Quantitative ultrastructural analyses of rabbit osteoclasts. The effect of in vivo treatment with sodium salicylate

Rabbit tibial osteoclasts were examined by electron microscopy and quantitative data on their ultrastructural morphology collected by methods described by Holtrop. Osteoclasts from the tibiae of two groups of rabbits were compared: one fed a commercial diet and the other fed the same diet containing 2% sodium salicylate (v/w). No changes in the total number of tibial osteoclasts were detected but average osteoclast size, numbers of nuclei per osteoclast and ruffled border and clear zone areas decreased (p less than 0.05), as did the proportion of osteoclasts directly attached to bone. These results suggest that osteoclast activity is inhibited by in vivo salicylate therapy.     

Correlation of an osteoclast antigen and ruffled border on giant cells formed in response to resorbable substrates

The osteoclast is the specialized multinucleated cell primarily responsible for the degradation of the organic and inorganic components of bone matrix. The functional and developmental relationship between osteoclasts and foreign body giant cells is unclear. The osteoclast plasma membrane ruffled border juxtaposed to the bone surface is a unique morphologic characteristic of active osteoclasts. In the studies reported here giant cell formation was induced in response to a variety of materials implanted onto the richly vascularized chick chorioallantoic membrane. Light and electron microscopic techniques were used to examine the morphologic characteristics of the giant cells. In addition, immunohistochemical methods were used to demonstrate the appearance of a 150 kD cell surface antigen on chicken osteoclasts recognized by monoclonal antibody 121F. Giant cells that formed in response to mineralized bone particles exhibited ruffled borders and stained positively with the 121F antibody. Many giant cells that formed in response to hydroxyapatite possessed ruffled borders similar to but not as extensive as those observed on giant cells formed on bone. Immunohistochemical localization of the 121F antigen on these cells suggested that the antigen was present, but staining intensity was reduced compared to that of bone-associated giant cells. The formation of mineral matrix complexes by the adsorption to hydroxyapatite of bone extract or osteocalcin enhanced ruffled borders and the presence of the 121F antigen on elicited giant cells. In contrast, giant cells that formed on non-resorbable materials, such as Sepharose beads, mica, and methacrylate, lacked ruffled borders and were negative for the 121F antigen. It appears that expression of the 121F osteoclast antigen correlates with the appearance and extent of ruffled membranes on giant cells. Furthermore, it appears that giant cell ruffled membrane development and the presence of the 121F osteoclast antigen are related to giant cell formation in response to resorbable materials that are subject to extracellular dissolution. Expression of this antigen may be indicative of the developmental and/or functional state of giant cells (osteoclasts) that form on resorbable substrates. In addition, components of the bone matrix, including osteocalcin, in association with bone mineral, lead to elevated levels of this osteoclast antigen.     

Abnormal osteoclast morphology and bone remodeling in a murine model of a lysosomal storage disease

Mucopolysaccharidosis type VII (MPS VII) is a heritable lysosomal storage disease caused by a deficiency in beta-glucuronidase (GUSB) activity, leading to progressive accumulation of undegraded glycosaminoglycans in many tissues. Clinical features include growth and mental retardation, hearing and visual defects, shortened lifespan, and skeletal deformities. A murine model of MPS VII has been described that shares many of the manifestations of the human disease, including the skeletal dysplasia. In this study we describe abnormalities in the cellular morphology and function of osteoclasts and a localized defect in bone formation rate in the MPS VII mouse. Ultrastructural analysis revealed that MPS VII osteoclasts fail to form ruffled border membranes and many appeared to be detached from the bone surface. Following bone marrow transplantation, osteoclasts derived from wild-type donors showed normal morphology and were closely associated with the bone surface in MPS VII recipients. In vitro bone resorption assays demonstrated that MPS VII osteoclasts formed significantly smaller and fewer pits than those formed by osteoclasts derived from normal mice of the same strain. Although osteoclast morphology and function appeared to be abnormal in the MPS VII mouse, interleukin-1 (IL-1)-induced osteoclastogenesis in vivo was not affected. In addition to the osteoclast defects, MPS VII mice demonstrated a slower rate of bone matrix deposition in the epiphysis by in vivo calcein labeling experiments. These data suggest that abnormal morphology and function of MPS VII osteoclasts, combined with deficient matrix deposition, may contribute to the skeletal defects observed in this lysosomal storage disease.     

Alendronate Disturbs Vesicular Trafficking in Osteoclasts

The nitrogen-containing bisphosphonate alendronate inhibits osteoclast-mediated bone resorption through inhibition of the mevalonate pathway. This results in impaired protein prenylation and may affect the function of small GTPases in osteoclasts. Since these proteins are important regulators of vesicle transport in cells, we investigated the possible interference of alendronate with these processes in isolated rat osteoclasts. We show here that alendronate-induced inhibition of bone resorption coincides with accumulation of tartrate-resistant acid phosphatase- and electron dense material-containing tubular vesicles in osteoclasts. Alendronate-induced changes in osteoclasts also included widening of the sealing zone areas and incomplete organization of tight attachments and ruffled borders. Osteoclasts also appeared partially detached from the bone surface, and organic matrix was typically dissolved only at the edges of the resorption pits on alendronate-coated bone slices. In contrast, resorption pits on the control and clodronate-coated bone slices were thoroughly resorbed. Inhibition of bone resorption by alendronate was not, however, related to a decrease in osteoclast number. In conclusion, our findings suggest that alendronate inactivates osteoclasts by mechanisms that impair their intracellular vesicle transport, apoptosis being only a secondary phenomenon to this.     

Ultrastructure, tartrate-resistant acid phosphatase activity and calcitonin responsiveness of osteoclasts at sites of demineralized bone matrix implant-induced osteogenesis

The morphology, ultrastructure, tartrate-resistance acid phosphatase reactivity, and calcitonin responsiveness of osteoclasts induced at sites of demineralized bone matrix (DBM) implant-induced osteogenesis in rats were determined. Osteoclasts at these ectopic sites had a morphologic and ultrastructural appearance similar to osteoclasts normally found in skeletal tissues. When observed by scanning electron microscopy, resorption surfaces on the implants had well-defined resorption pits (Howship's lacunae), indicative of active bone resorption. The osteoclasts stained intensely for tartrate-resistance acid phosphatase, an enzyme that is specific for osteoclasts. In response to human calcitonin, hypocalcemia occurred and osteoclasts lost their ruffled borders, indicating that these cells are responsive to exogenous hormonal stimulation. The osteoclasts induced by subcutaneous implantation of DBM had morphologic and functional characteristics similar to osteoclasts normally found in skeletal tissues.     

Bisphosphonate acts on osteoclasts independent of ruffled borders in osteosclerotic (oc/oc) mice

We examined the effects of a third generation bisphosphonate [YM-175; disodium dihydrogen (cycloheptylamino)-methylene-1,1-bisphosphonate] on osteoclasts in osteosclerotic (oc/oc) mice to elucidate the cellular mechanism for incorporation of the bisphosphonate. Osteoclasts of oc/oc mice were in direct contact with bone matrix but devoid of ruffled borders. Tartrate-resistant acid phosphatase (TRAPase) showed spotty localization intercellularly, whereas vacuolar H⁺-ATPase (V-ATPase) immunoreactivity was observed homogeneously in the cytoplasm. Upon injection of bisphosphonate, most osteoclasts lost cell polarity and were detached from bone surfaces. The detached osteoclasts underwent apoptosis as characterized by condensation of chromatin, absence of Golgi apparatus, and formation of many vesicles in the cytoplasm. Both TRAPase and V-ATPase were evenly distributed in the cytoplasm. The pyknotic nuclei of osteoclasts revealed DNA fragments as evidenced by the terminal deoxynucleotidyltransferase-mediated dUTP-biotin nick end-labeling (TUNEL) method. The results indicate that osteoclasts lacking ruffled borders in oc/oc mice incorporated the bisphosphonate from a site different from ruffled borders and that bisphosphonate may directly affect osteoclasts without mediating its deposition to the bone matrix.     

Ultrastructural localization of a tartrate-resistant acid ATPase in bone

Osteoclasts are effector cells in bone breakdown, and the active bone resorption is confined to the ruffled border zone of these cells. An acid milieu is maintained in this zone which is probably a prerequisite for bone resorption. Tartrate-resistant acid phosphatase (TRAP) activity has been recognized as a characteristic property of osteoclasts and in several studies proposed as a cytochemical marker of osteoclasts. We have previously isolated and characterized a tartrate-resistant and iron-activated acid ATPase (TrATPase) from rat bone, the enzyme being a member of the TRAP family. In the present study the ultrastructural localization of this enzyme was delineated by employing immunogold technique on low temperature-embedded maxillar rat bone. Intensive immunolabeling was seen on the bone surfaces facing the ruffled border zone while lower amounts of marker were seen in adjacent bone areas, that is, on the bone surfaces facing the clear zone and deeper-into the bone. Within the osteoclasts gold markers were observed mainly in vesicular structures interpreted as lysosomes. Immunolabeling was also observed in the recently described endocytic cells located near osteoblasts and osteoclasts. Also in these cells, the marker was confined to lysosomelike structures. The amount of label in bone facing osteoblasts was low, as was the amount within osteoblasts. Our observation of extracellular localization, in particular accumulation of TrATPase in bone matrix facing the ruffled border area of the osteoclasts, favors the view that the enzyme is exported to areas of active bone resorption, thereby indicating a potential role for the enzyme in this process.     

Cellular biology and biochemical mechanism of bone resorption. A review of recent developments on the formation, activation, and mode of action of osteoclasts

The newest knowledge on the osteoclast allows us to consider bone resorption in a global perspective, as the resultant of three successive steps that may each be individually regulated by physiopathologic or pharmacologic agents. The first involves the formation of osteoclast progenitors in hematopoietic tissues followed by their vascular dissemination and the generation of resting preosteoclasts and osteoclasts in bone. The second consists in the activation of osteoclasts at the contact of mineralized bone. Osteoblasts appear to control this step by exposing the mineral to osteoclasts and preosteoclasts and/or by releasing a soluble factor that activates these cells. In a third step, activated osteoclasts resorb both the mineral and the organic of mineralized bone through the action of agents that they secrete in the segregated zone underlying their ruffled border. The mineral appears to be solubilized by hydrogen ions secreted by an ATP-driven proton pump located at that border and fed by protons generated from CO2 by carbonic anhydrase. The removal of organic matrix, which could be prepared by osteoblast collagenase at the level of nonmineralized bone surfaces, appears dependent on acid proteinases, particularly cysteine-proteinases, secreted, together with other lysosomal enzymes, in the acid microenvironment of the resorption zone.     

Polarized osteoclasts put marks of tartrate-resistant acid phosphatase on dentin slices--a simple method for identifying polarized osteoclasts

Osteoclasts form ruffled borders and sealing zones toward bone surfaces to resorb bone. Sealing zones are defined as ringed structures of F-actin dots (actin rings). Polarized osteoclasts secrete protons to bone surfaces via vacuolar proton ATPase through ruffled borders. Catabolic enzymes such as tartrate-resistant acid phosphatase (TRAP) and cathepsin K are also secreted to bone surfaces. Here we show a simple method of identifying functional vestiges of polarized osteoclasts. Osteoclasts obtained from cocultures of mouse osteoblasts and bone marrow cells were cultured for 48 h on dentin slices. Cultures were then fixed and stained for TRAP to identify osteoclasts on the slices. Cells were removed from the slices with cotton swabs, and the slices subjected to TRAP and Mayer's hematoxylin staining. Small TRAP-positive spots (TRAP-marks) were detected in the resorption pits stained with Mayer's hematoxylin. Pitted areas were not always located in the places of osteoclasts, but osteoclasts existed on all TRAP-marks. A time course experiment showed that the number of TRAP-marks was maintained, while the number of resorption pits increased with the culture period. The position of actin rings formed in osteoclasts corresponded to that of TRAP-marks on dentin slices. Immunostaining of dentin slices showed that both cathepsin K and vacuolar proton ATPase were colocalized with the TRAP-marks. Treatment of osteoclast cultures with alendronate, a bisphosphonate, suppressed the formation of TRAP-marks and resorption pits without affecting the cell viability. Calcitonin induced the disappearance of both actin rings and TRAP-marks in osteoclast cultures. These results suggest that TRAP-marks are vestiges of proteins secreted by polarized osteoclasts.     

Osteoclast Fusion and Fission

Osteoclasts are specialized multinucleated cells with the unique capacity to resorb bone. Despite insight into the various steps of the interaction of osteoclast precursors leading to osteoclast formation, surprisingly little is known about what happens with the multinucleated cell itself after it has been formed. Is fusion limited to the short period of its formation, or do osteoclasts have the capacity to change their size and number of nuclei at a later stage? To visualize these processes we analyzed osteoclasts generated in vitro with M-CSF and RANKL from mouse bone marrow and native osteoclasts isolated from rabbit bones by live cell microscopy. We show that osteoclasts fuse not only with mononuclear cells but also with other multinucleated cells. The most intriguing finding was fission of the osteoclasts. Osteoclasts were shown to have the capacity to generate functional multinucleated compartments as well as compartments that contained apoptotic nuclei. These compartments were separated from each other, each giving rise to a novel functional osteoclast or to a compartment that contained apoptotic nuclei. Our findings suggest that osteoclasts have the capacity to regulate their own population in number and function, probably to adapt quickly to changing situations.     

An assay system utilizing devitalized bone for assessment of differentiation of osteoclast progenitors.

The present study provides a novel assay system to examine the differentiation of osteoclast progenitors on devitalized bone slices. We used the population of bone cells liberated enzymatically from 14-day-old mouse embryonal calvariae as a source of osteoclast progenitors. The analysis of differentiation of osteoclast progenitors into preosteoclasts and mature osteoclasts was assessed in terms of the formation of TRAP-positive cells and pits or resorption lacunae, respectively, on devitalized bone slices. Osteoclasts having bone-resorbing activity appeared when the calvarial cell population was cultured in the presence of 1 alpha,25-(OH)2D3 on devitalized bone slices. The resorbing activity increased in a 1 alpha,25-(OH)2D3 dose-related manner. However, calcitonin, a potent inhibitor of differentiation and activation of osteoclast lineage cells, reduced the area of the resorption lacunae in a dose-dependent fashion. The bone-resorbing cells on the bone slices expressed an obvious ruffled border and clear zone, structures specific to mature osteoclasts. These results suggest that osteoclast progenitors in the mouse calvarial population examined differentiated into mature osteoclasts in the presence of 1 alpha,25-(OH)2D3 on devitalized bone slices. Further, using this assay system we assessed the effect of some other osteotropic factors on the differentiation of osteoclast progenitors to mature osteoclasts. IL-1, IL-6, and PTH increased the formation of TRAP-positive cells and pits and the area of resorption lacunae in a dose-dependent fashion. However, prostaglandin E2 was unable to induce the formation of resorption lacunae, although a significant appearance of TRAP-positive cells was observed at a concentration of 200 ng/ml.(ABSTRACT TRUNCATED AT 250 WORDS)     

The rapid appearance of acid phosphatase activity at the developing ruffled border of parathyroid hormone activated medullary bone osteoclasts

Summary The localization of acid phosphatase (ACPase) activity in and near parathyroid hormone (PTH) activated osteoclasts was investigated using electron microscopic cytochemical methods. At 3 hours after oviposition in Japanese Quail hens, medullary bone osteoclasts were highly reactive for ACPase but lacked ruffled borders. There was no evidence of extracellular ACPase activity associated with these osteoclasts. At 20 minutes after PTH administration, osteoclasts had developing ruffled borders and ACPase activity was found in the matrix and extracellular space adjacent to most of these ruffled borders. ACPase activity was seldom observed beyond the resorption zone delineated eated by the osteoclast clear zones. These results provide direct cytochemical evidence that the ruffled border functions in the release and/or activation of ACPase. In addition, these results show that ACPase localization is rapidly responsive to exogenous PTH.     

Osteoclast TGF‐β Receptor Signaling Induces Wnt1 Secretion and Couples Bone Resorption to Bone Formation

Osteoblast‐mediated bone formation is coupled to osteoclast‐mediated bone resorption. These processes become uncoupled with age, leading to increased risk for debilitating fractures. Therefore, understanding how osteoblasts are recruited to sites of resorption is vital to treating age‐related bone loss. Osteoclasts release and activate TGF‐β from the bone matrix. Here we show that osteoclast‐specific inhibition of TGF‐β receptor signaling in mice results in osteopenia due to reduced osteoblast numbers with no significant impact on osteoclast numbers or activity. TGF‐β induced osteoclast expression of Wnt1, a protein crucial to normal bone formation, and this response was blocked by impaired TGF‐β receptor signaling. Osteoclasts in aged murine bones had lower TGF‐β signaling and Wnt1 expression in vivo. Ex vivo stimulation of osteoclasts derived from young or old mouse bone marrow macrophages showed no difference in TGF‐β–induced Wnt1 expression. However, young osteoclasts expressed reduced Wnt1 when cultured on aged mouse bone chips compared to young mouse bone chips, consistent with decreased skeletal TGF‐β availability with age. Therefore, osteoclast responses to TGF‐β are essential for coupling bone resorption to bone formation, and modulating this pathway may provide opportunities to treat age‐related bone loss. © 2015 American Society for Bone and Mineral Research.     

Osteoclast polarization is not required for degradation of bone matrix in rachitic FGF23 transgenic mice

Hypophosphatemic transgenic (tg) mice overexpressing FGF23 in osteoblasts display disorganized growth plates and reduced bone mineral density characteristic of rickets/osteomalacia. These FGF23 tg mice were used as an in vivo model to examine the relation between osteoclast polarization, secretion of proteolytic enzymes and resorptive activity. Tg mice had increased mRNA expression levels of the osteoblast differentiation marker Runx2 and mineralization-promoting proteins alkaline phosphatase and bone sialoprotein in the long bones compared to wild type (wt) mice. In contrast, expression of alpha1(I) collagen, osteocalcin, dentin matrix protein 1 and osteopontin was unchanged, indicating selective activation of osteoblasts promoting mineralization. The number of osteoclasts was unchanged in tg compared to wt mice, as determined by histomorphometry, serum levels of TRAP 5b activity as well as mRNA expression levels of TRAP and cathepsin K. However, tg mice displayed elevated serum concentrations of C-terminal telopeptide of collagen I (CTX) indicative of increased bone matrix degradation. The majority of osteoclasts in FGF23 tg mice lacked ultrastructural morphological signs of proper polarization. However, they secreted both cathepsin K and MMP-9 at levels comparable to osteoclasts with ruffled borders. Mineralization of bone matrix thus appears essential for inducing osteoclast polarization but not for secretion of osteoclast proteases. Finally, release of CTX by freshly isolated osteoclasts was increased on demineralized compared to mineralized bovine bone slices, indicating that the mineral component limits collagen degradation. We conclude that ruffled borders are implicated in acidification and subsequent demineralization of the bone matrix, however not required for matrix degradation. The data collectively provide evidence that osteoclasts, despite absence of ruffled borders, effectively participate in the degradation of hypomineralized bone matrix in rachitic FGF23 tg mice.     

Signaling axis in osteoclast biology and therapeutic targeting in the RANKL/RANK/OPG system

Bone integrity is maintained through a balance between bone formation and bone resorption, and osteoclasts are primary cells involved in bone resorption. Recent studies have revealed an essential role of macrophage colony-stimulating factor (M-CSF) and receptor activator of nuclear factor kappa-B ligand (RANKL) in the development of osteoclasts, and detailed molecular cascades that induce osteoclast differentiation, activation and apoptosis have been clarified. Osteoclasts are involved in various pathologic conditions, such as osteoporosis, rheumatoid arthritis and tumor-induced bone disease, which are characterized by abnormal bone resorption, and the finding of RANKL has provided us a good therapeutic target for such pathologic conditions.     

A scrutiny of matrix metalloproteinases in osteoclasts: evidence for heterogeneity and for the presence of MMPs synthesized by other cells

Genetic diseases and knockout mice stress the importance of matrix metalloproteinases (MMPs) in skeletal turnover. Our study aims at clarifying which MMPs are expressed by osteoclasts. Previous analyses of this basic question led to conflicting reports in the literature. In the present study, we used a variety of approaches: PCR, Northern blots, Slot blots, in situ hybridization, and immunohistochemistry. We analyzed osteoclasts in culture as well as osteoclasts in native bone at different locations and compared mouse and rabbit osteoclasts. Osteoclasts express MMP-9 and -14 in all conditions, although to a variable extent, and they are able to synthesize MMP-3, -10, and -12, at least under some circumstances. The induction of a given MMP in osteoclasts is influenced by its environment (e.g., osteoclast culture vs. native bone, and various sites within the same bone) and depends on the species (e.g., mouse vs. rabbit). Osteoclasts show high amounts of MMP-2 and -13 protein presumably made to a large extent by other cells, thereby documenting how proteinases of nonosteoclastic origin may contribute to osteoclast activities and giving insight in why the resorptive activity of purified osteoclasts appears insensitive to MMP inhibitors. Our study shows that the confusion about osteoclastic MMPs in the literature reflects the remarkable ability of osteoclasts to adapt to their environment, as required by the structural or functional diversity of bone tissue. Our observations provide basic information needed for understanding the emerging role of MMPs in controlling cell signaling and bone resorption.     

The mouse osteopetrotic grey-lethal mutation induces a defect in osteoclast maturation/function

The osteopetrotic grey-lethal (gl) mouse mutant displays many similarities to the human malignant autosomal-recessive form of osteopetrosis. In this study, we show that the gl osteopetrotic bone phenotype is characterized by the presence of numerous differentiated multinucleated osteoclasts. A significant increase in the number of tartrate-resistant acid phosphatase (TRAP)-positive osteoclasts was detected in vivo, suggesting induction of differentiation in the osteoclast lineage as a compensatory mechanism. These gl osteoclast cells demonstrated a defective cytoskeletal reorganization and an underdeveloped ruffled border, a membrane structure essential for active bone resorption. Accordingly, resorption activity of these cells is markedly impaired by four- to tenfold as evaluated with the pit formation assay. This low bone resorption in gl osteoclasts is highly reminiscent of the loss in key enzymes, V-ATPase or cathepsin-K, and in signaling factors, Src or TRAF-6, which were shown not to be significantly altered in gl osteoclasts. Thus, independently of a deficiency in V-ATPase, Src, cathepsin-K, and TRAF-6, the gl mutation results in increased number of osteoclasts, characterized by a disrupted cytoskeleton and an underdeveloped ruffled border.     

Electron microscopy of developing calvaria reveals images that suggest that osteoclasts engulf and destroy osteocytes during bone resorption

Summary It is generally accepted that osteoclasts are responsible for the breakdown and removal of bone matrix constituents. However, very little is known about the fate of osteocytes during bone resorption. In the present study we have examined sites of bone destruction in calvaria of young rats aged 4–9 days in the hope of obtaining information on the fate of osteocytes. Decalcified glutaraldehyde-formaldehyde-fixed specimens were prepared for ultrathin section electron microscopy. When sequentially arranged, the images obtained suggest that osteoclasts engulf and destroy osteocytes during bone degradation. We propose that the following sequence of events takes place when a lacuna is opened up by an osteoclast: (1) When the osteoclast comes in contact with an osteocyte, the villi of the ruffled border become flat and broad. (2) Long osteoclastic extensions surround the osteocyte. (3) The osteocyte is subsequently internalized with apparent degradation.     

Regulation of osteoclast differentiation and function by phosphate: potential role of osteoclasts in the skeletal abnormalities in hypophosphatemic conditions.

Mice fed with a low Pi diet exhibited decreased osteoclast number. Hyp mice also showed decreased osteoclasts, and high Pi reversed it. Low Pi reduced osteoclast formation and bone resorption in vitro. Hypophosphatemia may suppress osteoclast differentiation/function, leading to skeletal abnormalities. INTRODUCTION: Skeletal abnormalities seen in hypophosphatemic disorders indicate a critical role of phosphate (Pi) in skeletogenesis. However, the role of osteoclasts in the pathogenesis of the disturbed skeletogenesis is unclear. MATERIALS AND METHODS: Mice fed with a low-Pi diet and Hyp mice that are characterized by hypophosphatemia and impaired osteogenesis were studied. Effects of Pi on osteoclast formation and bone resorption were also examined in vitro. RESULTS: Histomorphometric examination showed that mice on a low-Pi diet exhibited decreased osteoclast number. Furthermore, osteoclast number in Hyp mice was also decreased compared with wildtype (WT) mice. Of note, feeding of Hyp mice with high-Pi diet significantly reversed hypophosphatemia, improved disturbed osteogenesis, and increased osteoclast number. Osteoclast-like cell (OLC) formation and bone resorption in Hyp bone marrow cells was not different from WT bone marrow cells. On the other hand, OLC formation and bone resorption were decreased in conjunction with reduced mRNA expression of RANKL in WT bone marrow cells cultured in the medium containing low Pi (0.5 mM). Recombinant human matrix extracellular phosphoglycoprotein (MEPE), a candidate for phosphatonin, also decreased osteoclast formation, whereas fibroblast growth factor 23 (FGF23), another phosphatonin candidate, showed no effects. CONCLUSIONS: Our results suggest that Pi controls the differentiation and function of osteoclasts. These actions of Pi on osteoclasts may be associated with the pathogenesis of the skeletal abnormalities in hypophosphatemic disorders.