Osteocalcin abnormalities in Hyp mice reflect altered genetic expression and are not due to altered clearance, affinity for mineral, or ambient phosphorus levels

PHEXL222P Mutation Increases Phex Expression in a New ENU Mouse Model for XLH Disease

PhexL222P mouse is a new ENU mouse model for XLH disease due to Leu to Pro amino acid modification at position 222. PhexL222P mouse is characterized by growth retardation, hypophosphatemia, hypocalcemia, reduced body bone length, and increased epiphyseal growth plate thickness and femur diameter despite the increase in PHEXL222P expression. Actually, PhexL222P mice show an increase in Fgf23, Dmp1, and Mepe and Slc34a1 (Na-Pi IIa cotransporter) mRNA expression similar to those observed in Hyp mice. Femoral osteocalcin and sclerostin and Slc34a1 do not show any significant variation in PhexL222P mice. Molecular dynamics simulations support the experimental data. P222 might locally break the E217-Q224 β-sheet, which in turn might disrupt inter-β-sheet interactions. We can thus expect local protein misfolding, which might be responsible for the experimentally observed PHEXL222P loss of function. This model could be a valuable addition to the existing XLH model for further comprehension of the disease occurrence and testing of new therapies.

FGF23 and its role in X-linked hypophosphatemia-related morbidity

Background

X-linked hypophosphatemia (XLH) is an inherited disease of phosphate metabolism in which inactivating mutations of the Phosphate Regulating Endopeptidase Homolog, X-Linked (PHEX) gene lead to local and systemic effects including impaired growth, rickets, osteomalacia, bone abnormalities, bone pain, spontaneous dental abscesses, hearing difficulties, enthesopathy, osteoarthritis, and muscular dysfunction. Patients with XLH present with elevated levels of fibroblast growth factor 23 (FGF23), which is thought to mediate many of the aforementioned manifestations of the disease. Elevated FGF23 has also been observed in many other diseases of hypophosphatemia, and a range of animal models have been developed to study these diseases, yet the role of FGF23 in the pathophysiology of XLH is incompletely understood.

Methods

The role of FGF23 in the pathophysiology of XLH is here reviewed by describing what is known about phenotypes associated with various PHEX mutations, animal models of XLH, and non-nutritional diseases of hypophosphatemia, and by presenting molecular pathways that have been proposed to contribute to manifestations of XLH.

Results

The pathophysiology of XLH is complex, involving a range of molecular pathways that variously contribute to different manifestations of the disease. Hypophosphatemia due to elevated FGF23 is the most obvious contributor, however localised fluctuations in tissue non-specific alkaline phosphatase (TNAP), pyrophosphate, calcitriol and direct effects of FGF23 have been observed to be associated with certain manifestations.

Conclusions

By describing what is known about these pathways, this review highlights key areas for future research that would contribute to the understanding and clinical treatment of non-nutritional diseases of hypophosphatemia, particularly XLH.

Calcium and vitamin D: what is known about the effects on growing bone

The objective of these investigations was to determine if the receptor-dependent effects of 1,25-dihydroxyvitamin D were essential for normal skeletal growth. Mice with targeted ablation of the vitamin D receptor were engineered, and the skeletal consequences of vitamin D receptor ablation were studied in the presence of normal and abnormal mineral ion homeostasis. Prevention of abnormal mineral ion homeostasis resulted in the development of a normal skeleton in the absence of a functional vitamin D receptor. The metabolic cause of rickets was found to be hypophosphatemia. The major receptor-dependent actions of 1,25-dihydroxyvitamin D on skeletal development are indirect and are a reflection of the role of this hormone on intestinal calcium absorption.

Correction of the mineralization defect in hyp mice treated with protease inhibitors CA074 and pepstatin

Increased expression of several osteoblastic proteases and MEPE (a bone matrix protein) occurs in X-linked hypophosphatemic rickets (hyp). This is associated with an increased release of a protease-resistant MEPE peptide (ASARM peptide), a potent inhibitor of mineralization. Cathepsin B cleaves MEPE releasing ASARM peptide and hyp osteoblast/osteocyte cells hypersecrete cathepsin D, an activator of cathepsin B. Our aims were to determine whether cathepsin inhibitors correct the mineralization defect in vivo and whether hyp-bone ASARM peptide levels are reduced after protease treatment. Normal littermates and hyp mice (n = 6) were injected intraperitoneally once a day for 4 weeks with pepstatin, CAO74 or vehicle. Animals were then sacrificed and bones plus serum removed for comprehensive analysis. All hyp mice groups (treated and untreated) remained hypophosphatemic with serum 1,25 vitamin D3 inappropriately normal. Serum PTH was significantly elevated in all hyp mice groups relative to normal mice (P = 0.0017). Untreated hyp mice had six-fold elevated levels of serum alkaline-phosphatase and two-fold elevated levels of ASARM peptides relative to normal mice (P < 0.001). In contrast, serum alkaline phosphatase and serum ASARM peptides were significantly reduced (normalized) in hyp mice treated with CA074 or pepstatin. Serum FGF23 levels remained high in all hyp animal groups (P < 0.0001). Hyp mice treated with protease inhibitors showed dramatic reductions in unmineralized osteoid (femurs) compared to control hyp mice (Goldner staining). Also, hyp animals treated with protease inhibitors showed marked and significant improvements in growth plate width (42%), osteoid thickness (40%) and cortical area (40%) (P < 0.002). The mineralization apposition rate, bone formation rate and mineralization surface were normalized by protease-treatment. High-resolution pQCT mineral histomorphometry measurements and uCT also confirmed a marked mineralization improvement. Finally, the growth plate and cortical bone of hyp femurs contained a massive accumulation of osteoblast-derived ASARM peptide(s) that was reduced in hyp animals treated with CA074 or pepstatin. This study confirms in vivo administration of cathepsin inhibitors improves bone mineralization in hyp mice. This may be due to a protease inhibitor mediated decrease in proteolytic degradation of the extracellular matrix and a reduced release of ASARM peptides (potent mineralization inhibitors).

The role of the PHEX gene (PEX) in families with X-linked hypophosphataemic rickets

For over a hundred years, the bane of rickets (a disease of bone), has been prominent in those countries that have participated in, and seeded, the industrial revolution. Industrialisation had major effects of the demography of populations, and many people moved to dark, heavily industrialised cities to find work. It soon became apparent that rickets could be cured by supplementing the diet with cod liver oil and exposure to sunlight. This in turn led to the discovery that photoactivation of 7-dehydrocholesterol was required to produce vitamin D, an indispensable regulator of bone mineral metabolism. Although inadequate exposure to light and poor dietary intake are the main causes of rickets and osteomalacia, recent research has confirmed the role of familial, and tumour forms of the disease. This review will describe the recent advances in our knowledge of the molecular defects in X-linked hypophosphataemic rickets (HYP), and oncogenic hypophosphataemic osteomalacia (OHO). Although HYP and OHO have different primary defects, both diseases have similarities that suggest a linked or overlapping pathophysiology. Also, without doubt, the recent cloning of the gene defective in HYP (the PHEX gene), has given researchers a new reagent to explore the molecular regulation of bone and its links to kidney endocrine function. The fact that the PHEX gene codes for a Zn metallopeptidase raises new and intriguing questions, and adds new momentum to the research on diseases of bone mineral metabolism.

Osteocalcin production in primary osteoblast cultures derived from normal and Hyp mice

Rickets and osteomalacia are characteristic features of the Hyp mouse model of human X-linked hypophosphatemia. Hyp mice demonstrate elevated circulating osteocalcin levels, as well as altered regulation of osteocalcin by 1,25(OH)2D3. Whether this osteocalcin abnormality is intrinsic to the osteoblast, or mediated by the in vivo milieu, has not been established. We therefore characterized osteocalcin production and its regulation by 1,25(OH)2D3 in primary cultures of murine osteoblasts and examined osteocalcin and its messenger RNA in response to 1,25(OH)2D3 in cultures of Hyp mouse-derived osteoblasts. Cell viability and osteocalcin production are optimal when murine cells are harvested within 36 h of age. Murine primary osteoblast cultures mineralize and produce osteocalcin in a maturation-dependent fashion (as demonstrated in other species), and continuous exposure to 1,25(OH)2D3, beginning at day 9 of culture, inhibits osteoblast differentiation and osteocalcin production and prevents mineralization of the culture. However, in contrast to other species, exposure to 1,25(OH)2D3, added later (days 17-25) in culture, does not stimulate osteocalcin but arrests osteocalcin production at current levels. Ambient media levels of osteocalcin were no different in cultures from Hyp mice and their normal litter mates, and the down-regulatory response to 1,25(OH)2D3 was comparable in cultures from normal and Hyp mice. Furthermore, expression of osteocalcin messenger RNA in murine cultures is reduced with exposure to 1,25(OH)2D3, and there is no difference between normal and Hyp cultures in this response. Thus, primary murine osteoblasts manifest a species-specific effect of 1,25(OH)2D3 on osteocalcin production. Furthermore, the increased serum osteocalcin production seen in intact Hyp mice, and the altered response to 1,25(OH)2D3 in Hyp mice, are not observed in osteoblast cultures derived from the mutant strain. These data indicate that abnormalities of osteocalcin described in intact Hyp mice require factors other than those present in cultured cells.

Regulation of osteocalcin production and bone resorption by 1,25-dihydroxyvitamin D3 in mouse long bones: interaction with the bone-derived growth factors TGF-beta and IGF-I

Bone cells produce multiple growth factors that have effects on bone metabolism and can be incorporated into the bone matrix. Interplay between these bone-derived growth factors and calciotropic hormones has been demonstrated in cultured bone cells. The present study was designed to extend these observations by examining the interactions between either transforming growth factor-beta (TGF-beta) or insulin-like growth factor-I (IGF-I) and 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) in a mouse long bone culture model with respect to osteocalcin production and bone resorption. In contrast to the stimulation in rat and human, in the fetal mouse long bone cultures, 1,25(OH)2D3 caused a dose-dependent inhibition of osteocalcin production. Both the osteocalcin content in the culture medium and in the extracts of the long bones was reduced by 1,25(OH)2D3. This effect was not specific for fetal bone because 1,25(OH)2D3 also reduced osteocalcin production by the neonatal mouse osteoblast cell line MC3T3. TGF-beta inhibited whereas IGF-I dose-dependently increased osteocalcin production in mouse long bones. The combination of TGF-beta and 1,25(OH)2D3 did not result in a significantly different effect compared with each of these compounds alone. The IGF-I effect was completely blocked by 1,25(OH)2D3. In the same long bones as used for the osteocalcin measurements, we performed bone resorption analyses. Opposite to its effect on osteocalcin, 1,25(OH)2D3 dose-dependently stimulated bone resorption. TGF-beta reduced and IGF-I did not change basal (i.e., in the absence of hormones) bone resorption. Our results show that 1,25(OH)2D3-enhanced bone resorption is dose-dependently inhibited by TGF-beta and IGF-I. Regression analysis demonstrated a significant negative correlation between 1,25(OH)2D3-induced bone resorption and osteocalcin production. The specificity for their effect on 1,25(OH)2D3-stimulated bone resorption was assessed by testing the effects of TGF-beta and IGF-I in combination with parathyroid hormone (PTH). Like 1,25(OH)2D3, PTH dose-dependently stimulates bone resorption. However, PTH-stimulated bone resorption was not affected by TGF-beta. Like 1,25(OH)2D3-stimulated bone resorption, IGF-I inhibited the PTH effect but at a 10-fold higher concentration compared with 1,25(OH)2D3. In conclusion, the present study demonstrates growth factor-specific interactions with 1,25(OH)2D3 in the control of osteocalcin production and bone. With respect to bone resorption, these interactions are also hormone specific. The present data thereby support and extend the previous observations that interactions between 1,25(OH)2D3 and bone-derived growth factors play an important role in the control of bone metabolism. These data together with the fact that TGF-beta and IGF-I are present in the bone matrix and potentially can be released during bone resorption support the concept that growth factors may control the effects of calciotropic hormones in bone in a localized and possibly temporal manner. Finally, in contrast to human and rat, in mice 1,25(OH)2D3 reduces osteocalcin production and this reduction is paralleled by stimulation of bone resorption by 1,25(OH)2D3. These data thereby show a dissociation between osteocalcin production and bone resorption.

Analysis of osteocalcin expression in transgenic mice reveals a species difference in vitamin D regulation of mouse and human osteocalcin genes

A line of transgenic mice expressing a human osteocalcin genomic fragment (hOClocus) and a murine MC3T3-E1 cell line containing a stably integrated human osteocalcin promoter construct have been developed to characterize the osteogenic and hormonal regulation of human osteocalcin in vivo and in vitro. In this study, we used these models to demonstrate a species difference in the regulation of the mouse and human osteocalcin genes by vitamin D. Repeated administration of 1,25-dihydroxyvitamin D3 (1,25(OH)2D3) to mice carrying the hOClocus transgene resulted in striking increases in serum human osteocalcin, whereas serum mouse osteocalcin levels were unchanged after 24 h and only modestly increased 48 h after the second dose of hormone. 1,25(OH)2D3 increased human calvarial mRNA expression by 1.8-fold and slightly decreased mouse osteocalcin mRNA levels by approximately 1.2-fold. Furthermore, treatment of primary calvarial osteoblasts from these mice with 1,25(OH)2D3 increased human osteocalcin production but inhibited mouse osteocalcin protein accumulation. To investigate further the mechanism for the apparent species difference in vitamin D3 induction of mouse and human osteocalcin, we examined the effect of 1,25(OH)2D3 in an MC3T3-E1 cell line (MC4) containing a stably integrated 3900 bp osteocalcin promoter-luciferase construct. Treatment of MC4 cells with ascorbic acid resulted in parallel increases of the endogenous mouse osteocalcin protein and luciferase reporter activity over a 12-day period. Continuous exposure of MC4 cells to 1,25(OH)2D3 resulted in time-and dose-dependent increases in the activity of the phOC3900 luciferase construct. By contrast, the hormone had no effect on mouse osteocalcin protein concentrations and inhibited its induction by ascorbic acid. However, when cells were treated acutely with 1,25(OH)2D3 at later times during growth in ascorbic acid, the induction of mouse osteocalcin protein was only partially inhibited. In conclusion, our results indicate that common osteogenic signals regulate both mouse and human osteocalcin gene expression, but the mouse gene is resistant to induction by vitamin D. This species difference in vitamin D regulation of osteocalcin appears to result from the failure of 1,25(OH)2D3 to transcriptionally activate the mouse osteocalcin gene.

Species-specific glucocorticoid and 1,25-dihydroxyvitamin D responsiveness in mouse MC3T3-E1 osteoblasts: dexamethasone inhibits osteoblast differentiation and vitamin D down-regulates osteocalcin gene expression

The mouse MC3T3-E1 cell line is nontumorigenic and undergoes a typical program of osteoblast differentiation in vitro, producing a bone-like mineralized extracellular matrix. We report responses of these cells to dexamethasone (Dex) and 1,25-(OH)2D3 that are in contrast to findings from other osteoblast culture systems. First, chronic exposure of both early- and late-passaged MC3T3-E1 cells to 10(-7) M Dex, initiated during the proliferation period, blocked osteoblast differentiation, in contrast to the enhanced differentiation observed in cultures of fetal rat calvarial-derived cells. Secondly, 1,25-(OH)2D3 did not up-regulate expression (messenger RNA or protein synthesis) of the endogenous mouse osteocalcin (OC) gene. Several lines of evidence are presented that suggest this response is caused by sequence specific properties of the mouse OC vitamin D response element. We also observed both qualitative and quantitative differences in expression of cell growth (histone H2B) and phenotype-related genes (collagen, OC, osteopontin, glucocorticoid receptor, and 1, 25-(OH)2D3 receptor), between pre- and postmineralization stage osteoblasts, in response to 24 h steroid hormone treatment. Our findings in MC3T3-E1 cells are consistent with current concepts of selective influences of 1,25-(OH)2D3 and glucocorticoids as a function of osteoblast maturation. However, the inhibition of osteoblast differentiation by chronic Dex at 10(-7) M and the down-regulation of OC by 1,25-(OH)2D3 are novel observations relevant to species-specific responsiveness of mouse bone-expressed genes to steroid hormones during osteoblast differentiation.