Elevated serum 1,25(OH)2-vitamin D3 level attenuates renal tubulointerstitial fibrosis induced by unilateral ureteral obstruction in kl/kl mice

Previous studies have suggested that Klotho provides reno-protection against unilateral ureteral obstruction (UUO)-induced renal tubulointerstitial fibrosis (RTF). Because the existing studies are mainly performed using heterozygous Klotho mutant (HT) mice, we focused on the effect of UUO on homozygous Klotho mutant (kl/kl) mice. UUO kidneys from HT mice showed a significantly higher level of RTF and TGF-β/Smad3 signaling than wild-type (WT) mice, whereas both were greatly suppressed in kl/kl mice. Primary proximal tubular epithelial culture cells isolated from kl/kl mice showed no suppression in TGF-β1-induced epithelial mesenchymal transition (EMT) compared to those from HT mice. In the renal epithelial cell line NRK52E, a large amount of inorganic phosphate (Pi), FGF23, or calcitriol was added to the medium to mimic the in vivo homeostasis of kl/kl mice. Neither Pi nor FGF23 antagonized TGF-β1-induced EMT. In contrast, calcitriol ameliorated TGF-β1-induced EMT in a dose dependent manner. A vitamin D3-deficient diet normalized the serum 1,25 (OH)2 vitamin D3 level in kl/kl mice and enhanced UUO-induced RTF and TGF-β/Smad3 signaling. In conclusion, the alleviation of UUO-induced RTF in kl/kl mice was due to the TGF-β1 signaling suppression caused by an elevated serum 1, 25(OH)2 vitamin D3.

vitamin D and its analogues have been shown to protect the kidneys from fibrosis due to various kidney injuries 21,22 . Also, 1, 25(OH) 2 vitamin D 3 reduces gene expression related to TGF-b-induced fibrosis in human uterine leiomyoma cells 23 . An active form of the vitamin D 3 analogue maxacalcitol recruits a PPM1A/VDR complex to phosphorylated Smad3 to accelerate its dephosphorylation 24 . Thus, the outcome of UUO-induced RTF becomes uncertain because of the opposing effect of an elevated vitamin D and Klotho deficiency. Because all of the existing studies on the effect of Klotho on UUOinduced RTF were performed using HT mice, whether the disrupted homeostasis in kl/kl mice has an effect is not known.
In this study, we compared the degree of UUO-induced RTF among wild-type, HT, and kl/kl mice and analyzed the underlying mechanisms.

Results
UUO-induced RTF is the mildest in kl/kl mice. Six week-old wildtype (WT), heterozygous Klotho mutant (HT), and kl/kl mice were subjected to UUO for 3 days or 1 week. The expression of RTF markers collagen I and aSMA was assessed by immunohistochemistry and quantitative real-time PCR. RTF developed after the UUO operation in a time dependent manner in WT and HT mice compared to the sham-operated control groups. The expression of RTF markers was much higher in HT mice than WT mice, which is corroborated by previous reports 18,20 . In the sham-operated groups, the expression of RTF markers was comparable between the WT and HT mice, but was increased more than 2-fold in the kl/kl mice. Conversely, kl/kl mice showed only a slight increase in the expression of fibrosis markers after UUO (Fig. 1).
TGF-b signaling is suppressed in kl/kl mice. Previous studies showed that Klotho protein interferes with TGF-b1 binding to its receptor, resulting in suppressed TGF-b signaling 18 . As a result, UUO-induced RTF is exaggerated in HT mice 20 . To examine whether the suppressed expression of RTF markers in kl/kl mice is a result of inhibited TGF-b signaling, the phosphorylation of Smad3 and TGF-b1 expression was assayed. In parallel with RTF markers, after UUO, phosphorylated Smad3 (pSmad3) and TGF-b1 expression were the highest in HT mice and were significantly suppressed (e, f ) Real-time PCR for Col1a1 (e) and aSMA (f ). Note that UUO-induced collagen I deposition and fibroblast activation were the most severe in HT mice at the protein and mRNA levels and were largely abolished in kl/kl mice. Data are presented as the means 6 S.D. *P , 0.05 (n 5 5). in kl/kl mice. Interestingly, in the sham-operated groups, the kl/kl mice had a higher percentage of pSmad3 positive staining cells than the other two groups, whereas this group had the lowest TGF-b1 content, suggesting that Smad3 phosphorylation in the shamoperated kl/kl mice was not a result of increased TGF-b1 (Fig. 2).
No suppression of the TGF-b1-induced epithelial-mesenchymal transition is observed in primary cultured proximal tubular epithelial culture cells isolated from kl/kl mice. Because kl/kl mice have disrupted phosphate and calcium (Pi/Ca) homeostasis, determining whether the suppressed TGF-b/Smad3 signaling is a result of complete depletion of Klotho or disrupted Pi/Ca homeostasis, most noticeably high phosphate, high FGF23, and high 1, 25(OH) 2 vitamin D 3 , is difficult. To address the underlying molecular mechanism, we compared the expression of molecules involved in epithelial-mesenchymal transition (EMT) using primary cultured proximal tubular epithelial cells (PTECs) from WT, HT, and kl/kl mice. TGF-b1-induced EMT was enhanced in HT and kl/kl cells compared to WT cells. The EMT markers showed higher aSMA and lower E-cadherin mRNA and protein expression at the indicated time point (Fig. 3). Smad3 phosphorylation was also enhanced in HT and kl/kl cells. Unlike UUO kidneys, kl/kl cells showed no suppression of TGF-b1-induced EMT and Smad3 phosphorylation compared to HT cells. These results suggest that the suppressed TGF-b/Smad3 signaling observed in UUO-induced RTF in kl/kl kidneys was a result of disrupted Pi/Ca homeostasis and not complete Klotho depletion (Fig. 3).
TGF-b1-induced EMT is suppressed not by high phosphate or FGF23 but by 1,25(OH) 2 vitamin D 3 in a dose-dependent manner. Previous studies showed that in kl/kl mice, serum phosphate and 1,25(OH) 2 vitamin D 3 increase sharply 1 and serum FGF23 concentration increased 2,000-fold compared to WT mice 4 . To identify an individual factor(s) that contribute to suppressed UUO-induced RTF expression in kl/kl mice, we used the normal kidney epithelial cell line NRK52E, and these cells were cultured in the presence of high phosphate (2 or 4 mM), FGF23 (0.2 or 2 mg/ml), or 1,25(OH) 2 vitamin D 3 (0.1 or 1 mg/ml) to evaluate their effect on TGF-b1-induced EMT. In the cells treated with TGF-b1 together with phosphate or FGF23 for 12 or 24 h, TGF-b1-induced EMT was not affected at the mRNA or protein level compared to cells treated with TGF-b1 alone. In contrast, 1,25(OH) 2 vitamin D 3 ameliorated TGF-b1-induced Smad3 phosphorylation, aSMA expression, and E-cadherin suppression in a dose dependent manner at the mRNA and protein levels. This suppressive effect of 1,25(OH) 2 vitamin D 3 on TGF-b-induced EMT lasted for 72 h after TGF-b treatment (supplementary Fig. 1).
Interestingly, the cells cultured with high phosphate (4 mM) for 48 h had more Smad3 phosphorylation and higher aSMA expression compared to cells cultured in normal DMEM, although TGF-b1 was absent. These results may explain the increased fibrosis markers observed in sham-operated kl/kl mice (Figs. 4 and 5).
Vitamin D-free diet normalizes serum 1,25(OH) 2 vitamin D 3 level and enhances the expression of UUO-induced RTF markers in kl/kl mice. A vitamin D-free diet (D-diet) has been reported to maintain phosphate homeostasis and improve the kl/kl phenotype 8,9 . In this study, we fed kl/kl mice on a D-diet to determine whether the decreased vitamin D level could enhance UUO-induced RTF. In parallel with previous studies, the D-diet normalized serum 1,25(OH) 2 vitamin D 3 and increased the body weight of the kl/kl mice, such that they were indistinguishable from their WT littermates (Fig. 6).
After the UUO operation, based on immunohistochemistry, expression of collagen I and aSMA was enhanced by the D-diet, with 3-and 4-fold increases, respectively, compared to kl/kl mice fed a standard diet, and mRNA expression showed a similar trend (Fig. 7).
In addition, the D-diet enhanced TGF-b/Smad3 signaling in UUO-induced RTF in kl/kl mice. The percentage of pSmad3-positive cells increased about 2.5-fold as a result of vitamin D withdrawal. ELISA was performed to measure the TGF-b1 content in kidney lysates. Protein and mRNA expression of UUO-induced TGF-b1 increased about 2-fold using the D-diet. Interestingly, in the shamoperated kl/kl mice fed the standard diet, TGF-b1 protein expression was lower than the kl/kl mice fed the D-diet, while there were more pSmad3 positive cells in kl/kl mice fed the standard diet (Fig. 8).
Note that in sham-operated kl/kl mice fed a standard diet, the expression of RTF markers and activation of TGF-b/Smad3 signaling are slightly but significantly higher than kl/kl mice fed a D-diet. The results suggest that increased RTF marker expression in sham-operated kl/kl mice was not a result of Klotho depletion but the disrupted Pi/Ca homeostasis (Figs. 7 and 8).

Discussion
In this study, we demonstrated that UUO-induced RTF was greatly suppressed in kl/kl mice compared to HT mice. Conversely, slightly but significantly more severe renal fibrosis was observed in shamoperated kl/kl kidneys than WT and HT kidneys. Because kl/kl mice exhibit a variety of abnormalities, among which phosphate retention and high serum 1, 25(OH) 2 vitamin D 3 are the most prominent changes, we hypothesized that disrupted Pi/Ca homeostasis in kl/kl mice may play an important role in the mild renal fibrosis in shamoperated mice and the suppression of RTF in UUO-operated mice.
We observed that HT mice exhibit more severe UUO-induced RTF compared to the WT group, which is consistent with previous reports 20 . Klotho has been proven to protect the kidneys from injuryinduced fibrosis by counteracting TGF-b/Smad3 and Wnt signal- ing 18,19 . Klotho deficiency renders the kidneys more susceptible to acute insults, while exogenous Klotho expression attenuates renal fibrosis due to various causes 13,14,18,20,25 . In rodent models, renal damage caused by several different kinds of stress, such as acute inflammatory stress and sustained circulatory stress, could suppress Klotho expression 26,27 . In the human kidneys, decreased Klotho expression occurs as early as CKD stage 2 28 . However, unexpectedly, UUO-induced RTF in kl/kl mice was greatly attenuated. Because TGF-b/Smad3 signaling is believed to be the main driving force toward post-UUO RTF 29 ,we examined TGF-b/Smad3 signaling, which turned out to be significantly suppressed in kl/kl kidneys. To identify the cause of this suppression, we examined TGF-b-induced EMT using primary cultured PTECs to exclude the humoral effects. Interestingly, unlike in vivo experiments, TGF-b1-induced EMT was not suppressed in cultured kl/kl PTECs, and there was an increase in Smad3 phosphorylation. These results suggest that the suppression of UUO-induced RTF in kl/kl mice might result from a disturbance in Pi/Ca homeostasis due to deficiency of Klotho. We have not investigated ERK/MAPK activation in cultured kl/kl PTECs because the increase in Smad3 phosphorylation was significant, but it might be important to examine non-canonical TGF-b signaling, which also plays an important role in EMT.
Next, because phosphate retention, high serum FGF23, and high serum 1, 25(OH) 2 vitamin D 3 are the most prominent changes in kl/kl mice 30 , we examined these factors individually using in vitro cultures. NRK52E cells were challenged with TGF-b1 in the presence of high phosphate, FGF23, or calcitriol (1, 25(OH) 2 vitamin D 3 ). TGF-b1-induced EMT was not affected by the presence of high phosphate or FGF23, but was inhibited by calcitriol in a dose dependent manner. These results suggest that a high level of 1, 25(OH) 2 vitamin D 3 but not phosphate or FGF23 may be the cause of the suppressed UUO-induced RTF in kl/kl mice.
To confirm this mechanism, we sought to eliminate the effect of vitamin D in kl/kl mice by feeding a D-diet and determining the degree of UUO-induced RTF. The D-diet normalized 1, 25(OH) 2 vitamin D 3 and phosphate levels in the blood of kl/kl mice and completely rescued the kl/kl phenotype, which is in agreement with previous reports 8 . As we predicted, the D-diet enhanced UUO- induced RTF in kl/kl mice to the same levels as HT mice. In addition, TGF-b signaling was enhanced by eliminating vitamin D. Paricalcitol, a synthetic vitamin D analogue, has been reported to attenuate renal interstitial fibrosis in obstructive nephropathy via repressing expression of TGF-b1 and its type I receptor 31 . Moreover, vitamin D analogs are reported to ameliorate proteinuria and kidney injury by blocking Wnt/b-catenin signaling 32 . We therefore conclude that an elevated level of 1, 25(OH) 2 vitamin D 3 in kl/kl   In sham-operated mice, there was no significant difference in the expression of RTF markers between WT and HT kidneys, whereas expression was significantly albeit mildly increased in kl/kl kidneys compared to WT and HT kidneys. Sugiura, H et al 20 suggested that this result may be due to failed TGF-b/Smad3 inhibition related to the absence of Klotho. However, our results did not support this hypothesis because the kl/kl PTECs expressed the same levels of EMT markers as WT and HT PTECs treated with TGF-b1. In addition, kl/kl mice fed a D-diet did not have any RTF onset in the shamoperated group. Interestingly, NRK52E cells cultured for 48 h in the medium containing a high concentration of phosphate (4 mM) showed Smad3 phosphorylation and increased aSMA expression compared to the control medium (Pi 1 mM), suggesting that an increased expression of RTF markers in sham-operated kl/kl mice may be a result of a disrupted Pi/Ca homeostasis, that is, a high concentration of phosphate and not a result of failed TGF-b/ Smad3 inhibition by Klotho. However, at this time, we cannot explain how a high concentration of phosphate phosphorylates Smad3.
In conclusion, we are the first to report that UUO-induced TGF-b/ Smad3 signaling is suppressed in kl/kl mice by an increased level of 1, 25(OH) 2 vitamin D 3 , resulting in amelioration of UUO-induced RTF. We identified 1, 25(OH) 2 vitamin D 3 to be a powerful suppressor of TGF-b/Smad3 signaling and protect the kidneys from RTF.

Methods
Animals, diets, UUO animal models, and tissue preparation. All experimental protocols were approved by the Animal Studies Committee of Wakayama Medical University. The methods were carried out in accordance with the approved guideline.
This study used klotho/Jcl mice, which were purchased from CLEA Japan and described elsewhere 1 . Mice were raised in a standard facility with free access to food and water.
Animals were fed a standard diet or vitamin D 3 free diet (D-diet). In the D-diet groups, heterozygous klotho female mice were maintained on a vitamin D 3 free diet (PMI nutrition international, Inc.) after pregnancy, in which the composition was identical to the standard diet except for missing vitamin D 3 . The offspring of D-diet heterozygous klotho females were also fed a D-diet throughout the experiments.
Six-to eight-week-old male wild-type (WT), heterozygous (HT), and kl/kl mice from the standard diet and D-diet groups were used to perform UUO. UUO surgery was performed as described elsewhere 33,34 . In short, male mice from each group (n 5 5) were anesthetized by an intraperitoneal injection of ketamine (100 mg/kg body weight) and xylazine (5 mg/kg body weight). A left-flank incision was made to expose the left ureter, and then ligation was done with a 5-0 silk suture. The sham groups (n 5 3) received the same procedure except for ureteral ligation.
At the time of sacrifice, the ligated kidneys (or sham kidneys) were removed and cut transversely. Tissues were fixed with 4% paraformaldehyde for histopathological examination or lysed by denaturing lysis buffer (50 mM Tris-HCl, pH 7.4, 1% triton X, 150 mM NaCl, 1 mM EDTA) with addition of a phosphatase inhibitor (2.5 mM sodium orthovanadate, 10 mM NaF, and 10 mM b-glycerol phosphate) and proteinase inhibitor (Roche). Lysed samples were stored in 280uC until use for western blot analysis or ELISA.
Immunohistochemistry and immunofluorescent staining. Immunohistochemistry was performed on paraffin sections using a microwave-based antigen retrieval technique. The antibodies used in this study include collagen I (Southern Biotechnology, Birmingham, AL), pSmad3 (Cell Signaling, Danvers, MA), aSMA (Sigma) and Fsp1 (Abcam). After incubation with the primary antibodies, sections were treated with a Vectastain ABC kit (Vector Laboratories, Birlingame, CA) according to the manufacturer's instructions. For immunofluorescent staining, a Cy3-anti-rabbit IgG (Sigma Aldrich) was used as a secondary antibody. TGF-b1 ELISA assay. Kidney cortices were dissected and homogenized in denaturing lysis buffer containing 100 mM NaCl, 1% Triton X-100, and 0.05 M Tris-HCl, pH7.4. Protein concentrations were measured using a BCA method according to the manufacturer's instruction. Total TGF-b1 was quantified using a TGF-b1 ELISA kit (Invitrogen). Values were expressed as pg/mg protein for the protein extracts.
Measurement of active serum 1, 25(OH) 2 vitamin D 3 in mice. Six-week-old male mice were anesthetized with ketamine and xylazine; blood was collected from the inferior vena cava, and the serum was separated by centrifugation at 7,500 g for 10 min at 4uC (n 5 4). The serum level of 1, 25(OH) 2 vitamin D 3 was measured by RIA (SRL Inc., Tokyo, Japan).
Cell culture. Primary proximal tubular epithelial cells were generated from the kidneys of WT, HT, and kl/kl mice using a method described previously 33   modification. Briefly, kidney cortices (6 wk male mice) were dissected, sliced, minced, and digested in 300 U/ml type II collagenase (Worthington Bio Corp) in FBS free DMEM in a shaking incubator at 37uC for 30 min. Collagenase was neutralized with 10% FBS/DMEM containing 100 U/ml penicillin and 0.1 mg/ml streptomycin. The suspension was triturated by pipetting and was passed through a 70 mm cell strainer (Becton Dickinson Labware, Franklin Lakes, NJ). The samples were centrifuged (500 rpm, 5 minutes) to pellet the tubules, washed with 10 ml FBS free-DMEM, centrifuged, and washed twice more. The final pellet, consisting mostly of renal tubules, was resuspended in culture medium (REBM bullet kit, Clonetics) and incubated at 37uC in a 5% CO 2 incubator with medium changes every 2 days until approximately 80% confluent.
NRK52E cells were obtained from ATCC and cultured in DMEM with 5% FBS (Gibco) without antibiotic supplementation.
For TGF-b1 stimulation, cells underwent 24 h serum starvation and then were exposed to 5 ng/ml TGF-b1 (Cell signaling) for 12 or 24 h in the presence or absence of 0.2 or 2 mg/ml murine FGF23 (Peprotech) or 0.1 or1 mg/ml Calcitriol (Cayman).
To make high phosphate growth medium, an equal volume of 100 mM NaH 2 PO 4 and 100 mM Na 2 HPO 4 were mixed to make 100 mM Pi stocks; 2 mM and 4 mM Pi medium were made by dilution of Pi stocks with DMEM (Pi concentration in DMEM is approximately 1 mM). NRK52E cells were grown in normal DMEM with 5% FBS or high Pi DMEM with 5% FBS until 80% confluent. TGF-b1 stimulation was performed with serum free normal DMEM or high Pi DMEM.
Quantitative Analysis. To quantify activated Smad3, the positive stained nuclei were counted in 20 consecutive high power fields (4003) and divided by the total number of nuclei. The number of collagen, and aSMA positive cells in the cortical interstitium was performed as described 36 . Band intensity in the western blots was measured using Image J software.
Statistical Analysis. The data were analyzed with Student's t-test and one-way ANOVA with a Student-Newman-Keuls test (SPSS, 13.0) and expressed as means 6 SD. For the data without Gaussian distribution, Kruskal-Wallis with post-hoc Mann-Whitney was used. P , 0.05 was considered to be statistically significant.