H, Iida A, Shiraki-Iida T, domain of the vitamin D receptor abrogates genomic and nongenomic Nishikawa S, Nagai R, Nabeshima YI (1997) Mutation of the mouse klotho functions of vitamin D. Mol Endocrinol 16: 1524 – 1537 gene leads to a syndrome resembling ageing. Nature 390: 45 – 51 Farrow EG, Davis SI, Summers LJ, White KE (2009) Initial FGF23-mediated Kurosu H, Ogawa Y, Miyoshi M, Yamamoto M, Nandi A, Rosenblatt KP, Baum signaling occurs in the distal convoluted tubule. J Am Soc Nephrol 20: MG, Schiavi S, Hu MC, Moe OW, Kuro O (2006) Regulation of fibroblast 955 – 960 growth factor-23 signaling by Klotho. J Biol Chem 281: 6120 – 6123 Farrow EG, Summers LJ, Schiavi SC, McCormick JA, Ellison DH, White KE Kurosu H, Yamamoto M, Clark JD, Pastor JV, Nandi A, Gurnani P, McGuinness (2010) Altered renal FGF23-mediated activity involving MAPK and Wnt: OP, Chikuda H, Yamaguchi M, Kawaguchi H, Shimomura I, Takayama Y, effects of the Hyp mutation. J Endocrinol 207: 67 – 75 Herz J, Kahn CR, Rosenblatt KP, Kuro-o M (2005) Suppression of aging in Fischer SS, Kempe DS, Leibrock CB, Rexhepaj R, Siraskar B, Boini KM, mice by the hormone Klotho. Science 309: 1829 – 1833 Ackermann TF, Foller M, Hocher B, Rosenblatt KP, Kuro O, Lang F (2010) Lambers TT, Bindels RJ, Hoenderop JG (2006) Coordinated control of renal Hyperaldosteronism in Klotho-deficient mice. Am J Physiol Renal Physiol Ca2 + handling. Kidney Int 69: 650 – 654 299:F1171 – F1177 Li H, Martin A, David V, Quarles LD (2011) Compound deletion of Fgfr3 and Goetz R, Beenken A, Ibrahimi OA, Kalinina J, Olsen SK, Eliseenkova AV, Xu C, Fgfr4 partially rescues the Hyp mouse phenotype. Am J Physiol Endocrinol Neubert TA, Zhang F, Linhardt RJ, Yu X, White KE, Inagaki T, Kliewer SA, Metab 300:E508 – E517 Yamamoto M, Kurosu H, Ogawa Y, Kuro-o M, Lanske B, Razzaque MS et al Li YC, Amling M, Pirro AE, Priemel M, Meuse J, Baron R, Delling G, Demay MB (2007) Molecular insights into the klotho-dependent, endocrine mode of (1998) Normalization of mineral ion homeostasis by dietary means action of fibroblast growth factor 19 subfamily members. Mol Cell Biol 27: prevents hyperparathyroidism, rickets, and osteomalacia, but not alopecia 3417 – 3428 in vitamin D receptor-ablated mice. Endocrinology 139: 4391 – 4396 Gonzalez-Mariscal L, Namorado MC, Martin D, Luna J, Alarcon L, Islas S, Martin A, David V, Quarles LD (2012) Regulation and function of the FGF23/ Valencia L, Muriel P, Ponce L, Reyes JL (2000) Tight junction proteins klotho endocrine pathways. Physiol Rev 92: 131 – 155 ZO-1, ZO-2, and occludin along isolated renal tubules. Kidney Int 57: Matsumura Y, Aizawa H, Shiraki-Iida T, Nagai R, Kuro-o M, Nabeshima Y 2386 – 2402 (1998) Identification of the human klotho gene and its two transcripts de Groot T, Lee K, Langeslag M, Xi Q, Jalink K, Bindels RJ, Hoenderop JG (2009) encoding membrane and secreted klotho protein. Biochem Biophys Res Parathyroid hormone activates TRPV5 via PKA-dependent phosphorylation. Commun 242: 626 – 630 J Am Soc Nephrol 20: 1693 – 1704 Mayan H, Vered I, Mouallem M, Tzadok-Witkon M, Pauzner R, Farfel Z (2002) He G, Wang HR, Huang SK, Huang CL (2007) Intersectin links WNK kinases to Pseudohypoaldosteronism type II: marked sensitivity to thiazides, endocytosis of ROMK1. J Clin Invest 117: 1078 – 1087 hypercalciuria, normomagnesemia, and low bone mineral density. J Clin Hellwig N, Albrecht N, Harteneck C, Schultz G, Schaefer M (2005) Homo- and Endocrinol Metab 87: 3248 – 3254 heteromeric assembly of TRPV channel subunits. J Cell Sci 118: 917 – 928 McCormick JA, Yang CL, Ellison DH (2008) WNK kinases and renal sodium Hesse M, Frohlich LF, Zeitz U, Lanske B, Erben RG (2007) Ablation of vitamin transport in health and disease: an integrated view. Hypertension 51: D signaling rescues bone, mineral, and glucose homeostasis in Fgf-23 588 – 596 deficient mice. Matrix Biol 26: 75 – 84 Michlig S, Mercier A, Doucet A, Schild L, Horisberger JD, Rossier BC, Firsov D Hoenderop JG, Vennekens R, Muller D, Prenen J, Droogmans G, Bindels RJ, (2004) ERK1/2 controls Na, K-ATPase activity and transepithelial sodium Nilius B (2001) Function and expression of the epithelial Ca(2+) channel transport in the principal cell of the cortical collecting duct of the mouse family: comparison of mammalian ECaC1 and 2. J Physiol 537: 747 – 761 kidney. J Biol Chem 279: 51002 – 51012 Hu MC, Shi M, Zhang J, Pastor J, Nakatani T, Lanske B, Razzaque MS, Nakatani T, Sarraj B, Ohnishi M, Densmore MJ, Taguchi T, Goetz R, Rosenblatt KP, Baum MG, Kuro-o M, Moe OW (2010) Klotho: a novel Mohammadi M, Lanske B, Razzaque MS (2009) In vivo genetic evidence phosphaturic substance acting as an autocrine enzyme in the renal for klotho-dependent, fibroblast growth factor 23 (Fgf23) -mediated proximal tubule. FASEB J 24: 3438 – 3450 regulation of systemic phosphate homeostasis. FASEB J 23: 433 – 441 ª 2014 The Authors The EMBO Journal Vol 33 |No 3 | 2014 245 The EMBO Journal FGF23 regulates renal TRPV5 Olena Andrukhova et al Okano T, Tsugawa N, Morishita A, Kato S (2004) Regulation of gene factor-23 results in hyperphosphatemia and impaired skeletogenesis, and expression of epithelial calcium channels in intestine and kidney of mice reverses hypophosphatemia in Phex-deficient mice. Matrix Biol 23: by 1alpha,25-dihydroxyvitamin D3. J Steroid Biochem Mol Biol 89–90: 421 – 432 335 – 338 Streicher C, Zeitz U, Andrukhova O, Rupprecht A, Pohl E, Larsson TE, Ring AM, Leng Q, Rinehart J, Wilson FH, Kahle KT, Hebert SC, Lifton RP (2007) Windisch W, Lanske B, Erben RG (2012) Long-term Fgf23 deficiency An SGK1 site in WNK4 regulates Na+ channel and K+ channel activity does not influence aging, glucose homeostasis, or fat metabolism in and has implications for aldosterone signaling and K+ homeostasis. Proc mice with a nonfunctioning vitamin D receptor. Endocrinology 153: Natl Acad Sci USA 104: 4025 – 4029 1795 – 1805 Sandulache D, Grahammer F, Artunc F, Henke G, Hussain A, Nasir O, Mack A, Tohyama O, Imura A, Iwano A, Freund JN, Henrissat B, Fujimori T, Nabeshima Friedrich B, Vallon V, Wulff P, Kuhl D, Palmada M, Lang F (2006) Y(2004) Klotho is a novel beta-glucuronidase capable of hydrolyzing Renal Ca2 + handling in sgk1 knockout mice. Pflugers Arch 452: steroid beta-glucuronides. J Biol Chem 279: 9777 – 9784 444 – 452 Tsujikawa H, Kurotaki Y, Fujimori T, Fukuda K, Nabeshima Y (2003) Klotho, a Schafer JA, Watkins ML, Li L, Herter P, Haxelmans S, Schlatter E (1997)A gene related to a syndrome resembling human premature aging, simplified method for isolation of large numbers of defined nephron functions in a negative regulatory circuit of vitamin D endocrine system. segments. Am J Physiol 273:F650 – F657 Mol Endocrinol 17: 2393 – 2403 Segawa H, Yamanaka S, Ohno Y, Onitsuka A, Shiozawa K, Aranami F, Furutani Urakawa I, Yamazaki Y, Shimada T, Iijima K, Hasegawa H, Okawa K, Fujita T, J, Tomoe Y, Ito M, Kuwahata M, Imura A, Nabeshima Y, Miyamoto K Fukumoto S, Yamashita T (2006) Klotho converts canonical FGF receptor (2007) Correlation between hyperphosphatemia and type II Na-Pi into a specific receptor for FGF23. Nature 444: 770 – 774 cotransporter activity in klotho mice. Am J Physiol Renal Physiol 292: Weber K, Erben RG, Rump A, Adamski J (2001) Gene structure and regulation F769 – F779 of the murine epithelial calcium channels ECaC1 and 2. Biochem Biophys Shimada T, Hasegawa H, Yamazaki Y, Muto T, Hino R, Takeuchi Y, Fujita T, Res Commun 289: 1287 – 1294 Nakahara K, Fukumoto S, Yamashita T (2004) FGF-23 is a potent regulator Weinman EJ, Steplock D, Shenolikar S, Biswas R (2011) FGF-23-mediated of vitamin D metabolism and phosphate homeostasis. J Bone Miner Res inhibition of renal phosphate transport in mice requires NHERF-1 and 19: 429 – 435 synergizes with PTH. J Biol Chem 286: 37216 – 37221 Shimada T, Mizutani S, Muto T, Yoneya T, Hino R, Takeda S, Takeuchi Y, White KE, Econs MJ (2008) Fibroblast Growth Factor-23.In Primer on the Fujita T, Fukumoto S, Yamashita T (2001) Cloning and characterization of Metabolic Bone Diseases and Disorders of Mineral Metabolism, Rosen CJ FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl (ed), pp 112 – 116. Washington, DC: Society of Bone and Mineral Acad Sci USA 98: 6500 – 6505 Research Shiraki-Iida T, Aizawa H, Matsumura Y, Sekine S, Iida A, Anazawa H, Nagai R, Woudenberg-Vrenken TE, Bindels RJ, Hoenderop JG (2009) The role of Kuro-o M, Nabeshima Y (1998) Structure of the mouse klotho gene and its transient receptor potential channels in kidney disease. Nat Rev Nephrol 5: two transcripts encoding membrane and secreted protein. FEBS Lett 424: 441 – 449 6 – 10 Zeitz U, Weber K, Soegiarto DW, Wolf E, Balling R, Erben RG (2003) Impaired Sitara D, Razzaque MS, Hesse M, Yoganathan S, Taguchi T, Erben RG, insulin secretory capacity in mice lacking a functional vitamin D receptor. Juppner H, Lanske B (2004) Homozygous ablation of fibroblast growth FASEB J 17: 509 – 511 The EMBO Journal Vol 33 |No 3 | 2014 ª 2014 The Authors

Journal

The EMBO Journal – Springer Journals

Published: Feb 3, 2014

Keywords: calcium homeostasis; fibroblast growth factor‐23; Klotho; renal calcium reabsorption; transient receptor potential vanilloid‐5

There are no references for this article.