Conditional Deletion of Fgfr1 in the Proximal and Distal Tubule Identifies Distinct Roles in Phosphate and Calcium Transport

A postnatal role of fibroblast growth factor receptor-1 (FGFR1) in the kidney is suggested by its binding to α-Klotho to form an obligate receptor for the hormone fibroblast growth factor-23 (FGF-23). FGFR1 is expressed in both the proximal and distal renal tubular segments, but its tubular specific functions are unclear. In this study, we crossed Fgfr1flox/flox mice with either gamma-glutamyltransferase-Cre (γGT-Cre) or kidney specific-Cre (Ksp-Cre) mice to selectively create proximal tubule (PT) and distal tubule (DT) Fgfr1 conditional knockout mice (designated Fgfr1PT-cKO and Fgfr1DT-cKO, respectively). Fgfr1PT-cKO mice exhibited an increase in sodium-dependent phosphate co-transporter expression, hyperphosphatemia, and refractoriness to the phosphaturic actions of FGF-23, consistent with a direct role of FGFR1 in mediating the proximal tubular phosphate responses to FGF-23. In contrast, Fgfr1DT-cKO mice unexpectedly developed hypercalciuria, secondary elevations of parathyroid hormone (PTH), hypophosphatemia and enhanced urinary phosphate excretion. Fgfr1PT-cKO mice also developed a curly tail/spina bifida-like skeletal phenotype, whereas Fgfr1DT-cKO mice developed renal tubular micro-calcifications and reductions in cortical bone thickness. Thus, FGFR1 has dual functions to directly regulate proximal and distal tubule phosphate and calcium reabsorption, indicating a physiological role of FGFR1 signaling in both phosphate and calcium homeostasis.


Introduction
There are 22 mammalian fibroblastic growth factors (FGFs) subgrouped into 7 subfamilies and four alternative spliced FGF receptor genes (FGFR1-FGFR4) that encode seven membraneassociated tyrosine kinase isoforms (FGFRs 1 b, 1c, 2b, 2c, 3b, 3c and 4). The majority of FGFs are retained in tissues and act as paracrine/autocrine activators of FGFRs in the presence of heparin sulfate proteoglycans. The more recently evolved subfamily of hormonal FGFs, include FGF-19, FGF-21 and FGF-23, have N-terminal FGF-homology domains linked to novel C-termini [1][2][3] that allow entry into the circulating and binding to a receptor complex formed by FGFRs and Klotho (KL), a type I membrane ß-glycosidase-like protein.
There a several knowledge gaps in FGF-23 regulation of renal tubular functions. For example, specific FGFRs that mediate the effects of FGF-23 in the kidney, the necessity for α-KL coexpression, and the precise tubular functions of FGFR activation by FGF-23 remain unclear. FGFR1c, 3c or 4, but not FGFR2, can form complexes with α-KL to constitute a functional FGF-23 receptor [13][14][15][16]. Immunohistochemistry analysis of the kidney originally identified FGFR3 in the proximal tubule [17,18], FGFR-1, -3 and -4 in the distal tubules, and FGFR2 in the distal straight tubules [18]; however, recent studies using more sensitive RT-PCR found that FGFR-1 and -4 transcripts are also expressed in the proximal tubule [9,19]. Mouse genetic approaches have so far failed to resolve the question of tubular specific functions of FGFRs [9,20,21]. Whereas Fgfr3 -/or Fgfr4 -/mice have no kidney phenotype, compound Fgfr3 -/-/Fgfr4 -/mice have increased serum 1,25(OH) 2 D levels, are completely resistant to FGF-23-mediated suppression of 1,25(OH) 2 D, and partially resistant to its phosphaturic actions [18,20,21]. Fgfr1 -/mice are embryonic lethal, but conditional deletion of FGFR1 in both the PT and DT using Pax3-Cre resulted in partial inhibition of phosphaturic responses to rFGF-23 associated with an increase in Npt2c expression in the kidney, but no effect on calcium or vitamin D metabolism [9,22]. Concomitant loss of FGFR4 and FGFR1 in Pax3-Cre/Fgfr1 cKO mice were required to disturb 1,25(OH) 2 D metabolism. While these studies show FGFRs have both overlapping and distinct functions in the kidney, they fail to define the tubular specific functions of individual receptors.
Uncertainty about the tubular expression of α-KL, the FGFR co-receptor, which imparts both function and cell selectivity to FGF-23 [13], further confounds the analysis of FGFR function in the kidney. Indeed, α-KL expression is mainly expressed in the distal tubule, but the principal function of FGF-23 is in the proximal tubule, a disparity leading to the conjecture of a distal-to-proximal paracrine feedback mechanism responsible for the proximal tubular effects of FGF-23 [20]. Subsequently, α-Kl transcripts were detected in the dissected PT segments, suggesting that FGF-23 directly activates FGFR/α-KL complexes in the proximal tubule [23], but whether FGF-23 directly activates FGFRs in the proximal tubule remains controversial. Indeed, signally responses after rFGF-23 administration are limited to the distal tubule [24]. Moreover, conditional deletion of α-KL from the DT using Ksp-Cre (α-Kl DT-cKO ) results in hyperphosphatemia and increased expression of vitamin D regulatory enzymes in association with elevated FGF-23 levels, indicating that loss of α-KL in the distal tubule can affect proximal tubule functions [25].
Studies of Fgf-23 -/and α-Kl -/mice also implicate FGF-23 in the regulation of distal tubular calcium transport, but the mechanism is controversial and the specific FGFRs mediating FGF-23 effects on distal tubule calcium have not been defined. FGF-23 activation of FGFR/α-KL coreceptors in the distal tubule is purported to stimulate TRPV5 trafficking and membrane insertion leading to increases calcium absorption [12]. On the other hand, FGF-23 suppresses the expression of α-KL, which stabilizes the membrane expression of TRPV5 through its glucuronidase/sialidase activity [26]. FGF-23-mediated reductions in α-KL and TRPV5 insertion should also inhibit distal tubular calcium reabsorption. Elucidating the mechanisms of FGF-23 effects on the distal tubule is confounded by limitations of Fgf-23 -/and Kl -/mouse models, which have systemic alterations in 1,25(OH) 2 D and PTH that preclude distinguishing between FGF-23 dependent and independent effects on distal tubule calcium transport.
In the current study, we tested the selective function of FGFR1 in the proximal and distal tubule and its role in mediating the effects of FGF-23 on proximal and distal tubular functions in mice with conditionally deletion of Fgfr1 from the either proximal or distal nephron segments. We confirm an important direct role of FGFR1 signaling in the regulation of phosphate transport in the proximal tubule. We also found an important role of FGFR1 in regulating distal tubule calcium transport. This suggests a new schema for FGF-23 physiological functions to coordinately regulate proximal and distal functions to maintain phosphate and calcium homeostasis.

Animals breeding and genotyping
All animal research was conducted according to guidelines provided by the National Institutes of Health and the Institute of Laboratory Animal Resources, National Research Council. The University of Tennessee Health Science Center's Animal Care and Use Committee approved all animal studies (Protocol number: 12-168.0). CMV-Cre mice were originally purchased from the Jackson Laboratory (Bar Harbor, ME, USA) and maintained in C57BL/6J background. The floxed Fgfr1 mice (Fgfr1 flox/flox ) were obtained from Dr. Chuxia Deng at National Institute of Diabetes and Digestive and Kidney Diseases and maintained in C57BL/6J background for at least six generations. γGT-Cre [27] and Ksp-Cre [28] mice were used to delete the floxed Fgfr1 in kidney as described previously and maintained in C57BL/6J background for at least five generations. All mice were maintained on a standard diet (7912, Harlan Teklad, Madison, WI, USA). First, we crossed Fgfr1 flox/+ to CMV-Cre to obtain a germline-specific deletion of Fgfr1 (Fgfr1 null/+ ). The Fgfr1 null/+ mice were crossed to γGT-Cre mice or Ksp-Cre mice to obtain heterozygous γGT-Cre;Fgfr1 null/+ or Ksp-Cre;Fgfr1 null/+ mice, respectively. Then, Fgfr1 flox/flox females were crossed to γGT-Cre;Fgfr1 null/+ males or Ksp-Cre;Fgfr1 null/+ males to obtain the kidney-specific deletion of Fgfr1 in proximal tubule or distal tubule of these mice. For the entire study, samples were collected from 4-month-old Fgfr1 flox/+ (control equivalent) control, conditional γGT-Cre;Fgfr1 null/flox (Fgfr1 PT-cKO )-null mice or Ksp-Cre;Fgfr1 null/flox (Fgfr1 DT-cKO )-null mice, respectively. Tail clips were collected to genotype the mice. REDExtract-N-Amp Tissue PCR Kit (Sigma-Aldrich, St. Louis, MO, USA) was used for DNA extraction and PCR amplification. Mice were genotyped for the Fgfr1 flox allele using forward primer 5 0 -CTG GTA TCC TGT GCC TAT C-3 0 and reverse primer 5 0 -CCA ATC TGA TCC CAA GAC CAC-3 0 (325 bp product for the Fgfr1+ control allele, 400 bp product for the Fgfr1flox floxed allele), and for the Fgfr1null allele using forward primer 5 0 -GTA TTG CTG GCC CAC TGT TC-3 0 and reverse primer 5 0 -CCA ATC TGA TCC CAA GAC CAC-3 0 (300 bp product for the Fgfr1 null null allele). We collected 12 h urine from control, Fgfr1 PT-cKO and Fgfr1 DT-cKO mice. We also collected 12 h urine from control, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice injected with vehicle (PBS) or rFGF-23 (100 ng/g/mouse). Mice were euthanized by exposing overdose of isoflurane followed by cervical dislocation. Power calculations to determine minimal sample size were performed as previously described [29]. All animal procedures are approved by The University of Tennessee Health Science Center's Animal Care and Use Committee (Protocol number: 12-168.0).

Serum and urine biochemistry
Blood samples were collected by retro-orbital bleeding before the endpoint and by intracardiac exsanguinations at the endpoint. Urine samples were collected overnight (from 6:00 pm to 6:00 am) in mice housed in metabolic cages. Calcium was measured using a Calcium CPC Liquicolor Kit (Stanbio Laboratories, Boerne, TX, USA) and phosphorus was measured using the phosphomolybdylate-ascorbic acid method. Serum parathyroid hormone (PTH) levels were measured using the Mouse Intact PTH ELISA kit (Immutopics, Carlsbad, CA, USA). Urea nitrogen (BUN), and creatinine were measured using individual kits from Stanbio laboratories. Serum 1,25 dihydroxy vitamin D (1,25(OH) 2 D) was measured using a ELISA kit from MyBio-Source.com. Serum full-length fibroblastic growth factor 23 (FGF-23) levels were measured using the FGF-23 ELISA kit (Kainos Laboratories, Tokyo, Japan).

Immunohistochemistry
After fixation in 10% formalin neutral buffered solution immediately after sacrifice, kidneys were embedded in paraffin. Immunohistochemistry was performed using an UltraSensitive ABC Rabbit IgG staining kit and following the manufacturer's specifications (ThermoScientific, Rockford, IL). Samples were rehydrated in decreasing ethanol solutions, then immersed in phosphate buffered saline for 10 min followed by quenching in 3% hydrogen peroxide. Slides were blocked for 30 minutes in the kit's blocking solution. Primary antibodies [Klotho antibody, 1:100 from R&D Laboratory, Minneapolis, MN; FGFR1 antibody, 1:100 from Abcam, Cambridge, MA; Npt2a antibody, 1:50 from Alpha Diagnostic International, San Antonio, TX [30], TRPV5 antibody 1:1000 from Alomone Labs, Jerusalem, Israel) were applied to slides and incubated for an hour. Secondary horseradish peroxidase (HRP)-conjugated antibodies were applied following another PBS wash and incubated for 1 hour as indicated. Slides were rewashed for 10 min with PBS, and then the ABC reagent was applied for 30 minutes. Immunostaining was detected using a Metal Enhanced DAB Substrate Kit (Thermo Scientific, Rockford, IL). For Klotho and TRPV5 staining, the secondary antibody was Alexa Fluor 568 1:1000, or 488 1:1000 dye, respectively (Thermo Scientific). Controls were performed by omitting primary antibodies. DAPI was used for counterstaining of the kidney sections.

Von Kossa and Alizarin Red S staining
Kidney sections of mice were stained for calcium deposition using a Von Kossa Stain kit following manufactures' instructions (American Mastertech Scientific, Lodi, CA). Alizarin Red S staining was performed as described previously [31].

High resolution 3D microtomography
The femurs were collected, fixed and dehydrated in 70% ethanol. High-resolution micro-Computed Tomography (μCT40, Scanco Medical, Basserdorf, Switzerland) was used to scan and evaluate the metaphyseal trabecular bone microarchitecture and the midshaft cortical bone parameters. The entire femurs and tibia were scanned in a 12.3 mm diameter sample holder at 6 μm resolution: energy level of 55 KeV and intensity of 145 μA. The trabecular bone volume (mm 3 ) and bone volume fraction (BV/TV) was measured within the secondary spongiosa on a set of 50 sections (0.6 mm) underneath the growth plate at a threshold of 200 as previously described [32]. The cortical bone thickness (CtTh, mm) was analyzed from 100 sections chosen at the midshaft of each femur at a threshold of 350.

Kidney RNA isolation and real-time reverse transcriptase (RT)-qPCR
Total RNA was isolated from whole kidney of mouse at 4 months of age using an RNeasy Mini Kit (Qiagen, Germany). For quantitative real-time RT-PCR, 1.0 μg total RNA isolated from kidney of three genotypes mice was reverse transcribed using an iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) by following the manufacturer's instructions. PCR reactions contained 1 μl of cDNA (equivalent to 50 ng of total RNA), 300 nM each primers, and 1× iQ SYBR Green supermix (Bio-Rad, Hercules, CA, USA) in a total of 25 μl reaction volume performed with CFX96 Real-Time PCR Detection Systems (Bio-Rad). Relative expression values were evaluated with the 2 -ΔΔ Ct method using GAPDH as housekeeping gene (forward primer: 5'-CAC CACCAACTGCTTAGCC -3', and reverse primer:

Western blot analysis
Kidney tissue (~10 mg) was transferred into T-PER tissue protein extraction reagent (Thermo Scientific, Rockford, IL) and 1x protease inhibitor cocktail with 1mM phenylmethylsulfonyl fluoride (PMSF)(Cell Signaling, Danvers, MA, USA). After three 30-second sonication, samples were centrifuged at 13,000x g for 10 minute and protein contents in the supernatants were quantified and samples were stored at -80°C until use. Total cell membrane protein was isolated as described previously [12]. For electrophoresis, samples were prepared by mixing 3x SDS loading buffer (Cell signaling) with 1x DTT. About 50 μg of protein were loaded onto NuPAGE 4-12% Bis-Tris Gel (Invitrogen, Carlsbad, CA, USA). Proteins were separated at 150 V for 60 minutes and transferred to nitrocellulose membrane (Invitrogen, Carlsbad, CA). Membranes were blocked with Superblock blocking buffer in TBST (Thermo Scientific, Rockford, IL, USA) for 30 minutes and then incubated with primary antibodies (Klotho antibody, 1:500, from Gwen King's lab; FGFR1 antibody 1:1000, Cell Signaling Technology, Danvers, MA; TRPV5 1:1000 from Santa Cruz Biotechnology, Dallas, TX; and, Npt2a 1:1000, Alpha Diagnostic International, San Antonio, TX [33]) with gentle agitation overnight at 4°C. After 3 washes with TBST (15 min once and 2x 5 min), membrane was incubated with secondary antibody in Superblock blocking buffer at room temperature for 1 hour. Membrane was then washed 4 times (15 min and 3x 5 min) and subjected to ECL (Thermo Scientific) and analyzed with the FOTO/Analyst Luminary/FX imaging workstation (FOTODYNE INCORPORATED, Hartland, WI, USA). The intensity of bands was quantified using Image J software (http://rsb. info.nih.gov/ij/).

Statistics
We evaluated differences between two groups by unpaired t-test and multiple groups by oneway ANOVA followed by Tukey's post-test for multiple comparisons. All values are expressed as means ± SD. All statistical tests are performed with an alpha of 0.05 as the significance threshold. All computations were performed using GraphPad Prism5 (GraphPad Software Inc. La Jolla, CA, USA).

Results
Proximal and distal tubule specific deletion of Fgfr1 in the kidney Gamma-glutamyltransferase-Cre (γGT-cre) and kidney specific-Cre (Ksp-Cre) to conditionally delete Fgfr1 in proximal tubule or distal tubule of the kidney successfully created Fgfr1 PT-cKO and Fgfr1 DT-cKO mice (Fig 1A), respectively. Both Fgfr1 PT-cKO mice and Fgfr1 DT-cKO mice were born with the expected Mendelian frequency. There were no differences in survival between the genotypes produced by this breeding strategy. All mice survived to adulthood.
To confirm γGT-Cre and Ksp-Cre mediated recombination occurred selectively in the kidney we performed PCR analysis of genomic DNA using primer sets that specifically detect the excised flox alleles. We found γGT-Cre and Ksp-Cre excision of the Fgfr1 floxed alleles occurred in the kidney in both Fgfr1 PT-cKO mice ( Fig 1B) and Fgfr1 DT-cKO mice ( Fig 1C). We found no evidence of the excised flox alleles in a wide variety of other tissues tested using Ksp-Cre, but found expression of γGT-Cre in a sample obtained from spine but not long bone.
Expression of Ffgr1 message and protein levels were reduced by~75% in the kidney of both Fgfr1 PT-cKO and Fgfr1 DT-cKO mice as determined by qRT-PCR ( Fig 1D) and Western blot analysis (Fig 1E and 1F). We confirmed the specific deletion of FGFR1 in the proximal tubule of Fgfr1 PT-cKO mice and in the distal tubule of Fgfr1 DT-cKO mouse by immunohistochemical staining with an anti-FGFR1 antibody (Fig 1G). FGFR1 immuno-reactivity was detected in both distal and proximal tubules in control mice (Fig 1G), but FGFR1 expression was selectively reduced in the morphologically distinct proximal tubular segments in Fgfr1 PT-cKO mice and in the distal tubular segments in Fgfr1 DT-cKO mice, respectively (Fig 1G). Differences in FGFR1 expression in morphologically distinct proximal and distal tubular segments are apparent in γGT-Cre and Ksp-Cre conditional knockout mice (S1 Fig). Indeed, FGFR1 staining was only seen in the distal tubules of Fgfr1 PT-cKO mice and in the proximal tubules of Fgfr1 DT-cKO mice (Fig 1G and S1 Fig). FGFR1 staining was most predominant on the luminal membranes S1 Fig).
Fgfr1 PT-cKO mice exhibited gross skeletal abnormalities characterized by curled tails, smaller body size and reductions in body weight (Fig 1H, 1I and 1J). Fgfr1 DT-cKO mice exhibited no gross abnormalities and body weight no different from control mice (Fig 1K, 1L and 1M).

Serum biochemistries in mice with proximal and distal tubule deletion of Fgfr1
We observed no significant differences in circulating concentrations of FGF-23, calcium, or 1,25(OH) 2 D, between control, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice under basal conditions (Fig  2A, 2B and 2E). Serum BUN were not different between Fgfr1 PT-cKO , Fgfr1 DT-cKO and control mice ( Fig 2F). We found that the level of serum phosphate was significantly higher (p<0.05) in Fgfr1 PT-cKO mice and significantly lower (p<0.05) (Fig 2C) in Fgfr1 DT-cKO compared to controls. In addition, serum PTH levels were similar to controls in Fgfr1 PT-cKO mice, but significantly increased in Fgfr1 DT-cKO mice ( Fig 2D).
Urinary calcium and phosphate excretion were not different between Fgfr1 PT-cKO and controls; however, both urinary calcium and phosphate excretion were significantly higher in Fgfr1 DT-cKO mice than control mice (p<0.05) (Fig 3A and 3B). Urine volume and urine creatinine levels were similar in control, Fgfr1 PT-cKO, and Fgfr1 DT-cKO mice (Fig 3C and 3D). Both Fgfr1 PT-cKO and Fgfr1 DT-cKO had normal renal architecture (Fig 3E), indicating that loss of FGFR1 had no effect on kidney development.
Kidney micro-calcifications and abnormal bone structure in Fgfr1 PT-cKO and Fgfr1 DT-cKO mice Kidney sections from Fgfr1 DT-cKO mice showed evidence of small dense deposits foci in the tubular lumen mainly in the medullary regions that were positively by both Von Kossa and Alizarin Red staining (Fig 3F). We observed discrete intratubular dense deposits in the distal tubules in Fgfr1 DT-cKO but not control mice. The intra-luminal deposits were not associated with cell injury or inflammation. We also did not observed soft-tissue calcification of the kidneys by μCT analysis (data not shown). These unusual appearing structures resembled calcium phosphate microcrystal masses previously observed in renal tubules following intestinal resection in rats [34].
We further characterized the skeletal abnormalities in Fgfr1 PT-cKO mice by X-ray and μCT analysis (Fig 4A). The radiograph and three dimensional spine images demonstrated that the spinal deformities of vertebrae. The involved vertebral bodies lost normal anatomic structures with irregular wedge-shaped morphology or fusion of several segments. Curly tails and vertebral deformities were observed in~88% and~44% of Fgfr1 PT-cKO mice, indicating incomplete penetrance. Since the spinal defect resembles spina bifida, which can be caused by folic acid deficiency, we measured serum folic acid in Fgfr1 PT-cKO mice. Serum folic acid concentrations (mean± SEM, ng/ml) were not different between Fgfr1 PT-cKO (11.64±2.63), Fgfr1 DT-cKO (10.72 ±0.88), and control (11.04±1.18) mice. In addition to the skeletal deformities of the spine, we found that both trabecular bone volume and trabecular bone fraction of femoral bone were significantly decreased in Fgfr1 PT-cKO mice (Table 1 and Fig 4B). In contrast, Fgfr1 DT-cKO mice exhibited no gross abnormalities (Fig 1K). However, trabecular thickness, trabecular material, and cortical thickness of femur bone was significantly reduced in Fgfr1 DT-cKO mice compared to c006Fntrols (Table 1 and Fig 4B).   Effect of Fgfr1 conditional deletion on proximal and distal tubule gene expression Next, we examined the effects of proximal tubule deletion of Fgfr1 on renal tubular transporters involved phosphate transport and vitamin D metabolism (Fig 5). In Fgfr1 PT-cKO mice, messenger RNA levels for the sodium phosphate cotransporters Npt2a and Npt2c were increased 1.5-fold and 2.5-fold, respectively (Fig 5A and 5B); but only expression of Npt2c achieved significance compared to controls (p<0.05). Expression of 25-hydroxyvitamin D-1 alpha hydroxylase (Cyp27b1) was not affected, but expression of Cyp24a1 was significantly reduced (p<0.01) in Fgfr1 PT-cKO mice compared to controls (Fig 5C and 5D). In spite of observing no alterations in urinary calcium compared to controls in Fgfr1 PT-cKO mice, loss-of-Fgfr1 in the PT resulted in significant increments message levels for transient receptor potential cation channel subfamily V member 5 (TRPV5), TRPV6, and calcium binding protein 28k (CaBP28k) in the kidney (p<0.05) (Fig 5E, 5F, and 5G). Levels of sodium-calcium exchanger (NCX1), plasma  membrane calcium pump-1b (PMCA1b), and α-Kl transcripts, however, were not changed in Fgfr1 PT-cKO mice compared to control (Fig 5H, 5I, and 5J).
In Fgfr1 DT-cKO mice, the expression of sodium phosphate cotransporters mRNA, Npt2a and Npt2c, were both significantly decreased compared to controls (p<0.05) (Fig 5A and 5B). Although expression of Cyp27b1 was not affected, expression of Cyp24a1 was significantly reduced in Fgfr1 DT-cKO mice compared to controls (p<0.01) (Fig 5C and 5D). Since PTH suppresses sodium phosphate transporters and inhibits Cyp24a1, the observed changes in gene expression are most likely indirectly due to the effects of elevated serum PTH on the proximal tubule.
We also examined the effects of distal tubule deletion of FGFR1 on renal tubular transporters involved in calcium transport (Fig 5E-5J). In spite of the increased urinary calcium, we found that expression of messenger RNA levels for TRPV5, TRPV6, CaBP28k, NCX1, and PMCA1b were significantly increased in Fgfr1 DT-cKO compared to control mice (p<0.05) ( Fig  5E-5I). In addition, α-Kl transcripts were significantly decreased in Fgfr1 DT-cKO kidneys compared to controls.

Expression of TRPV5 and α-KL proteins
We evaluated the expression of α-Klotho (α-KL) and TRPV5 using proteins isolated from whole kidney and kidney cell membrane. We found that expression of α-KL in the whole kidney protein was not affected in Fgfr1 PT-cKO or Fgfr1 DT-cKO mice (Fig 6A and 6B). The increase in TRPV5 message expression observed in Fgfr1 PT-cKO was associated with only a slight increase in TRPV5 protein by Western blot analysis (Fig 6C and 6D). Interestingly, we found that expression of both Klotho isoforms (α-KL and s-KL) and TRPV5 in isolated kidney membrane proteins were significantly (p<0.05) decreased in Fgfr1 DT-cKO mice as compared to that of WT or Fgfr1 PT-cKO mice, respectively (Fig 6E, 6F, 6G, and 6H). Down-regulation of TRPV5 in membrane derived kidney protein from Fgfr1 DT-cKO mice was also confirmed using a different anti-TRPV5 antibody ( S2 Fig). TRPV5, a marker for the distal tubule, was detected only in the distal tubule of control, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mouse kidneys by immunohistochemical staining (Fig 7). The distal tubules that express TRPV5 also had morphological features distinct from the proximal tubules exhibited brush borders and lacked expression of TRPV5. α-KL immunoreactivity was detected in the distal tubules of control, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice (Fig 7). α-KL immunoreactivity co-localized with TRPV5 in the distal tubule. We observed no appreciable staining for TRPV5 or Klotho in morphologically distinct proximal tubules characterized by brush border membranes. High power images demonstrated localization of TRPV5 and Klotho to the apical membranes of the distal tubules (Fig 7, lower panels with 600X magnification). Both TRPV5 and α-KL were decreased in Fgfr1 DT-cKO . Expression of TRPV5 and α-KL in Fgfr1 PT-cKO was similar to wild-type mice.

Expression of Npt2a protein
We further examined the expression of sodium-depended phosphate co-transporter gene Npt2a in the kidney of control, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice. Western blot analysis of membrane protein showed that expression of Npt2a was significantly downregulated in Fgfr1 DT-cKO mice (Fig 8A and 8B, and S3 Fig). In contrast, expression of Npt2a protein was significantly upregulated in Fgfr1 PT-cKO mice (Fig 8A and 8B, and S3 Fig). Immunohistochemical staining of kidney sections showed that Npt2a was located at apical brush border membrane of the proximal tubule (Fig 8C). We showed that staining of Npt2a was increased in Fgfr1 PT-cKO mice, and decreased in Fgfr1 DT-cKO mice, respectively (Fig 8C). Thus, the increase in Npt2a protein expression was greater than the increase in Npt2a message expression. Because of the lack of a validated anti-Npt2c antibody, we were unable to assess the effects on Npt2c protein expression.
Effects of rFGF-23 in Fgfr1 PT-cKO and Fgfr1 DT-cKO mice Finally, we examined the effect of selective loss of FGFR1 in the distal tubule on the renal response to FGF-23 administration. Administration rFGF-23 at 100 ng/g IP to 16 week-old controls resulted in a 2-fold increase in circulating FGF-23 concentrations 12 hours after injection. rFGF-23 resulted in a significant 32% reduction in serum phosphate concentrations, a 4-fold increase in urinary phosphate excretion, and a 7-fold reduction in Npt2a message expression in control mice. rFGF-23 administration had no significant effects of serum calcium, PTH or 1,25(OH) 2 D levels or urinary calcium excretion in control mice. rFGF-23 also suppressed Cyp27b1, α-Kl (4-fold), and Fgfr1 (2-fold) message expression in control mice. rFGF-23 treatment resulted in non-significant increments in TRPV5, TRPV6 and CaBP28k.
We observed marked differences in proximal tubular responses to rFGF-23 in Fgfr1 PT-cKO and Fgfr1 DT-cKO mice. Administration of rFGF-23 resulted in a 2.5 fold increase in serum FGF-23 concentrations 12 hours after dosing in Fgfr1 PT-cKO mice. rFGF-23 did not reduce serum phosphate in Fgfr1 PT-cKO mice, but resulted in a~20% increase urinary phosphate excretion, likely due to residual FGFR function. The magnitude of the phosphaturic effect of FGF-23 in Fgfr1 PT-cKO mice, however, was much lower than observed in controls. In addition, rFGF-23 resulted in an~50% reduction in Npt2a and Npt2c expression, but again, the magnitude of this reduction was less that observed in controls treated with rFGF-23.
With regards to distal tubular functions, in Fgfr1 PT-cKO mice, we found that rFGF-23 increased urinary calcium 4-fold in association with significant reductions in TRPV5, TRPV6 and α-Kl expression. In Fgfr1 DT-cKO mice, rFGF-23 had a slight, but non-significant effect to further increase in urinary calcium excretion. Interestingly, however, high urinary calcium persisted, in spite of the effect of FGF-23 to suppress TRPV5, TRPV6, and CaBP28k transcripts in Fgfr1 DT-cKO mice, suggesting an uncoupling between transport and gene expression.

Discussion
Our studies are the first to assess the separate functions of FGFR1 in the proximal and distal tubules of the kidney in mice. By selectively deleting FGFR1 in the proximal tubule using γGT-Cre, which expresses Cre recombinase in kidney proximal tubules after nephrogenesis is complete [27], we were able to assess the postnatal and proximal tubule specific functions of this receptor. Reductions of FGFR1 expression in the proximal tubule resulted in hyperphosphatemia and resistance to the phosphaturic effects of administered recombinant FGF-23. These findings are consistent with the known actions of FGF-23 to regulate Npt2-mediated phosphate transport, but also suggest that FGFR1 has a greater effect to regulate Npt2c message expression. Interestingly, selective loss of FGFR1 in the proximal tubule in Fgfr1 PT-cKO mice did not alter serum 1,25(OH) 2 D levels or Cyp27b1 expression at baseline or in response to recombinant FGF23 administration, but did reduce Cyp24a1 expression.
Since FGF-23 activation of FGFRs reduces 1,25(OH) 2 D production in the proximal tubule through suppression of Cyp27b1 and stimulation of Cyp24a1, our findings suggest possible involvement of other FGFRs in mediating renal responses. Indeed, FGFR3 and FGFR4 are expressed in the proximal tubule and have been implicated in mediating FGF-23 effects. In prior studies, we found that compound mutant Fgfr3 -/-/Fgfr4 -/mice completely reversed the effects of FGF-23 to suppress 1,25(OH) 2 D, but only partially corrected the renal phosphate wasting [20]. These results, along with current findings in Fgfr1 PT-cKO mice, suggest that FGFR1, FGFR3 and FGFR4 have overlapping roles to regulate phosphate transport and distinct functions in the proximal tubule to regulate vitamin D metabolism [18,36]. Although our studies did not examine the down-stream signaling pathways, FGF-23 activation of FGFRs regulates phosphate transport in the proximal tubule through stimulation of extracellular signalregulated kinase 1 and 2 (ERK1/2) and serum and glucocorticoid-regulated kinase 1 (SGK1) phosphorylation of the Na+/H+ exchange regulatory cofactor 1 (NHERF-1) [23]. How FGFR1, FGFR3 and FGFR4 in the proximal tubule, which have common down-stream signaling pathways, concordantly regulate phosphate transport and differentially regulate vitamin D metabolism remain to be determined.
To our knowledge, no other studies have directly assessed the functions of FGFR1 in the distal tubule. Using Ksp-Cre to conditionally delete FGFR1 from the distal tubule, we unexpectedly observed dramatic alterations of both calcium and phosphate homeostasis in mice. Indeed, conditional deletion of FGFR1 in the distal tubule resulted in a marked increase in urinary excretion of calcium that was severe enough to cause secondary hyperparathyroidism. The increase in urinary calcium was accompanied by an increase in urinary phosphate excretion and suppression of Npt2a and Npt2c expression in the proximal tubule in association with significant increments in PTH levels, but no alterations in 1,25(OH) 2 D or changes in circulating FGF-23 levels. In addition, we found evidence for microcrystal formation in the tubular lumen, likely due to the increased calcium and phosphate excretion by the kidney. Our findings that FGFR1 regulates distal tubular calcium transport are consistent with recent reports showing that FGF-23 regulates calcium reabsorption in the distal nephron through stimulation of ERK/SKG1/WNK4-depedendent trafficking of TRPV5 to the plasma membrane [12]. Since hypercalciuria and secondary hyperparathyroidism were not observed in compound mutant Fgfr3 -/-/Fgfr4 -/mice [20], our data suggest that FGFR1 is the principal FGFR regulating calcium transport in the distal tubule. Consistent with this notion, we found that loss of FGFR1 in the distal tubule diminished membrane expression of TRPV5. Indeed, calcium wasting occurred in Fgfr1 DT-cKO mice in spite of compensatory increases in PTH and increments in expression of genes that would be predicted to increase the transcellular calcium transport in the distal tubule, including calbindin D28K (CaBP28k), the sodium-calcium exchanger (NaCX), PMCA1b, calcium channel transient receptor potential vannilloid-5 and -6 (TRPV5 and TRPV6). FGFR1 effects to diminish membrane expression of TRPV5 in Fgfr1 DT-cKO mice likely accounts for both the inability of PTH and intracellular calcium transport pathways to compensate for the hypercalciuria. Interestingly, we also found that loss of FGFR1 had greater effects of TRPV5 message than protein expression. Similarly, FGFR1 also had greater effects on Npt2a protein than message expression in the proximal tubule. This may be due to known effects of FGFR1 to recruit RNA to polysomes to increase in protein expression in other systems [37]. Nevertheless, a role for FGFR1 in the regulation of renal calcium handling and the activation of this receptor by FGF-23 creates a possible regulatory endocrine loop, since circulating calcium levels are an important regulator of FGF-23 expression and secretion from bone [11,38].
α-KL is an obligate co-receptor for FGF-23 activation of FGFR1, FGFR3 and FGFR4 in renal tubules [13]. Our findings that FGFR1 mediated FGF-23 functions in the distal tubule are consistent with the observation that α-KL is predominantly expressed in the distal tubule, and in vivo pulse chase experiments showing FGF-23 activation of FGFR signaling occurs only in the distal tubule [24,25]. The role of α-KL in mediating the effects of FGFR1 in the proximal tubule is less certain. Loss of FGFR1 in the proximal tubule blocked FGF-23 phosphaturic effects, in spite of our inability to detect α-KL in the proximal tubule by immunohistochemistry. However, α-Kl message is present isolated proximal tubule segments [9], suggesting that functional FGFR/ α-KL complexes are present in the proximal tubule. This explanation, however, does not account for the hyperphosphatemia and compensatory increments in serum FGF-23 concentrations in mice with conditional deletion of α-Kl in the distal tubule, which appears to support the presence of a distal to proximal tubule feedback mechanism [25]. Indeed, we would have predicted that the conditional deletion of Fgfr1 and α-Kl from the distal tubule using the same Ksp-Cre would have result in similar phenotypes. Rather, selective deletion of Fgfr1 in the distal tubule inhibited renal calcium transport with secondary effects to inhibit proximal tubular phosphate transport, whereas conditional deletion of α-Kl in the distal tubule resulted in hyperphosphatemia and no reported effect on renal calcium handling, a phenotype resembling Fgfr1 PT-cKO mice. Additional studies will be needed to reconcile these discrepancies, but endocytoses by the proximal tubule of α-KL released into the circulation by ectodomain shedding from the distal tubule has been proposed to sequester Npt2a transporters, leading to FGF-23 independent phosphaturia [39]. It is possible, but not tested, that soluble KL uptake in the proximal tubule may also function as a co-factor for FGFR1 activation [39][40][41][42].
Finally, the mechanism underlying the observed curly tail and spina bifida-like malformation in Fgfr PT-cKO mice are not clear. A similar phenotype has been observed by loss of Fgfr1 in the neural tissues in the posterior portion of the spinal cord in Fgfr1 -/mice [43]. Although we did not test for FGFR1 deletion the specifically in the developing neural tube of Fgfr1 PT-cKO , the tail distortion and spina bifida in the setting of normal serum folate and normal trabecular and cortical bone development, suggests that the γGT-Cre may be expressed in critical sites controlling spinal development. Importantly, however, the development of other bones were normal, and γGT-Cre was not expressed in long bones. We also observed no reductions in FGF-23, which would be expected if we had deleted FGFR1 in FGF-23 producing cells in bone [32]. Fgfr1 DT-cKO , in contrast, had normal spinal and tail development, but demonstrated diminished cortical thickness, similar to TRPV5 -/mice with calcium wasting and elevated PTH levels [44]. Interestingly, elevations in PTH observed in Fgfr1 DT-cKO mice did not result in secondary increments in serum FGF-23 levels. PTH effects on FGF-23, however are variable and context dependent [45]. Although we don't have a specific explanation for why elevations in PTH did not stimulate FGF-23 in Fgfr1 DT-cKO , recent studies implicate a role for α-Klotho as a factor that regulates FGF-23 in osteoblasts [46]; consequently it is tempting to speculate that the reductions in α-Klotho observed in Fgfr1 DT-cKO may have contributed to the refractoriness to PTH stimulation of FGF-23.
In conclusion, the current observations suggest the following physiological schema. FGF-23 directly inhibits phosphate reabsorption in the proximal tubule through activation of FGFR1 (and to a lesser extent FGFR3 and FGFR4), and suppresses 1,25(OH) 2 D production mainly through activation of FGFR3 and FGFR4 [20]; in turn, reduced 1,25(OH) 2 D production by the kidney limits gastrointestinal calcium absorption and leads to secondary increments in PTH. Increased FGF-23 and PTH work together to inhibit proximal tubular phosphate reabsorption through FGFRs and PTH receptors in the proximal tubule. Similarly, both FGF-23 and PTH have concurrent effects to enhance renal calcium reabsorption in the distal tubule through their respective receptors. The consequent increase in the renal reabsorption of calcium offsets the hypocalcemic effects of reduced 1,25(OH) 2 D, and in doing so, allows PTH and FGF-23 to increase the renal excretion of phosphate, while maintaining calcium homeostasis. In this schema, increments in serum calcium or excess 1,25(OH) 2 D suppress PTH production by the parathyroid gland, but stimulate FGF-23 production by bone [4], creating a scenario where elevations of FGF-23 can offset the loss of PTH on urinary phosphate and calcium excretion by directly inhibiting proximal tubule phosphate, stimulating distal tubular calcium reabsorption, and reducing the renal production of 1,25(OH) 2 D.
Supporting Information S1 Fig. Immunohistochemistry of FGFR1 expression in the kidney of WT, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice. FGFR1 staining was found in both proximal and distal tubules in kidney sections derived from wild-type mice. Fgfr1 DT-cKO mice showed loss of FGFR1 staining in the distal tubules and persistent FGFR1 staining in the proximal tubule. The FGFR1 expression pattern was reciprocal in Fgfr1 PT-cKO mice, with loss of FGFR1 in the proximal tubule and continued expression in the distal tubular segments. The proximal tubules are identified by FGFR1 expression in luminal brush border membranes. The different structures are indicted as follows: proximal tubules (PT), distal tubules (DT), glomeruli (G). (PDF) S2 Fig. Comparison of TRVP5 antibodies. A and B. Western blot analysis of expression of TRPV5 from membrane proteins prepared from kidney cortex of 16-weeks-old control, Fgfr1 PT-cKO , or Fgfr1 DT-cKO mice. TRPV5 expression in membrane protein was detected by both the anti-TRPV5 antibody from Alomone (A) and from Santa Cruz (B). C and D. Immunohistochemical staining of TRPV5 expression in kidney sections of 16-weeks-old control, Fgfr1 PT-cKO , or Fgfr1 DT-cKO mice. TRPV5 expression was decreased in distal tubule of Fgfr1 DT-cKO mice compared to WT control mice using either Alomone (C) or Santa Cruz TRPV5 (D) antibodies. (PDF) S3 Fig. Npt2a protein expression in membrane-derived proteins from the kidney cortex of 16-weeks-old WT, Fgfr1 PT-cKO , and Fgfr1 DT-cKO mice. Western blot analysis was performed to assess expression of Npt2a using the Abcam anti-Npt2a antibody. To validate the antibody, protein samples isolated from Npt2a overexpressing MDCK cells and from pcDNA3.1 vectortransfected MDCK cells that lack Npt2a expression [35] were used as positive and negative controls, respectively. Expression of Npt2a was decreased in Fgfr1 DT-cKO mice and increased in Fgfr1 PT-cKO mice. (PDF)