Histochemical Examination on Periodontal Tissues of Klotho-Deficient Mice Fed With Phosphate-Insufficient Diet

Abstract

To elucidate which of elevated serum concentration of inorganic phosphate (Pi) or disrupted signaling linked to αklotho/fibroblast growth factor 23 (FGF23) is a predominant regulator for senescence-related degeneration seen in αKlotho-deficient mice, we have examined histological alteration of the periodontal tissues in the mandibular interalveolar septum of αKlotho-deficient mice fed with Pi-insufficient diet. We prepared six groups of mice: wild-type, kl/kl, and αKlotho^−/− mice with normal diet or low-Pi diet. As a consequence, kl/kl^norPi and αKlotho^−/−norPi mice showed the same abnormalities in periodontal tissues: intensely stained areas with hematoxylin in the interalveolar septum, dispersed localization of alkaline phosphatase–positive osteoblasts and tartrate-resistant acid phosphatase–reactive osteoclasts, and accumulation of dentin matrix protein 1 in the osteocytic lacunae. Although kl/kl^lowPi mice improved these histological abnormalities, αKlotho^{−/− lowPi} mice failed to normalize those. Gene expression of αKlotho was shown to be increased in kl/kl ^lowPi specimens. It seems likely that histological abnormalities of kl/kl mice have been improved by the rescued expression of αKlotho, rather than low concentration of serum Pi. Thus, the histological malformation in periodontal tissues in αKlotho-deficient mice appears to be due to not only increased concentration of Pi but also disrupted αklotho/FGF23 signaling.

Introduction

The periodontal tissues including periodontal ligament (PDL) and alveolar bone are a compound tissue assembled adequately for oral function, for example, in response to occlusion, pronunciation, mechanical stress, aging, and so forth. Alveolar bone has unique histological properties harmonized with mechanical stress: The physiological distal movement of the molars causes the continuous bone resorption and bone formation on the mesial and distal sides of the interradicular septum, due to the compression and tensile force. PDL encompassing the alveolar bone is a typical site of dense fibrous connective tissue containing many extracellular matrix proteins—abundant type I collagen, type III collagen, fibronectin, proteoglycans, and oxytalan fiber.^1,2 PDL seems to dynamically serve a major remodeling function by providing cells that are able to form and resorb the attachment apparatus, that is, alveolar bone, cementum, and the PDL itself. Indeed, there are several reports on the continuous and fast turnover of periodontal collagen in PDL.^1,3–9

Aging is one physiological factor for chronological alteration of morphology and function on the periodontal tissue. Investigations on senescent rats, marmosets, and human verified that the periodontal tissue showed atrophic and degenerative changes on the histology—narrowed periodontal spaces, irregularly oriented periodontal fibers, and so forth.^10–14 A circumstance of estrogen deficiency, that is, a postmenopausal state, in addition to mechanical stress was reported to affect alveolar bone remodeling.^15–20 In ovariectomized rats, consistently, a decrease in bone volume of mandible has been demonstrated.²¹ Thus, periodontal tissue changes according to senescence.

One of the important factors for aging is a klotho molecule: Genetic variants in Klotho have been associated with human aging,²² and a klotho protein has been shown to be a circulating factor detectable in serum that declines with age.²³ There are two types of klotho: αklotho is a circulating hormone regulating mineral homeostasis, whereas its homolog, βklotho, is related to bile acid/cholesterol metabolism.^24,25 In a normal state, biological functions of αklotho have been reported to regulate serum calcium (Ca) and inorganic phosphate (Pi) homeostasis, serving as a cofactor for the fibroblast growth factor (FGF) receptor 1c (FGFR1c) in FGF23 signaling.^26–29 αKlotho expression has been detected in some tissues including kidney,²⁹ and the signaling linked to αklotho/FGF23 is predominantly found in the proximal renal tubules, where it inhibits Pi reabsorption and 1α-hydroxylase activity, thereby reducing serum Pi and activation of vitamin D3.^{26,28,30–32} Mutations of the αKlotho in mice (kl/kl mice) produce a syndrome that seems to considerably accelerate the multiple age-sensitive traits in mice.³³ The phenotype of kl/kl mice is osteoporosis, skin atrophy, ectopic calcifications, pulmonary emphysema, gonadal dysplasia, defective hearing, hypervitaminosis D, hypercalcemia, and hyperphosphatemia.^33,34 These morphological features are similar to those found in the FGF23-deficient (Fgf23^−/−) mice,^33,35,36 and in mice with deletion of αKlotho gene (αKlotho^−/− mice) generated by Nabeshima’s group.³⁷ Taken together, bone abnormalities seen in kl/kl mice may result from the highly elevated serum Pi consequent to the defective αklotho/FGF23 axis.

Excessive retention of Pi in the body of the experimental animals could accelerate the aging process by inflicting widespread tissue atrophy, including hypogonadism and reduced overall survival.³⁸ Kl/kl and αKlotho^−/− mice show excessive Pi retention in serum, as early as 3 weeks of age.^{32–34,39,40} But, it is veiled whether abnormalities of age-sensitive traits seen in these kl/kl and αKlotho^−/− mice are due to disrupted signaling linked to αklotho/FGF23 or resultant elevated concentration of serum Pi. Kl/kl mice are generated by miss-sensing mutation in the promoter region of αKlotho gene which makes αklotho molecule unstable, whereas αKlotho^−/− mice artificially lack the αKlotho gene by gene targeting technique and never express αKlotho gene in any circumstance.³⁷ If the histological abnormalities of αKlotho-deficient mice are caused by elevated concentration of Pi, it seems possible the low-Pi diet would rescue the abnormality of these mice.

In our study, we have investigated histological alteration of the PDL and mandibular alveolar bone of kl/kl mice when fed with Pi-insufficient diet, to elucidate which, serum concentration of Pi or a αklotho molecule, is a predominant regulator for senescence-related degeneration of periodontal tissue.

Materials and Methods

Animals and Tissue Preparation

Four-week-old male wild-type, kl/kl, and αKlotho^−/− mice (n=12 for each; CLEA Japan, Tokyo, Japan), were used in this study, which followed the principles for care and research use of animals set by Hokkaido University (Approval No. 16-0027). Some mice were kept with normal diet containing 1.17% Ca and 0.69% Pi for another 3 weeks, and the other mice were kept fed with a diet containing 1.17% Ca and 0.415% Pi and 3.0% sevelamer hydrochloride, a polymeric amine that binds Pi which is used in treating high Pi levels in the blood (Chugai Pharmaceutical Co., Ltd., Tokyo, Japan), for the same duration. Therefore, six groups of mice were prepared: wild-type mice with normal diet (wild-type ^norPi mice), wild-type mice with low-Pi diet (wild-type ^lowPi mice), kl/kl mice with normal diet (kl/kl ^norPi mice), kl/kl mice with low-Pi diet (kl/kl ^lowPi mice), αKlotho^−/− mice with normal diet (αKlotho^{−/− norPi} mice), and αKlotho^−/− mice with low-Pi diet (αKlotho^{−/− lowPi} mice).

Before fixation, these mice were anesthetized with an intraperitoneal injection of chloral hydrate for measuring body weight and collecting blood to estimate the concentration of serum Pi, and then fixed by perfusion with 4% paraformaldehyde diluted in 0.1-M cacodylate buffer (pH 7.4) through the cardiac left ventricle. Left mandibles were immediately removed and immersed with the same fixative for 18 hr at 4C. The samples were decalcified with 10% ethylenediaminetetraacetic acid disodium salt (EDTA-2Na) solution. They were dehydrated in ascending alcohol solutions before paraffin embedding and sectioning. Before perfusion, left kidneys and right mandibles were extracted from all mice for RT-PCR and real-time PCR to examine αKlotho and Fgfr1c gene expression.

Histochemisry for Tissue-Nonspecific Alkaline Phosphatase (ALPase), Dentin Matrix Protein 1 (DMP-1), Osteopontin, FGF23, and Tartrate-Resistant Acid Phosphatase (TRAPase)

After inhibition of endogenous peroxidase activity with methanol containing 0.3% hydrogen peroxidase for 30 min, dewaxed paraffin sections were pretreated with 1% BSA (Serologicals Proteins Inc., Kankakee, IL) in PBS (1% BSA–PBS) for 30 min. Sections were then incubated for 2 to 3 hr at room temperature with rabbit polyclonal antisera against tissue-nonspecific ALPase⁴¹ at a dilution of 1:300 with 1% BSA–PBS, rabbit polyclonal antibody against DMP-1 (Code No. M176; TaKaRa Bio Inc., Otsu, Japan) at a dilution of 1:500, or rabbit polyclonal antisera to osteopontin (Code No. 25715-1-AP; Cosmo Bio Co., Ltd., Tokyo, Japan) at a dilution of 1:2000 for 1 hr. The sections thus treated were followed by incubation with horseradish peroxidase (HRP)–conjugated anti-rabbit IgG (Chemicon International Inc., Temecula, CA) for 1 hr. For detection of FGF23, the dewaxed paraffin sections were reacted with rat monoclonal anti-mouse FGF23 (Catalog No. MAB 26291; Clone No. 283507; R&D systems Inc., Minneapolis, MN) diluted at 1:100 for 2 hr. These sections were subsequently incubated in HRP-conjugated anti-rat IgG (Chemicon International Inc.). For visualization of all immunoreactions, diaminobenzidine tetrahydrochloride was used as a substrate. Normal rabbit serum or normal rat serum was used for negative control experiments concerning DMP-1, osteopontin and ALPase, or FGF23 IHC, and nonspecific binding was virtually absent.

TRAPase was detected as previously described⁴²; in short, slides were rinsed with PBS and incubated in a mixture of 2.5 mg of naphthol AS-BI phosphate (Sigma-Aldrich Co., LLC., St. Louis, MO), 18 mg of red violet LB (Sigma-Aldrich) salt, and 100-mM l-(+)-tartaric acid (0.76 g) diluted in 30 ml of a 0.1-M sodium acetate buffer (pH 5.0) for 15 min at 37C. All sections were counterstained with methyl green and observed under light microscopy (Eclipse E800; Nikon Instruments Inc., Tokyo, Japan).

Statistical Analysis

Statistical analyses on body weight and the serum concentration of Pi were performed using Microsoft Excel 2003 (Microsoft Corporation, Redmond, WA), with differences among groups being assessed by unpaired Student’s t-tests and considered statistically significant at p<0.05.

Polymerase Chain Reaction and Real-Time PCR for αKlotho and Fgfr1c

Left kidneys and right mandibles were harvested from wild-type ^norPi, wild-type ^lowPi, kl/kl ^norPi, and kl/kl ^lowPi mice. Mandibles were immediately frozen by liquid nitrogen and crushed into small pieces. The kidneys and crushed mandibles were homogenized in 10-ml TRIzol reagent (Life Technologies Co., Carlsbad, CA) per 1-g tissue to extract total RNAs. The mixture was centrifuged at 15,000 rpm for 5 min at 4C, allowing for removal of small debris. The supernatant was transferred to a new tube, which was vortexed for 15 sec after addition of 2-ml of chloroform. The lysate was then transferred to a new tube and incubated for 5 min at room temperature. After phase separation, the aqueous phase containing the RNA was transferred to a fresh new tube and RNA was precipitated by adding 5-ml isopropyl alcohol per 10-ml TRIzol reagent. After 10-min incubation at room temperature, the mixture was centrifuged for 60 min at 15,000 rpm at 4C. The resulting RNA pellet was washed with 1-ml 75% ethanol and briefly air-dried. RNA was dissolved in 30-µl diethyl pyrocarbonate (DEPC)–treated water. First-strand cDNA was synthesized from 2 µg of total RNA by SuperScript VILO cDNA Synthesis Kit (Life Technologies).

The primer sequences used for PCR were as follows: mouse Gapdh forward: TGTCTTCACCACCATG GAGAAGG, reverse: GTGGATGCAGGGATGATGTT CTG; mouse Fgfr1c forward: CTTGACGTCGTGGAAC GATCT, reverse: AGAACGGTCAACCATGCAGAG; and mouse αKlotho forward: GGGTCACTGGGTCA ATCT, reverse: GCAAAGTAGCCACAAAGG. The PCR was performed using a thermal cycler (GeneAmp PCR System 2700; Applied Biosystems, Foster City, CA) as follows: denaturation at 94C for 30 sec, annealing at 60C (for Gapdh) and 55C (for Fgfr1c and αKlotho) for 30 sec, extension at 72C for 30 sec, and a final incubation at 72C for 10 min. RT-PCR products were subjected to 2% agarose gel electrophoresis, stained with ethidium bromide, and detected using E-Gel Imager (Life Technologies).

Real-time PCR assays were performed in a final reaction volume of 50 µl that consisted of 25.0 µl of 2× QuantiFast SYBR Green PCR Master Mix (QIAGEN GmbH, Hilden, Germany), 2.5-µl of each primer, 1.0 ml of template DNA, and 19.0 µl of DEPC-treated water. All reactions were performed thrice. The primer sequences used for real-time PCR were as follows: mouse Fgfr1c forward: 5′-GACTGCTG GAGTTAATACCA-3′, reverse: 5′-CTGGTCTCTCTT CCAGGGCT-3′ and mouse αKlotho forward: 5′-G AGGGCAATAAGGTAGTGAACAGA-3′, reverse: 5′-A AGCCATACTGGTTTGATAGCTCG-3′. Cycling conditions involved an initial activation step of 50C for 2 min and 94C for 15 min, followed by 50 cycles of 94C for 15 sec and annealing at 55C (Gapdh) or 53C (Fgfr1c and αKlotho) for 30 sec, extension at 72C for 35 sec, and a final incubation at 95C for 15 sec, 60C for 1 min, and 95C for 15 sec, using ABI 7500 (Applied Biosystems). Real-time PCR on mouse β-actin was performed using QuantiTect Primer Assay (Mm_Actb_2_SG; Catalog No. QT01136772; QIAGEN GmbH).

Results

Body Weights and the Serum Concentration of Pi

Consistent with the appearance of whole body size at the 7 weeks of age (Fig. 1A), the average body weight of wild-type ^lowPi mice was lower than wild-type ^norPi mice, whereas that of kl/kl ^lowPi mice was markedly increased compared with kl/kl ^norPi mice (Fig. 1B). Serum concentration of Pi tended to be attenuated in wild-type ^lowPi and kl/kl ^lowPi mice when compared with their counterparts fed with normal diet though no significant difference (Fig. 1C).

Figure 1.

Body weights and the serum concentration of Pi in wild-type and kl/kl mice fed with normal and Pi-insufficient diet. (A) Appearance of whole body size at the 7 weeks of age in wild-type ^norPi, kl/kl ^norPi, wild-type ^lowPi, and kl/kl ^lowPi mice. The average body weight of wild-type ^lowPi mice significantly decreases compared with wild-type ^norPi mice. In contrast, the body weight of kl/kl ^lowPi mice increases compared with kl/kl ^norPi mice (B; **p<0.01). Serum concentration of Pi shows no significant difference in kl/kl ^lowPi mice and kl/kl ^lowPi mice (C). The values are expressed as mean ± SE. Abbreviation: Pi, phosphate.

Histology of Interalveolar Septum of Wild-Type Mice and kl/kl Mice Fed With Normal and Pi-Insufficient Diet

Interalveolar septum between the first and second molars of wild-type ^norPi and wild-type ^lowPi mice showed intact histology of alveolar bone and the similar thickness of PDL, therefore indicating no histological change with low-Pi diet (Fig. 2A, B, E, and F). Alternatively, kl/kl ^norPi mice exhibited intensely stained areas with hematoxylin in the alveolar bone and narrow PDL (Fig. 2C and G). However, kl/kl ^lowPi mice made the histology of interalveolar septum and the thickness of PDL normalized as can be seen in the wild-type mice, though leaving a few patchy staining of hematoxylin (Fig. 2D and H).

Figure 2.

Histological analysis of interalveolar septum in wild-type and kl/kl mice fed with normal and Pi-insufficient diet. Panels (E)–(H) are highly magnified images of panels (A)–(D). Hematoxylin–eosin staining demonstrates the similar histology of alveolar bone and PDL in the interalveolar septum between the first and second molars of wild-type ^norPi and wild-type ^lowPi mice (A, B, E, F). The regions intensely stained with hematoxylin (arrows in C and G) are observed in the alveolar bone and narrow PDL of kl/kl ^norPi mice (C, G). However, these histological alterations are rescued by feeding low-Pi diet as shown in kl/kl ^lowPi mice (D, H). Note only a few regions of intensely stained with hematoxylin in alveolar bones of kl/kl ^lowPi mice (arrows in D and H). Abbreviations: Pi, phosphate; PDL, periodontal ligament; is, interalveolar septum. Bars: A–D = 150 µm; E–H = 50 µm.

Interalveolar septum is physiologically subjected to compression and tensile force in the mesial and distal regions, respectively, and therefore, we investigated those regions of wild-type and kl/kl mice (Fig. 3). Both wild-type ^norPi and wild-type ^lowPi mice displayed scalloped surfaces in the mesial regions of the interalveolar septum, whereas they revealed smooth bone surfaces in the distal regions (Fig. 3A–D). Unlikely, kl/kl ^norPi mice demonstrated the irregular shapes at both mesial and distal regions, and amorphous materials intensely stained with hematoxylin in the superficial layer of alveolar bone and in the cementum (Fig. 3E and F). In kl/kl ^norPi mice, PDL cells were irregularly dispersed in the narrow periodontal space, and several osteocytic lacunae were empty. With low-Pi diet, however, kl/kl ^lowPi mice came to show scalloped and smooth surfaces at the mesial and distal regions, respectively (Fig. 3G and H).

Figure 3.

Histology of mesial and distal regions in the interalveolar septum of wild-type and kl/kl mice fed with normal and Pi-insufficient diet. In wild-type ^norPi and wild-type ^lowPi mice, the mesial regions of the interalveolar septum demonstrate scalloped bone surfaces (A, C), whereas the distal regions demonstrate smooth bone surfaces (B, D). Kl/kl ^norPi mice shows the abnormal features at both mesial and distal regions—amorphous materials intensely stained with hematoxylin (arrows) in the superficial layer of alveolar bone and in the cementum (E, F). There are several empty osteocytic lacunae in the alveolar bone. Note the disturbed orientation of PDL cells in the narrow periodontal space. However, kl/kl ^lowPi mice show the intact histology of alveolar bone and PDL as can be seen in the wild-type mice (G, H). Abbreviations: Pi, phosphate; PDL: periodontal ligament; is, interalveolar septum. Bars: A–H = 20 µm.

Improved Localization of ALPase-Positive Osteoblasts and TRAPase-Reactive Osteoclasts in the Interalveolar Septum in kl/kl Mice With Low-Pi Diet

Distal smooth surfaces of wild-type ^norPi and wild-type ^lowPi interalveolar septum had many osteoblasts with an intense ALPase positivity, therefore clearly demonstrating no change in the osteoblasts’ distribution by Pi-insufficient diet (Fig. 4A, B, E, and F). Alternatively, kl/kl ^norPi mice revealed the scattered distribution of osteoblasts weakly positive for ALPase on the mesial and distal surfaces of the interalveolar septum (Fig. 4C and G). But, being fed with low-Pi diet, kl/kl ^lowPi mice localized ALPase-positive osteoblasts on the distal surface (Fig. 4D and H). TRAPase-reactive osteoclasts tended to accumulate on the mesial surfaces of wild-type ^norPi and wild-type ^lowPi alveolar bones (Fig. 5A, B, E, and F). However, kl/kl ^norPi mice did not show many TRAPase osteoclasts, but instead, some loci of TRAPase-positive cement lines could be seen in the central regions of the interalveolar septum (Fig. 5C and G). After feeding with low-Pi diet, TRAPase-reactive osteoclasts came to accumulate on the mesial surface of kl/kl ^lowPi mice (Fig. 5D and H).

Figure 4.

Immunolocalization of ALPase in the interalveolar septum of wild-type and kl/kl mice fed with normal and Pi-insufficient diet. Panels (E)–(H) are the highly magnified images of the distal walls of interalveolar septum (A–D). Distal walls of interalveolar septum of wild-type ^norPi mice (arrows, E), wild-type ^lowPi mice (arrows, F), and kl/kl ^lowPi mice (arrows, H) display intense ALPase immunoreactivity (brown color, A, B, D, E, F, H) and smooth bone surfaces. However, in kl/kl ^norPi interalveolar bone, weak ALPase positivity can be seen (C, G). Abbreviations: ALPase, alkaline phosphatase; Pi, phosphate; is, interalveolar septum. Bars: A–D = 50 µm; E–H = 20 µm.

Figure 5.

Distribution of TRAPase-positive osteoclasts in the interalveolar septum of wild-type and kl/kl mice fed with normal and Pi-insufficient diet. Panels (E)–(H) are the highly magnified images of the mesial walls of interalveolar septum (A–D). TRAPase-reactive osteoclasts (red color) are located on the mesial surfaces of wild-type ^norPi, wild-type ^lowPi, and kl/kl ^lowPi interalveolar septum (arrows, A, B, D, E, F, H). However, TRAPase-positive osteoclasts are hardly seen in kl/kl ^norPi mice (C, G), but instead, some TRAPase-positive cement lines (arrows, G) are discernible in the central regions of the interalveolar septum. Abbreviations: TRAPase, tartrate-resistant acid phosphatase; Pi, phosphate; is, interalveolar septum. Bars: A–D = 50 µm; E–H = 20 µm.

The Normalized Distribution of DMP-1 and Osteopontin in kl/kl Mice With Low-Pi Diet

DMP-1 immunoreactivity was detected in the regions of osteocytes and their canaliculi of wild-type ^norPi and wild-type ^lowPi interalveolar septum (Fig. 6A, B, E, and F). In kl/kl ^norPi mice, DMP-1 strongly accumulated in the osteocytic lacunae and in the cementum, and sometimes in spherical materials in the periodontal space (Fig. 6C and G). Alternatively, low-Pi diet resulted in the normally distributed DMP-1 as can be seen in the wild-type mice, although some regions kept DMP-1 accumulation in the osteocytic lacunae in kl/kl ^lowPi mice (Fig. 6D and H).

Figure 6.

Immunolocalization of DMP-1 in the interalveolar septum of wild-type and kl/kl mice fed with normal and Pi-insufficient diet. DMP-1 positivity (brown color) is found in osteocytes and their canaliculi located in the interalveolar septum of wild-type ^norPi and wild-type ^lowPi (A, B, E, F). In contrast, kl/kl ^norPi interalveolar septum displays many patchy materials with DMP-1 immunoreactivity in osteocytic lacunae and on cementum (arrows, C, G). When being fed with low-Pi diet, DMP-1 immunoreactivity of kl/kl ^lowPi looks similar to that seen in the wild-type mice, although some regions show DMP-1 accumulation in the osteocytic lacunae in kl/kl ^lowPi mice (arrows, D, H). Abbreviations: DMP-1, dentin matrix protein 1; Pi, phosphate; is, interalveolar septum. Bars: A–D = 50 µm; E–H = 20 µm.

Wild-type ^norPi and wild-type ^lowPi mice localized osteopontin immunoreactivity mainly on the scalloped surfaces of the mesial regions of the alveolar bone (Fig. 7A, B, E, and F), whereas unlikely, kl/kl ^norPi mice had amorphous materials with abundant osteopontin in their alveolar bone (Fig. 7C and D). However, kl/kl ^lowPi mice revealed the normalized distribution of osteopontin on the mesial surface of the alveolar bone, which looked similar to that seen in the wild-type mice (Fig. 7D and H). Thus, low-Pi diet appeared to normalize the histological abnormalities in kl/kl mice.

Figure 7.

Immunolocalization of osteopontin in the interalveolar septum of wild-type and kl/kl mice fed with normal and Pi-insufficient diet. Panels (E)–(H) are the highly magnified images of the mesial regions of interalveolar septum (A–D). The mesial surfaces of interalveolar septum are intensely stained with osteopontin in wild-type ^norPi and wild-type ^lowPi mice (brown color, arrows, E, F). Alternatively, amorphous materials bearing an intense osteopontin reactivity (arrows) are present in kl/kl ^norPi alveolar bone (C, G). In kl/kl ^lowPi mice, however, immunolocalization of osteopontin (arrows) shows the same distribution as wild-type mice in the mesial region of interalveolar septum (D, H). Abbreviations: Pi, phosphate; is, interalveolar septum. Bars: A–D = 50 µm; E–H = 20 µm.

Abnormal Histology of the Interalveolar Septum Is Not Improved With Pi-Insufficient Diet in αKlotho^−/− Mice

We wondered which of αklotho/FGF23-driven signaling or serum concentration of Pi controlled by kidney is essential for improvement of histological abnormalities seen in kl/kl mice. Therefore, we performed the same experiments on αKlotho^−/− mice by feeding a low-Pi diet. Unlike kl/kl mice, however, αKlotho^−/− mice did not improve their histological abnormalities with low-Pi diet (Fig. 8). There were intensely stained amorphous materials with hematoxylin in the interalveolar septum both of αKlotho^{−/− norPi} and αKlotho^{−/− lowPi} mice (Fig. 8A, B, C, and D). TRAPase-reactive osteoclasts were scattered without showing the tendency to locate on the mesial surface (Fig. 8E and F). DMP-1 accumulated in the osteocytic lacunae, and intense osteopontin was discernible in the amorphous material in αKlotho^{−/− norPi} and αKlotho^{−/− lowPi} mice (Fig. 8G–J). Thus, there could not be seen any significant difference in histology of αKlotho^−/− mice when fed with low-Pi diet.

Figure 8.

Histochemical analyses of interalveolar septum in αKlotho^−/− mice fed with normal and Pi-insufficient diet. Panels (C)–(J) are the magnified images of the region indicated by arrows in panels A and B. Both of αKlotho^−/−norPi and αKlotho^−/−lowPi mice exhibit the areas intensely stained with hematoxylin in the interalveolar septum (arrows, C, D). TRAPase-positive osteoclasts are not obvious in the mesial region (E, F). Note that DMP-1 accumulates in the osteocytic lacunae (brown color, arrows, G, H), and abundant osteopontin is discernible in the amorphous material (brown color, arrows, I, J). Note the distribution of amorphous materials strongly stained with hematoxylin (C, D), DMP-1 (G, H), and osteopontin (I, J) is similar between αKlotho^−/−norPi and αKlotho^−/−lowPi interalveolar septum. Abbreviations: Pi, phosphate; TRAPase, tartrate-resistant acid phosphatase; DMP-1, dentin matrix protein 1; is, interalveolar septum. Bars: A, B = 150 µm; C–F = 50 µm; G–J = 30 µm.

Elevated αKlotho Gene Expression in Kidney and Mandible of kl/kl Mice Fed With Low-Pi Diet

RT-PCR demonstrated an intense band representing αKlotho expression in the kl/kl ^lowPi mice, when compared with kl/kl ^norPi mice (Fig. 9A). Consistently, real-time PCR showed the elevated level of αKlotho gene in kl/kl ^lowPi mice compared with kl/kl ^norPi mice, whereas it did not reach the levels seen in wild-type specimens (Fig. 9B). There was no obvious difference in the expression level of Fgfr1c mRNA among the specimens from wild-type ^norPi, wild-type ^lowPi, kl/kl ^norPi, and kl/kl ^lowPi mice (Fig. 9C).

Figure 9.

Gene expression of Fgfr1c and αKlotho in kidney and mandibles of kl/kl mice fed with normal and phosphate (Pi)-insufficient diet. RT-PCR shows relatively intense band representing αKlotho gene in the kidney of kl/kl ^lowPi mice compared with that of kl/kl ^norPi mice (A). In real-time PCR analysis, αKlotho gene is elevated in kl/kl ^lowPi mice compared with kl/kl ^norPi mice, whereas it does not reach the levels of wild-type (WT) specimens (B). No obvious difference can be seen in the expression level of Fgfr1c mRNA among the specimens from WT ^norPi, WT ^lowPi, kl/kl ^norPi, and kl/kl ^lowPi mice (C). Panels (D)–(G) show FGF23 immunolocalization in the interalveolar septum. FGF23 immunoreactivity is observed in osteocytes (arrows in D, E, and G). Real-time PCR shows an elevated expression of αKlotho gene in kl/kl ^lowPi specimens compared with kl/kl ^norPi mice (H). Notice no obvious differences concerning Fgfr1c mRNA among all groups (I). Bars: D–G = 50 µm.

FGF23 immunoreactivity was observed in osteocytes in the wild-type ^norPi and wild-type ^lowPi mice (Fig. 9D and E). In contrast, kl/kl ^norPi mice showed less immunoreactivity for FGF23 in alveolar osteocytes (Fig. 9F). With the low-Pi diet, however, osteocytes in the alveolar bone of kl/kl ^lowPi mice were intensely FGF23 immunoreactivity, similar to be seen in wild-type specimens (Fig. 9G). Real-time PCR verified an elevated expression of αKlotho gene in kl/kl ^lowPi specimens compared with kl/kl ^norPi mice (Fig. 9H). There were no obvious differences concerning Fgfr1c mRNA among all groups (Fig. 9I).

Discussion

In our study, the disrupted signaling linked to αklotho/FGF23 in kl/kl and αKlotho^−/− mice resulted in abnormal histology of the periodontal tissue of interalveolar septum—senescence-related alterations including narrowed periodontal spaces, amorphous structures in the PDL, and irregular distribution of periodontal fibers, which can be consistently seen in human.^11,14 Therefore, it is meaningful to examine a mouse model for providing a clue for understanding the mechanism of senescence in human periodontal tissue. To our knowledge, this is the first report that the low-Pi diet rescued the abnormalities in the periodontal tissues of kl/kl mice. We found that histological abnormalities were not improved by feeding low-Pi diet to αKlotho^−/− mice and also that αKlotho gene was elevated in kidney and mandibles of kl/kl ^owPi mice. Therefore, we hypothesize the possibility that histological abnormalities seen in the interalveolar septum are due to not only increased concentration of Pi but also disrupted αklotho/FGF23 signaling in the periodontal tissue.

The average body weight of wild-type ^lowPi mice was decreased more than that of wild-type ^norPi mice. However, the wild-type ^lowPi interalveolar septum did not histologically change. Unexpectedly, there was no significant decrease of serum concentration of Pi corrected before sacrificing kl/kl ^lowPi mice. To accurately measure the serum Pi concentration, it seems necessary to correct blood in the same time every day. But, it seems impossible to get enough volume of blood to measure the serum concentration of Pi from mice every day. In our own study, kl/kl ^lowPi mice grew well like wild-type counterparts, and the average body weight of kl/kl ^lowPi mice increased close to the wild-type ^lowPi and wild-type ^norPi mice. Unlike kl/kl ^norPi mice, the mobility of kl/kl ^lowPi mice was very active as well as the wild-type mice (data not shown). Therefore, though no significant reduction of serum Pi when sacrifice, kl/kl ^lowPi mice looked to improve their health.

In kl/kl ^norPi mice, less numbers of ALPase-positive osteoblasts and TRAPase-positive osteoclasts were seen on the interalveolar septum, and in addition, the PDL cells were irregularly arranged unlike those in wild-type mice. It seems therefore suspicious for the occurrence of physiological tooth movement in kl/kl ^norPi mice. However, once normalized with low-Pi diet, kl/kl ^lowPi mice featured many ALPase-positive osteoblasts and TRAPase-reactive osteoclasts in the distal and mesial regions of the interalveolar septum, which presumably implies the restart of tooth movement in kl/kl mice.

The histological improvement was not only the localization of ALPase-positive osteoblasts and TRAPase-reactive osteoclasts but also the distribution of DMP-1 and osteopontin. Consistent with kl/kl incisors and femora in our recent study,^43,44 alveolar bone showed several empty lacunae and accumulation of DMP-1 in the osteocytic lacunae and some DMP-1-positive cells in the periodontal space. Our findings suggest that histological abnormalities such as the unusual synthesis of DMP-1 and osteopontin are unique to disrupted signaling of αklotho/FGF23, which may permit the inference that αklotho deficiency renders osteocytes dysfunctional. However, it is veiled why DMP-1 and osteopontin, which have high affinity to crystalline Ca²⁺, have been overproduced in the circumstance of the absence of αklotho.

Our striking finding is that the abnormal periodontal tissues of the interalveolar septum of kl/kl ^lowPi mice were rescued by the low-Pi diet. As bone abnormalities of kl/kl mice have been reported to link to highly elevated serum Pi levels, due to defects in the FGF23–klotho axis,^{33,34,38–40} it is our surprise that αKlotho^−/− mice did not rescue the abnormalities by the low-Pi diet. Kl/kl ^lowPi mice were shown to express more αKlotho in both kidney and mandibles—It is higher than in kl/kl ^norPi mice but not as high as in the wild-type counterparts. This apparently implies that the synthesis of αklotho molecules has been partially, at least, improved by feeding with low-Pi diet, consistent with the in vitro reports.^45,46 However, the average of serum Pi concentration was not significantly decreased in kl/kl ^lowPi mice. It seems possible that temporal or intermittent decrease of serum Pi may be enough to facilitate the promoter activity of αKlotho gene which is usually inhibited in kl/kl mice.⁴⁵ Elevated synthesis of αklotho, even though it does not attain the level of normal mice, appears to begin to transduce the signaling of αklotho/FGF23 in kidney and periodontal tissues. In kidney, this may accelerate Pi reabsorption by mediating NaPi IIa/IIc in the proximal tubules and also normalize the activity of 1α-hydroxylase for synthesis of active form of vitamin D3.^{26,28,30–32} The findings that low-Pi diet did not rescued the histological abnormalities in αKlotho ^lowPi mice implicate the possibility that other signaling in addition to inhibition of serum Pi may be important for improving the abnormalities in the kl/kl mice.

The recovery of αKlotho gene expression in the mandible of αKlotho ^lowPi mice suggests an explanation for the local function of αklotho/FGF23 signaling. Rhee et al. have reported that both FGFR1c and klotho transcripts are expressed by osteocytes and osteoblasts.⁴⁷ This study also confirmed the expression of Fgfr1c and αKlotho gene in intact alveolar bone, thereby supporting the idea that αklotho/FGF23 signaling may act in an autocrine/paracrine manner in bone. Based on our results, we hypothesize that while the renal αklotho/FGF23 signaling may control the mineral balance in serum, αklotho/FGF23 signaling in the periodontal tissues including the alveolar bone may be involved in tissue maintenance, as shown by our results. Indeed, we have shown that FGF23 immunoreactivity was markedly reduced in the alveolar bone of kl/kl ^norPi mice. Such reduced FGF23 immunoreactivity may be due to empty osteocytic lacunae seen in kl/kl ^norPi mice; on the contrary, the histological recovery of the periodontal integrity in kl/kl ^lowPi mice rescued the synthesis of FGF23 in osteocytes. However, this putative local action of FGF23/αklotho in bone and in the periodontium may not be so simple. Recent investigations showed that FGFR4 is expressed in bone and in periodontal tissues^48,49 and that a Klotho-dependent, FGFR1- and FGFR4-mediated signaling mechanism that may also involve Janus kinase 3 is at work in renal tubules.⁵⁰ Therefore, a different experimental model seems necessary to clarify the molecular mechanism of FGF23/αklotho signaling mediated by FGFR4.

In summary, our study demonstrated that a low-Pi diet improved the histological abnormalities of the periodontal tissues and increased the gene expression of αKlotho in kidney and mandible of kl/kl mice. In contrast, a low-Pi diet failed to normalize the abnormalities in αKlotho^−/− mice. These findings indicate that the senescence-related histological abnormality of the periodontal tissues in αKlotho-deficient mice may be due to not only increased concentration of Pi but also disrupted αklotho/FGF23 signaling in the periodontal tissue.