The authors state that they have no conflicts of interest
Article first published online: 27 APR 2009
Copyright © 2009 ASBMR
Journal of Bone and Mineral Research
Volume 24, Issue 9, pages 1586–1597, September 2009
How to Cite
de Freitas, P. H. L., Li, M., Ninomiya, T., Nakamura, M., Ubaidus, S., Oda, K., Udagawa, N., Maeda, T., Takagi, R. and Amizuka, N. (2009), Intermittent PTH Administration Stimulates Pre-Osteoblastic Proliferation Without Leading to Enhanced Bone Formation in Osteoclast-Less c-fos−/− Mice. J Bone Miner Res, 24: 1586–1597. doi: 10.1359/jbmr.090413
Published online on April 27, 2009
- Issue published online: 4 DEC 2009
- Article first published online: 27 APR 2009
- Manuscript Accepted: 21 APR 2009
- Manuscript Revised: 25 MAR 2009
- Manuscript Received: 11 SEP 2008
- PTH/PTH-related peptide;
- c-fos deficiency;
This study aimed to investigate the behavior and ultrastructure of osteoblastic cells after intermittent PTH treatment and attempted to elucidate the role of osteoclasts on the mediation of PTH-driven bone anabolism. After administering PTH intermittently to wildtype and c-fos−/− mice, immunohistochemical, histomorphometrical, ultrastructural, and statistical examinations were performed. Structural and kinetic parameters related to bone formation were increased in PTH-treated wildtype mice, whereas in the osteoclast-deficient c-fos−/− mice, there were no significant differences between groups. In wildtype and knockout mice, PTH administration led to significant increases in the number of cells double-positive for alkaline phosphatase and BrdU, suggesting active pre-osteoblastic proliferation. Ultrastructural examinations showed two major pre-osteoblastic subtypes: one rich in endoplasmic reticulum (ER), the hypER cell, and other with fewer and dispersed ER, the misER cell. The latter constituted the most abundant preosteoblastic phenotype after PTH administration in the wildtype mice. In c-fos−/− mice, misER cells were present on the bone surfaces but did not seem to be actively producing bone matrix. Several misER cells were shown to be positive for EphB4 and were eventually seen rather close to osteoclasts in the PTH-administered wildtype mice. We concluded that the absence of osteoclasts in c-fos−/− mice might hinder PTH-driven bone anabolism and that osteoclastic presence may be necessary for full osteoblastic differentiation and enhanced bone formation seen after intermittent PTH administration.
PTH-driven anabolism in bone involves the osteoblastic and the osteoclastic lineages. Increases in osteoblastic number have been reported after intermittent PTH administration, and putative explanations include reduction in osteoblastic apoptosis(1,2) and recruitment of bone lining cells.(3,4) Whereas some have shown that osteoblastic cell replication increases after 6 h of exposure to PTH and PTH-related peptide (PTHrP) in vitro,(5) others have reported that intermittent PTH stimulates osteoblastic differentiation by inducing exit from the cell cycle.(6) Until now, it has been unclear which of the two processes, osteoblastic proliferation or differentiation, is a more or less important determinant of intermittent PTH-driven bone anabolism. Comprehensive in vivo examinations and characterization of the cells involved in the bone anabolic responses can contribute to a deeper understanding of the events that follow PTH administration.
Osteoblastic cells are the likely main actors in PTH-driven bone anabolism, because mature osteoblasts and their precursors express the PTH receptor (PTHR).(7) Nonetheless, the function and behavior of pre-osteoblastic cells, which were first described by Pritchard,(8) have not yet been clarified. Scott(9,10) has reported on the phenotypical diversity of bone cells. In his work, spindle-shaped “A cells” were shown to incorporate [3H]thymidine, a marker for cell proliferation, and were classified as osteoblastic precursors. In another study, Martineau-Doizé et al.(11) identified three preosteoblastic phenotypes in the rat femur: an endocytic cell, a rough endoplasmic reticulum (rER)-rich cell and an undifferentiated cell. Rouleau et al.(12,13) identified and characterized PTHR-bearing pre-osteoblasts and named them “PT cells.”
The bone-resorbing osteoclast may also play an important role in PTH-driven anabolism. As previously suggested,(14) PTH administration affects bone formation and bone resorption. Given that the activities of osteoclasts and osteoblasts are intertwined during normal bone remodeling,(15) it is plausible that the anabolic action of PTH is either directly or indirectly related to the osteoclast. A single subcutaneous injection of PTH(1-38) promoted transient increases in RANKL mRNA levels accompanied by decreased osteoprotegerin (OPG) gene levels in rats,(16) which could lead to brief osteoclast activation. Li et al.(17) have shown significant upregulation of genes that stimulate osteoclastogenesis after anabolic PTH treatment, suggesting that osteoclast generation, even if transient, may be necessary for PTH-induced anabolism in bone.
One of the best ways to advance the understanding of the interactions between osteoblasts and osteoclasts and the mechanisms of the anabolic response to PTH is experimenting with animal models featuring defective bone remodeling.(18) An example of such is the c-fos knockout mouse, an osteopetrotic strain that lacks osteoclasts and their precursors.(19,20) The defect in these mice seems to be intrinsic to hematopoietic precursors, which have their progress into the osteoclastic lineage blocked in a very early stage.(21)Fos ablation does not seem to affect osteoblastic proliferation and differentiation, however; calvarial cells derived from mice heterozygous and homozygous for c-fos deletion do not differ regarding their differentiation capabilities nor their response to PTH.(18) Demiralp et al.(22) reported on the absence of an anabolic response to PTH in the c-fos knockout mouse during bone growth, whereas anabolism was rescued once c-fos−/− “vossicles” were implanted in nude mice that were subsequently treated with intermittent PTH.(18)
Interactions between osteoclasts and osteoblastic cells that are relatively early in their differentiation process may be necessary for an optimal response to intermittent PTH. In this work, we used wildtype and c-fos–deficient mice to (1) examine the ultrastructure and behavior of the osteoblastic cells that are affected by intermittent PTH and (2) attempt to elucidate the role of osteoclasts in the anabolic actions of PTH.
MATERIALS AND METHODS
Animals, protocols for PTH, calcein, and BrdU injections, and tissue preparation
Twelve 8-wk-old C57BL/6J (wildtype) mice and 12 age-matched mice homozygous for c-fos gene deletion were purchased from The Jackson Laboratory(Bar Harbor, ME, USA). Mice were kept under standard conditions, and experimental protocols were approved by Niigata University and Matsumoto Dental University. Because c-fos deletion renders mice edentulous, the same food ration given to wildtypes had to be crushed and moistened with distilled water for appropriate nutrition. Feeding trays were monitored frequently, and the food ration was kept soft and moist by adding water to it when necessary. Within each strain, mice were divided into injection and control groups receiving, respectively, 80 μg/kg of human PTH(1-34) (Sigma-Aldrich, St. Louis, MO, USA) or vehicle (0.9% saline) subcutaneously, daily, for 14 days. For bone labeling, 20 mg/kg of calcein (Dojindo Laboratories, Kumamoto, Japan) was injected intraperitoneally 72 and 24 h before death. All mice were administered with 0.2 ml of BrdU (5 mg/ml; Sigma) intraperitoneally 24 and 4 h before death.(23) After perfusion with 4% paraformaldehyde solution (pH 7.4), femora and tibias from wildtype and c-fos−/− mice were extracted, decalcified, and embedded in paraffin as previously described.(24) Specimens were observed under a Nikon Eclipse E800 microscope (Nikon Instruments, Tokyo, Japan). Light microscopy images were acquired with a digital camera (DXM1200C; Nikon). For transmission electron microscopy (TEM), fixed specimens were decalcified, postfixed with OsO4, dehydrated, and embedded in epoxy resin (Epon 812; Taab, Berkshire, UK) as previously described.(24) Specimens were observed under TEM (H-7100; Hitachi, Tokyo, Japan) at 80 kV.
Immunohistochemistry and TRACP detection
Dewaxed paraffin sections were treated for endogenous peroxidase inhibition with 0.3% H2O2 in PBS for 20 min and nonspecific staining blocking with 1% BSA in PBS (1% BSA-PBS) for 30 min at room temperature. Sections were incubated with rabbit antisera against tissue nonspecific alkaline phosphatase (ALP)(24,25) or goat anti-mouse EphB4 (R&D Systems, Minneapolis, MN, USA) for 2 h at room temperature, and incubated with horseradish peroxidase (HRP)-conjugated anti-rabbit or anti-goat secondary antibodies (Chemicon International, Temecula, CA, USA) for 1 h at room temperature. Immunoreactions were detected with 3,3′-diaminobenzidine tetrahydrochloride (DAB; Dojindo). For double immunostaining procedures, the sequential approach was used. For Runx2/ALP detection, sections were pretreated as described above with the addition of a 10-min trypsin treatment before the blocking step. Sections were incubated with mouse anti-Runx2/Cbfa1 (Medical & Biological Laboratories, Nagoya, Japan) overnight at 4°C followed by HRP-conjugated anti-mouse IgG (Chemicon) overnight at 4°C. ALP immunoreactivity was detected as described above, only with ALP-conjugated anti-rabbit IgG (Sigma) as the second antibody and with a visualization procedure described previously.(26) For ephrinB2/ALP, the sections were incubated with goat anti-mouse ephrinB2 (R&D Systems) followed by HRP-conjugated anti-goat IgG (Chemicon). ALP positivity was detected as in the double staining mentioned above. For EphB4/TRACP, goat anti-mouse EphB4 (R&D Systems) was used as the primary antibody, with HRP-conjugated anti-goat IgG (Chemicon) as the secondary antibody. TRACP activity was detected as described previously.(24) All sections were faintly counterstained with methyl green.
BrdU/ALP double immunohistochemistry and quantification of double-positive cells
BrdU was detected with an immunohistochemistry kit (Calbiochem, Merck KGaA, Darmstadt, Germany) according to the manufacturer's protocol. Subsequently, ALP immunolocalization was performed as described above. Double-stained cells were counted with the aid of the ImagePro Plus 6.2 software (Media Cybernetics, Silver Spring, MD, USA), and the results are shown in cell number per tissue area (200 × 200-μm square portion of the metaphyseal region, 150 μm below the growth plate, excluding the cortical bone).
Quantification of the area occupied by ALP- and EphB4-positive cells
Images of sections single stained for ALP or for EphB4 (400 × 400-μm square portion of the metaphyseal region, 150 μm below the growth plate, excluding the cortical bone) were taken from PTH-injected and noninjected samples (n = 6/group). The area occupied by the positive cells was measured with ImagePro Plus 6.2 software (Media Cybernetics), and values are described in terms of percentage of area.
Characterization and quantification of preosteoblastic phenotypes under TEM
Ultrathin sections were obtained from the metaphyseal area (a longitudinal section, 1 mm below the most inferior point of the growth plate, 1.5 × 2 mm) and collected on 100-mesh cupper grids for TEM observation. Based on the works of Martineau-Doizé et al.(11) and Rouleau et al.,(12) pre-osteoblastic cells found between mature osteoblast/bone-lining cells and the bone marrow were reclassified into three subtypes. Ultrastructural features used for classification are summarized in Table 1. Results are expressed in number of cells per square millimeter of tissue area.
Immunoelectron microscopy for PTH-R and EphB4
Decalcified femoral sections (100 μm thick) were obtained as reported(27) and incubated with goat anti-mouse EphB4 (R&D Systems) or with rabbit anti-serum against PTHR(28) for 48 h at 4°C. These were followed by incubation with HRP-conjugated anti-goat IgG or anti-rabbit IgG (Chemicon), respectively, for 24 h at 4°C. After visualization with DAB, the sections were embedded in Epon as described above. Ultrathin sections were stained only with lead citrate before TEM observation (Hitachi H-7100).
Bone histomorphometrical parameters were quantified using ImagePro Plus 6.2 software (Media Cybernetics). For determination of structural parameters, H&E-stained paraffin sections were used. For kinetic parameters, 10-μm-thick sections embedded in glycol methacrylate (GMA) were observed under a fluorescent microscope (Nikon Eclipse E800). Images (400 × 400-μm square portion of the metaphyseal region, 150 μm below the growth plate and excluding the cortical bone) were obtained for all groups (n = 6/group). Abbreviations and calculations were done according to the recommendations of the ASBMR Histomorphometry Nomenclature Committee.(29)
Statistical analysis was performed using Microsoft Excel 2003 (Microsoft, Redmond, WA, USA), with differences among groups being assessed by unpaired Student's t-tests and considered statically significant at p < 0.05.
Intermittent PTH administration enhances pre-osteoblastic proliferation and bone formation in wildtype mice
Clearer calcein labeling was detected in PTH-treated mice (Figs. 1A and 1B), a finding consistent with the increases in structural and kinetic parameters presented in Table 2. PTH-treated specimens showed a thicker layer of pre-osteoblastic cells surrounding cuboidal osteoblasts, and these cells were ALP+ in both PTH-injected and noninjected groups (Figs. 1C and 1D). The percentage of ALP+ area was significantly higher in PTH-treated specimens (7.69 ± 5.84% versus 0.97 ± 0.6%, p < 0.05; Fig. 1I). Double staining for BrdU/ALP showed an increase of BrdU incorporation in ALP+ cells after PTH injection (Figs. 1E, 1F, and 1J; 6.67 ± 1.37 versus 3.97 ± 0.83, p < 0.05). In both groups, many ALP+ mature osteoblasts and pre-osteoblastic cells were shown to be also Runx2+ (Figs. 1G and 1H).
Pre-osteoblastic phenotype with fewer, dispersed ER (misER cell) is the most abundant after intermittent PTH
Ultrastructural observations of control wildtype bones showed spindle pre-osteoblastic cells between the bone marrow spaces and the mature osteoblasts lying on the bone surface (Fig. 2A). Three pre-osteoblastic phenotypes were identified in wildtype control bones after TEM observations: (1) an undifferentiated cell, which is mononuclear, has few organelles, and appears in proximity to osteoblasts or capillaries (not shown); (2) a cell with dispersed, relatively enlarged ER cisterns and many long cytoplasmic processes (Fig. 2B). Its Golgi apparatuses were scattered throughout the cytoplasm, and it bore several endosomes and lysosomes; and (3) a cell whose abundant ER cisterns were assembled in parallel and whose Golgi apparatuses were adjacent to the nuclei (Fig. 2C), as seen in the mature osteoblast. Given these ultrastructural characteristics and for the sake of convenience, we coined the terms misER cell (from the Latin miser, poor or reduced) to describe the phenotype with fewer and dispersed ER, and hypER cell (from the Greek hiper, abundant) for those whose abundant ER content was mostly arranged in parallel. Whereas hypER cells were seen adjacent to mature osteoblasts, misER cells and their long processes overlaid mature osteoblasts and hypER cells. Despite the morphological differences, the presence of the PTHR was confirmed in both phenotypes by means of immunoelectron microscopic examinations (Figs. 2D and 2E). In PTH-treated samples, an exuberant layer of misER cells and their processes populated the intertrabecular regions including bone marrow cells within their cellular network (Figs. 3A and 3B). Statistical analysis showed a significant increase in the number of misER cells (1377.03 ± 1020.47 versus 488.38 ± 650.08, p < 0.005) and mature osteoblasts (1697.56 ± 1567.42 versus 614.3 ± 859.7, p < 0.05) after PTH injection as shown in Fig. 3C. In contrast, the numbers of undifferentiated cells (394.44 ± 346.21 versus 646.06 ± 625.24, not significant) and hypER cells (814.73 ± 1066.11 versus 369.43 ± 397.55, not significant) were not significantly different between groups.
Proliferation of ALP+ cells is significantly increased but not accompanied by enhanced bone formation in PTH-injected c-fos −/− mice
There was no detectable calcein labeling within the areas of interest in control and in PTH-injected c-fos−/− specimens (Figs. 4A and 4B). The absence of labeling made the calculation of histomorphometrical kinetic parameters impossible in c-fos−/− specimens; as for structural parameters, differences between groups were unimportant (Table 2). However, PTH administration in c-fos−/− mice led to more intense expression of ALP (Figs. 4C and 4D) and to a significantly higher percentage of area occupied by ALP+ cells (8.12 ± 0.32% versus 1.83 ± 0.02%, p < 0.05; Fig. 4I). In addition, the number of cells positive for ALP and BrdU incorporation was significantly higher in the PTH injection group (Figs. 4E, 4F, and 4J; 3.54 ± 0.52 versus 2.13 ± 0.75, p < 0.01). As in wildtype mice, Runx2/ALP+ cells were detected in the bones of c-fos−/− mice regardless of the group (Figs. 4G and 4H).
misER cell is the most abundant pre-osteoblastic phenotype after intermittent PTH administration to c-fos−/− mice
Control c-fos−/− bone surfaces featured flattened cells with dispersed ER, many vesicles, and extended cytoplasmic processes representative of the misER phenotype. Some of the control c-fos−/− bone surfaces were not covered by cells; instead, blood vessels and cytoplasmic processes of misER cells were partially attached to the bone matrix (Figs. 5A and 5C). On the other hand, several misER cells were seen between blood vessels and bone surfaces in PTH-treated specimens (Figs. 5B and 5D). Interestingly, there were virtually no cells representative of either the hypER or the mature osteoblast phenotypes in the control and PTH-injected c-fos−/− specimens. Contrary to undifferentiated cells (252.47 ± 561.26 versus 291.67 ± 642.62), misER cell numbers were increased (2076.5 ± 1695.29 versus 1064.46 ± 1695.29), making these the most abundant pre-osteoblastic phenotype in PTH-injected c-fos−/− mice (Fig. 5E).
EphB4+ misER cells are seen in the proximity of osteoclasts after PTH administration
The relation between misER cells and hypER cells and osteoclasts was examined by means of immunohistochemistry for ephrinB2 and its receptor EphB4, because these molecules have been proposed as mediators of osteoblast–osteoclast interactions.(30) EphrinB2 was localized in osteoclasts and in osteoblasts of control and PTH-injected wildtype specimens (Figs. 6A and 6B), whereas EphB4 positivity was seen slightly in pre-osteoblasts (Figs. 6C and 6E). After PTH administration, several EphB4+ pre-osteoblasts were seen surrounding osteoclasts (Figs. 6D and 6F). Immunoelectron microscopy showed EphB4+misER cells in controls (Figs. 7A–7C), an immunoreactivity that grew more intense after PTH administration (Figs. 7D and 7E). According to light microscopy data, the percentage of area occupied by EphB4+ cells was significantly increased after treatment (7.21 ± 2.98% versus 2.13 ± 0.88%, p < 0.005; Fig. 7F). In contrast, mature osteoblasts did not show obvious expression of EphB4 in either control or PTH-treated specimens.
After submitting wildtype and c-fos−/− mice to intermittent PTH administration, our main findings were (1) ALP+ pre-osteoblastic cells were proliferating, and the misER cell was the most abundant pre-osteoblastic phenotype after PTH treatment regardless of the mice strain used; (2) despite significant increases in the number of misER cells after PTH injection, fully differentiated mature osteoblasts were absent in c-fos−/− mice, and consequently, bone anabolism was not seen in these mice; and (3) in wildtype mice, EphB4-positive misER cells were consistently seen in the proximity of osteoclasts, suggesting that these cells could be interacting with osteoclasts in the circumstance of PTH administration. Therefore, osteoclast presence may be necessary to support complete osteoblastic differentiation and subsequently promote the enhanced bone formation seen with intermittent PTH treatment.
The data obtained from our PTH-treated wildtype mice corroborates previous animal studies showing significant increases in structural and kinetic histomorphometrical parameters.(31–34) Whereas in our c-fos−/− mice, PTH-driven bone anabolism was not observed, such an absence of an anabolic response has been reported after intermittent PTH administration in growing mice.(22) Our regimen is therefore adequate for examining the effects of PTH treatment in cells of the osteoblastic lineage and the possible interactions of these cells with the bone-resorbing osteoclast.
ALP is an important bone metabolic marker, and increases in its levels are a common finding after intermittent PTH treatment.(35–37) ALP is also expressed in the very early stages of the osteoblastic lineage(38) and, agreeably, our PTH regimen has promoted significant increases in the percentage of area positive for ALP in wildtype and c-fos−/− mice. Osteoblastic cells in wildtype and c-fos−/− specimens were also shown to express Runx2, a transcriptional factor essential for osteoblastic differentiation.(39) Runx2 positivity was detected in cells representative of the misER phenotypes, as well as in hypER cells and mature osteoblasts. In addition, our immunostaining against ALP and BrdU was consistent with previous reports of increased proliferation of osteoprogenitor cells(40) and increased BrdU labeling in young rats treated with PTH for 3–5 days(41) and with the notion that PTH and PTHrP provoke an upregulation of cyclin D1 that may promote proliferation in the early stages of osteoblastic differentiation.(42) The increases in ALP expression and in BrdU/ALP double-positive cells point to enhanced proliferation of cells of the osteoblastic lineage, but such enhanced proliferation was not enough to promote PTH-driven bone anabolism in c-fos−/− bones.
The phenotypical variety of cells of the osteoblastic lineage(11–13) may provide a rationale for the paradox seen in the PTH-injected, osteoclast-less c-fos−/− mouse (i.e., increases in ALP expression and in osteoblastic cell proliferation unaccompanied by enhanced bone formation). In control wildtype mice, three major pre-osteoblastic phenotypes were identified: undifferentiated cells, misER cells, and hypER cells. The number of misER cells was significantly increased in PTH-treated wildtype and c-fos−/− mice, making them the most abundant pre-osteoblastic phenotype after intermittent PTH treatment. Of notice, misER and hypER cells consistently expressed the PTH receptor. It is therefore reasonable that ALP+misER cells in wildtype mice may respond to PTH essentially by proliferating and forming an extended cellular network with their long cytoplasmic processes. Because the difference in the number of hypER cells was not significantly increased after PTH treatment, but the number of mature osteoblasts was, we propose that misER cells might be pre-osteoblasts in an earlier differentiation stage and that hypER cells may be further into the differentiation ladder. The hypER cells may represent a rather transitional stage, readily differentiating into mature osteoblasts in the course of events after PTH administration.
Recently, ephrin signaling has been highlighted as crucial to the mutual communication between osteoclasts and osteoblasts. The work by Zhao et al.(30) reframed the discussion on osteoblast–osteoclast coupling by proposing that a bidirectional interaction between osteoclast-borne ephrinB ligands and EphB receptors expressed by osteoblasts may enhance the shift from resorption to formation that occurs at each resorption cycle. Such a shift would occur as osteoclastic formation is lessened by reverse ephrin signaling and osteoblastic differentiation is enhanced by forward ephrin signaling. Allan et al.(43) reported on PTH- and PTHrP-induced ephrinB2 mRNA increases in different osteoblastic cell lines and proposed that regulated ephrinB2 production might affect pre-osteoblasts in an autocrine or paracrine mode, causing them to differentiate and form bone. Also, the authors showed that EphB4 blockade significantly inhibited mineralization and the expression of genes usually seen late in the differentiation of osteoblastic Kusa 4b10 cells. Although differences in methodology between the two works do not allow for a clear comparison of the results, our experiments showed that misER cells (1) consistently expressed EphB4, (2) occupied a higher percentage of tissue area, and (3) were seen in the proximity of osteoclasts after PTH administration, findings that suggest an interaction at the pre-osteoblast–osteoclast level. Because both ephrinB2(43) and RANKL(16) are transiently expressed shortly after single PTH injection, we hypothesized that intermittent PTH might be able to sustain small but frequent cycles of osteoclast activation/formation while supporting the differentiation of proliferating pre-osteoblastic cells (such as the misER cell) through forward ephrin signaling, which would ultimately lead to bone anabolic responses.
Regardless of the limitations of this study, our results point to an apparent block in the differentiation cascade of c-fos−/− osteoblastic cells even after an intermittent PTH regimen. Such differentiation block may be a consequence of the nonexistence of osteoclasts and osteoclast precursors in these animals, reinforcing the notion that osteoblast–osteoclast interactions might be necessary for the pharmacological effects of intermittent PTH.
This work was partially supported by grants from the Japanese Society for the Promotion of Science and Naito Memorial Foundation (N.A., P.H.L.F.). We thank Dr. Hisashi Murayama (Kureha Special Laboratory) for invaluable assistance with our bone histomorphometry.
- 81956 The osteoblast. In: BourneGH (ed.) The Biochemistry and Physiology of Bone. Academic Press, New York, NY, USA, 179–212.
- 151964 Dynamics of bone remodelling. In: FrostHM (ed.) Bone biodynamics, 1st ed. Little, Brown & Co, Boston, MA, USA, 315–333.
- 332004 Skeletal unloading alleviates the anabolic action of intermittent PTH(1-34) in mouse tibia in association with inhibition of PTH-induced increase in c-fos mRNA in bone marrow cells. J Bone Miner Res 19: 1813–1820., , , , , ,