Klotho Supplementation Reverses Renal Dysfunction and Interstitial Fibrosis in Remnant Kidney

Klotho is a member of type 1 membrane proteins and mainly expressed in the kidney, especially in distal nephrons but not resident fibroblasts [1, 2]. Extracellular domain of klotho is enzymatically cleaved and released in the renal interstitium and then enters systemic circulation [3]. Klotho interacts with various membrane proteins, including receptors and channels, to modulate their functions. Klotho binds to Wnt, insulin-like growth factor (IGF) receptor [4, 5], exerting anti-aging and anti-oxidative actions. Klotho also binds to transforming growth factor β (TGFβ) receptor, inhibiting fibrosis [6]. Klotho converts canonical fibroblast growth factor (FGF) receptor into a specific receptor for FGF23 [7], regulating phosphate excretion.

Among one-fifth of patients with end-stage renal diseases, underlying chronic kidney diseases (CKDs) are unclear [8]. Final common pathway to end-stage renal disease is glomerular sclerosis with interstitial fibrosis, in which myofibroblasts produce massive matrix deposits and express all components of the renin-angiotensin system (RAS) [9, 10]. A large population of human kidney myofibroblasts is derived from pericytes and fibroblasts [11], and myofibroblast apoptosis triggers the reversal of fibrosis [12]. Although both pericytes and resident fibroblasts have high plasticity, the local microenvironment appears to determine their fate [13].

Remnant rat kidney is a typical model of common human progressive CKD, with glomerular sclerosis, interstitial fibrosis, and diminished expression of klotho [14, 15]. In remnant kidney, renal function is markedly decreased immediately after nephrectomy and stabilized with partial recovery 2–4 weeks after nephrectomy, associated with compensatory hypertrophy. The inhibition of RAS shows renoprotection in this model [16]. Molecular mechanisms underlying interstitial fibrosis are extensively investigated. Among them, TGFβ and its target gene and downstream mediator, connective tissue growth factor (CTGF) are central for profibrotic reactions in this model [17]. TGFβ also prevents myofibroblast apoptosis [12]. In addition, Wnt expression is enhanced in long-term remnant kidney, contributing to fibrosis [18]. Although IGF plays an important role in compensatory hypertrophy [19], IGF inhibition does not alter interstitial fibrosis in remnant kidney [20]. In mesangial cells, TGFβ and IGF enhance fibrogenesis, contributing to the development of glomerular sclerosis [21]. CTGF activates Wnt signaling through low-density lipoprotein receptor-related protein 6 (LRP6) in pericytes as well as mesangial cells [22, 23].

Bone morphogenetic protein (BMP) 7 is expressed in both metanephric mesenchyme and ureteric bud, maintaining metanephric mesenchyme by preventing early epithelization during nephrogenesis, which involves epigenetic and transcriptional regulation of genes [24, 25]. BMP7 is localized mainly in podocyte, loop of Henle and distal nephrons of adult kidney. Our previous data showed that side population (SP) cells, an enriched population of stem cells, were obtained from adult kidneys and that the inhibition of histone deacetylase (HDAC) elevated BMP7 expression predominantly in SP cells and reversed renal dysfunction in CKD model [26]. Although BMP signals are important to reverse renal interstitial fibrosis [27], it has not been assessed whether klotho protein supplementation is able to enhance BMP7 and reverse renal fibrosis [28]. Furthermore, while klotho retards the progression of kidney diseases [29], it has not been shown whether klotho protein supplementation is able to reverse renal dysfunction. The present study was performed to examine whether klotho supplementation reversed renal dysfunction and fibrosis in remnant kidney.

Eighteen 6-week-old male Wistar rats were purchased from Sankyo Laboratory Service Corporation (Tokyo, Japan). All efforts were made to minimize animal suffering. Each rat was given free access to water and chow containing 1.06% calcium and 0.99% phosphate (CE-2, Nihon CLEA, Tokyo, Japan), housed separately in a metabolic cage that was kept in a temperature-controlled room with a 12/12-h light/dark cycle [2, 30].

After 1 week, ligations of all but one posterior branch of the left renal artery were performed under pentobarbital anesthesia (50 mg/kg ip). After another week, animals were again anesthetized and right nephrectomy was performed with flank incision [16]. After surgery, rats were divided into 3 groups (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000530469). The first group was treated with vehicle alone and sacrificed at 12 weeks of age (SR). Second group was also administered vehicle and harvested at 20 weeks of age (LR). The third group received subcutaneous administration of klotho daily from 12 weeks of age and continued for 8 weeks (KL). The study was designed to examine whether klotho protein supplementation was able to reverse renal dysfunction. Regarding klotho administration, 4 weeks were allowed after nephrectomy to stabilize renal function [14]. Exogenous supplementation of recombinant human klotho protein (rh-klotho, 20 μg/kg/day, PeproTech Inc., Rocky Hill, NJ, USA) or a vehicle (0.1 mL of saline containing 0.1% bovine serum albumin [BSA]) was used. Rh-klotho, which consists of 516 amino acid residues from the KL1 domain, ameliorated calcium and phosphate abnormalities in klotho-deficient animals, indicating that it is physiologically active [3, 31]. Systolic blood pressure (SBP) was measured every 4 weeks using the tail-cuff method, and urine was collected every 4 weeks to measure albumin and prostaglandin F2α (PGF) by enzyme-linked immunosorbent assay.

At the end of experiment, rats were anesthetized with intra-peritoneal Inactin (100 mg/kg, Byk Gulden, Konstanz, Germany). To directly measure mean blood pressure and obtain blood samples, femoral artery was cannulated [19]. Jugular vein was also cannulated (PE50) to infuse saline containing 1% BSA and 7.5% Inutest (Laevosan-Gesellschaft, Linz/Donau Austria). The ureter was cannulated (PE10) with a midline abdominal incision to collect urine in a preweighed tube. Rats were allowed to breathe air enriched with 100% oxygen (aimed at FiO₂ of 40%), which markedly improves the stability of arterial blood pressure. After completion of surgery, 1 h of equilibration was allowed. Two separate 20-min clearance studies were performed to measure glomerular filtration rate (GFR). After completion of clearance study, the renal artery and vein were ligated, and the kidney was removed and weighed. The animals were euthanized by administering an overdose of barbital. The kidney was cut in half. Half of the kidney from each rat was quickly frozen in liquid nitrogen. The other half was fixed in 4% formalin solution. Blood samples were centrifuged at 4°C for 10 min. Serum, urine, and tissues were deep-frozen until usage. Klotho and angiotensin II (ANGII) were measured, using enzyme-linked immunosorbent assay and radioimmunoassay, respectively [32].

For microscopic evaluation, renal tissue specimens were embedded in paraffin and cut into approximately 3 μm-thick sections, which were stained with periodic acid-Schiff or Masson trichrome. The fraction of the renal cortex occupied by interstitial fibrosis was quantified [2]. Twenty-five consecutive microscopic fields, at a final magnification of ×100, were randomly selected for each kidney section, and interstitial fibrosis was evaluated according to the following scale: 0 = no fibrosis, 1 = interstitial fibrosis less than one quarter of a visual field, 2 = one quarter to one half of a visual field, 3 = one half to three quarters of a visual field, and 4 = more than three quarters of a visual field. A fibrosis index was calculated for each rat as the arithmetic mean of these values. As detailed previously [33], glomerulosclerosis score (GS) was assessed by grading glomerulus between 0 and 3, based on the percentage surface area of involvement in each glomerulus: grade 1 represented an involvement of <30% of a sectioned glomerular area, whereas grade 3 indicated that >60% of the sectioned area was affected. One hundred glomeruli were assessed for each animal. Immunohistochemistry was also performed to demonstrate the localization of BMP7 (rabbit anti-BMP7 antibody, Novus Biochemical) and klotho (goat anti-klotho antibody, Novus Biochemical) in 3 groups on the same day. Two independent pathologists examined the tissues in an observer-blinded fashion.

Human embryonic kidney (HEK) 293 cells (Funakoshi, Tokyo, Japan), a cell line that constitutively expressed klotho [34], were grown and maintained in DMEM (Sigma-Aldrich, St Louis, MO, USA) supplemented with 10% fetal bovine serum (Biowest, Nuallie, France), 1% penicillin-streptomycin-glutamine (Nacalai Tesque, Kyoto, Japan) on 100-mm culture plate (Corning, Phenix, AZ, USA). For transfection, HEK 293 cells were plated in 24-well plates in 300 μL of the medium mentioned above [35], and 100 nm of klotho small interfering RNA (Invitrogen, Waltham, MA, USA) or scrambled RNA was mixed with 4 μL of Lipofectamine RNAiMAX (Invitrogen) in 30 μL OPTI-MEM (Gibco, Waltham, MA, USA). After 10 min, the solution was added dropwise to the cells and gently mixed. On the next day after transfection, the medium was replaced with 300 μL of DMEM containing antibiotics. In the first series of experiments, basal expression of BMP7 and klotho was checked by real-time polymerase chain reaction (RT-PCR) in cells grown for 48 h after transfection. The experiments were repeated 4 separate times.

In the second study, HEK 293 cells were divided into 4 groups 1 day before transfection: first group was treated with TGFβ (10 ng/mL, R&D Systems, Minneapolis, MS, USA), second group was treated with SB525334 (SB, 10 μm, R&D Systems), third group received combined treatment with TGFβ and SB, and the fourth group was treated with vehicle [32, 36]. To prepare stock solution, TGFβ was dissolved in DMEM containing 1% BSA (Wako Pure Chemical) and antibiotics. Next day, transfection was performed. One day later, the medium was replaced with 300 μL of DMEM containing antibiotics and respective reagents in 4 groups. The experiments were performed 4 separate times. The expressions of BMP7 and HDAC1 were checked 48 h after transfection.

In the third series of experiments, intervention was given with or without recombinant human FGF23 (R&D Systems) 48 h after transfection in HEK 293 cells. To prepare stock solution, FGF23 was dissolved in DMEM containing 1% BSA (Wako Pure Chemical) and antibiotics. Cells were treated with 10 nm or 100 nm of FGF23 in DMEM with antibiotics [31]. Samplings were performed after 1, 6, and 24 h. The experiments were carried out 4 separate times.

Data are expressed as mean ± SEM, unless otherwise stated. Statistical analysis was performed using analysis of variance (ANOVA) and Student’s t test with or without Bonferroni correction. p values <0.05 were considered statistically significant. Please note that detailed methods for RT-PCR, immunoprecipitation, and Western blot were described in supplemental materials.

Albuminuria, SBP, and PGF excretion were similar among 3 groups at 4 weeks of the experimental period (Fig. 1a–c). Compared to long remnant kidney (LR) group, SBP, albuminuria, and PGF excretion were reduced in klotho-treated (KL) group at 12 weeks of the experimental period. At the time of GFR measurement, mean blood pressure in short remnant kidney (SR) group (94 ± 2 mm Hg) was lower than that of LR group (106 ± 2 mm Hg, p < 0.05) and that of KL group (85 ± 2 mm Hg, p < 0.05) was even lower than that of SR group. While GFR in LR group tended to be smaller than that of SR group (p = 0.08), GFR in KL group was higher than that of SR or LR group (p < 0.05), indicating that klotho supplementation reversed renal dysfunction (Fig. 1d). Although GS in LR group was greater than that in SR and KL group (p < 0.05), GS was similar between KL and SR group (Fig. 2). It may take a longer period for klotho supplementation to reverse GS. Body and kidney weights were similar among 3 groups (online suppl. figures 2 and 3). While renal ANGII concentration and renal expression of angiotensinogen in LR group were elevated compared to those of SR group, they were decreased by klotho supplementation (p < 0.05 for both, Table 1).

Compared to SR and LR group, serum concentration of klotho, renal expression of klotho, and urinary klotho in KL group were elevated (p < 0.05 for all, Table 1). Although serum phosphate and FGF23 levels in KL group were smaller than those of SR and LR group (p < 0.05), fractional excretion of phosphate in KL group was greater than that in SR and LR group (p < 0.05). Regarding serum calcium and parathyroid hormone levels, no significant differences were detected among 3 groups. Preliminary studies indicated that the dose of klotho used in this study was required and sufficient to increase serum klotho levels in remnant kidney (online Suppl. Table 1).

Fibrosis index in LR group was higher than that in SR group, whereas KL group showed a smaller fibrosis index to SR group (p < 0.05 for both, Fig. 2), implicating the reversal of fibrosis by klotho. Consistently, renal expression of TGFβ and CTGF showed similar trends with fibrosis index (Table 2), supporting that albuminuria and ANGII are strong inducers of TGFβ [32]. Renal expression of lymphoid enhancer-binding factor 1 (LEF1), a target gene of Wnt, also exhibited tendency similar to fibrosis index, which was accordant with the notion that klotho binds Wnt to inhibit its signaling [4]. Tissue inhibitor of metalloprotease 1 (TIMP1) expression in LR group was greater than that in SR and KL groups (p < 0.05), suggesting its contribution to the progression of fibrosis in LR group. Klotho protein supplementation diminished renal expression of collagen I and HDAC1 in remnant kidney (p < 0.05 for both). In contrast, klotho supplementation enhanced renal expression of BMP7 and superoxide dismutase (Table 2, SOD, p < 0.05 for both): the latter may account for the reduction of PGF excretion, an index of oxidative stress, in KL group (5). Klotho supplementation elevated BMP7 and klotho expression in remnant kidney, so we attempted to specify the sites of their expression (Fig. 3). Immunohistochemical studies revealed that klotho and BMP7 colocalized in interstitial cells as well as tubular epithelium in KL group.

All in vitro studies were performed to support the in vivo findings and further investigate the potential mechanisms. Klotho binds to transient receptor potential channel (TRPC) 1 [37], and TRPC1 shows the similarity with TRPC6, which plays a pivotal role in glomerular sclerosis. Because TRPC6 elicits calcium entry into podocyte [29], and because klotho supplementation enhanced GFR in vivo, immunoprecipitation studies were performed to examine possible binding of klotho with TRPC6. As shown in online supplemental Figure 4, however, Western blot did not show that klotho exhibited considerable dose-dependent bindings with TRPC6.

Since exogenous klotho protein supplementation increased endogenous renal expression of klotho and BMP7 in vivo, cell studies were performed to examine whether the downregulation of klotho influenced BMP7 expression (Fig. 4a–b). Surprisingly, small interfering RNA against klotho significantly downregulated BMP7 as well as klotho expression in HEK 293 cells (p < 0.01 for both).

Following studies were performed to examine the interactions between klotho and TGFβ (or FGF23) in a direct manner. As both BMP7 and klotho counteract TGFβ [6, 27], and because klotho supplementation enhanced BMP7 and reduced TGFβ expression in vivo, we assessed whether TGFβ influenced BMP7 expression and whether klotho affected this response (Fig. 4c). Three-way ANOVA revealed that TGFβ reduced BMP7 expression (online suppl. Table 2, p < 0.001), and SB525334 (SB) increased it in HEK cells (p < 0.001). The presence of klotho decreased the magnitude of the above responses (p < 0.001). As shown in Figure 4d, 3-way ANOVA demonstrated that TGFβ upregulated HDAC1 (online suppl. Table 3, p < 0.001), and SB decreased it in HEK cells (p < 0.001). The presence of klotho declined the degree of these responses (p < 0.001).

Considering that klotho serves as the co-receptor for FGF23 [7] and that klotho supplementation induced BMP7 in association with reductions of FGF23 levels in vivo, we assessed whether FGF23 influenced BMP7 expression and whether the downregulation of klotho affected this (Fig. 4e–f). Three-way ANOVA showed that in HEK cells, FGF23 reduced BMP7 expression in time- and dose-dependent manner (online suppl. Table 4, p < 0.001 for both), and the presence of klotho facilitated the downregulation of BMP7 by FGF23 (p < 0.001).

We have demonstrated that exogenous klotho protein supplementation inactivates renal RAS in various CKD models, by modulating oxidative stress, AT1 receptor internalization, Wnt and IGF signaling [19, 29, 32, 37, 38]. Similar results were obtained in the present study that klotho protein supplementation reduced renal angiotensinogen expression and ANGII levels, underlying decreases in blood pressure and albuminuria. Our present data constitute a new demonstration that klotho supplementation reversed decrements of inulin-based GFR in the remnant kidney, suggesting that the kidney damage suffered during the first 4 weeks of kidney deficit is partially reversible. Immunoprecipitation data were inconsistent with the hypothesis that klotho binds to TRPC6 to augment GFR (online suppl. Fig. 4). Certainly, klotho-induced suppression of blood pressure and RAS would enhance GFR by increasing ultrafiltration coefficient [39]. As discussed (bide infra), klotho increased BMP7 expression in remnant kidney. Both BMP7 and klotho inhibit the actions of TGFβ [6, 27], and TGFβ elicits calcium influx into mesangial cells to induce contraction [40]. These raise the postulate that klotho and BMP7 ameliorate ultrafiltration coefficient to increase GFR by antagonizing TGFβ. Collectively, the present data indicated that klotho supplementation arrested the progression of glomerulosclerosis in remnant kidney and suggest that klotho reversed renal dysfunction partly involving paracrine actions of BMP7.

In the present study, klotho protein supplementation increased serum klotho levels toward those of healthy controls (Table 1), verifying the doses of klotho used in this study. Oxidative stress, ANGII, TGFβ, and albuminuria decline endogenous klotho expression [32, 41, 42]. Indeed, exogenous klotho protein supplementation enhanced endogenous renal expression of klotho, in association with the inactivation of the abovementioned factors in remnant kidney. The main source of urine klotho is transcellular transport of klotho protein from renal interstitium to the lumen by tubular cells [43]. Our data implicated that klotho supplementation returned urine klotho to nearly normal levels [31], suggesting that both an exogenous klotho protein and an enhanced endogenous klotho participated in accumulating sufficient klotho protein in the interstitium.

Using klotho as the co-receptor, FGF23 transduces its signal to increase phosphate excretion [7]. We demonstrated that exogenous klotho protein supplementation induced endogenous klotho expression to reduce FGF23 resistance by recovering FGF23-klotho signaling, facilitating phosphate excretion to decrease serum phosphate and FGF23 levels. In contrast, an increase in phosphate excretion with high phosphate diet reduces klotho expression via Wnt pathway [44]. However, klotho is nearly equivalent to Dickkopf-related protein 1 in the ability to suppress Wnt’s biological activity [4]. In CKD models, kidney SP cells did not express high levels of renoprotective factors, including BMP7 [45]. Our in vitro data provided novel evidence that FGF23 downregulated BMP7 in HEK cells. Together, these observations indicated that klotho supplementation elevated BMP7 expression, partly through the decrease in FGF23 resulting from reduction of serum phosphate in remnant kidney, and suggest that phosphate toxicity such as FGF23 elevation participates in preventing adequate expression of renoprotective molecules in CKD.

To the best of our knowledge, this is the first demonstration that BMP7 and klotho colocalized in interstitial cells as well as tubular epithelium. Our in vitro data indicated that TGFβ induced HDAC1 and reduced BMP7 expression and that klotho antagonized these actions of TGFβ. TGFβ not only induces HDAC1 but exhibits Warburg effects to reduce acetyl-CoA availability, making it difficult to maintain normal acetylation status [46]. The reductions of CTGF and HDAC1 in KL group were consistent with that klotho counteracted TGFβ in remnant kidney [17]. Of interest, kidney SP cells are located in the interstitium of adult kidneys [45]. Furthermore, FGF23 affects the fate of mesenchymal stem cells that express klotho [47]. Taken together, these have provided the evidence that klotho supplementation enhances BMP7 expression in remnant kidney and suggest that klotho affords the microenvironment favorable for both interstitial and tubular cells to express BMP7 partially via epigenetic modulation.

The present study comprised a new presentation that klotho supplementation suppressed fibrosis index and collagen expression, with elevating BMP7 and klotho expression. Our in vitro study uncovered that klotho deficiency diminished BMP7 expression. In addition, SB increased BMP7 expression in HEK cells, being compatible with the notion that BMP7 was already reduced by basal TGFβ. Similar to hepatic growth factor, BMP7 downregulates CTGF [17, 48]. It supports chronic hypoxia hypothesis that the stimulation of pericytes by CTGF directly triggers migration, which induces capillary rarefaction [23, 49]. Klotho acts as an extracellular antagonist of Wnt [4], whose engagement of LRP6 at the cell surface is necessary for responses to multiple fibrogenic signaling pathways in pericytes [23]. Indeed, renal expression of LEF1, collagen I, and TIMP1 was reduced in KL group. These could partly account for fibrosis reversal in KL group. Consequently, these results denoted that klotho supplementation reversed interstitial fibrosis in remnant kidney and suggest that klotho supplementation represses various fibrotic signals including TGFβ and Wnt to allow BMP7 expression, participating in the reversal of fibrosis [27].

The study has limitations. First, healthy control and remnant kidney with RAS inhibition were not included [50]. Thus, only relative benefit of the treatment was assessed. Caution is needed before generalizing the results. However, the in vitro studies showed that klotho directly inhibited responses to TGFβ. Second, the contribution of epithelial-mesenchymal transition to fibrosis varies among kidney disease models [51]. It remains to be determined whether klotho improved the phenotypes themselves or attenuated the development of disease-related phenotypes of interstitial cells. Although remnant kidney does not show typical progressive CKD in mice [52], further studies using fate-tracing methods are necessary to draw final conclusions [53]. Third, FGF23 activates ERK [54]. It is likely that in addition to HDAC1, ERK may reduce BMP7 in the adult kidney. Finally, the complexity of growth factors in pericyte/fibroblast biology may limit the efficacy of therapy for fibrosis with single factor. Klotho protein supplementation inhibits not only TGFβ but also Wnt, IGF, and RAS, raising the possibility that it may result in better outcomes.

In summary, the present data indicated that klotho supplementation reduced blood pressure and albuminuria with inactivation of renal RAS. Our findings provided the evidence that exogenous klotho supplementation increased endogenous klotho expression to recover FGF23-klotho signaling, facilitating phosphate excretion with reductions in serum phosphate and FGF23. The current results suggest that klotho protein supplementation reversed renal dysfunction and fibrosis at least partly by elevating BMP7 expression in remnant kidney. This study may provide the translationally important information of klotho protein supplementation, which ethically appears more acceptable than gene transfer, for clinical trials in managing progressive CKD.

The authors thank Ms. Maho Yamashita, Hiroko Sano, Yumi Sakane, and Maiko Sato for providing technical assistance. We thank Editage (www.editage.com) for English language editing. Parts of the present data were presented at Renal Week 2019 in Washington, DC, November 2019, and published as an abstract.

All animal experiments were performed at Keio University, and the Keio University Institutional Animal Care and Use Committee approved the experimental protocol (16078-(0)). This study was performed by strictly complying with recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health.

The authors have no conflicts of interest to declare.

This study was supported by a grant from the Japanese Ministry of Education, Culture, Sports, Science and Technology (JSPS KAKENHI 20K08641).

Tsuneo Takenaka and Hiromichi Suzuki designed the study; Arif Hasan, Takashi Miyazaki, Akira Nishiyama, and Naohito Ishii carried out the experiments; Tsutomu Inoue and Yoshifumi Kurosaki analyzed the data; Takeshi Marumo provided the reagents; and Tsuneo Takenaka and Matsuhiko Hayashi drafted and revised the paper. All authors approved the final version of the manuscript.

The data supporting the findings of this study are available within the article and its supplementary information file. Further inquiries can be made to the corresponding author.