Osteopontin Deficiency Induces Parathyroid Hormone Enhancement of Cortical Bone Formation

Intermittent PTH treatment increases cancellous bone mass in osteoporosis patients; however, it reveals diverse effects on cortical bone mass. Underlying molecular mechanisms for anabolic PTH actions are largely unknown. Because PTH regulates expression of osteopontin (OPN) in osteoblasts, OPN could be one of the targets of PTH in bone. Therefore, we examined the role of OPN in the PTH actions in bone. Intermittent PTH treatment neither altered whole long-bone bone mineral density nor changed cortical bone mass in wild-type 129 mice, although it enhanced cancellous bone volume as reported previously. In contrast, OPN deficiency induced PTH enhancement of whole-bone bone mineral density as well as cortical bone mass. Strikingly, although PTH suppressed periosteal bone formation rate (BFR) and mineral apposition rate (MAR) in cortical bone in wild type, OPN deficiency induced PTH activation of periosteal BFR and MAR. In cancellous bone, OPN deficiency further enhanced PTH increase in BFR and MAR. Analysis on the cellular bases for these phenomena indicated that OPN deficiency augmented PTH enhancement in the increase in mineralized nodule formation in vitro. OPN deficiency did not alter the levels of PTH enhancement of the excretion of deoxypyridinoline in urine, the osteoclast number in vivo, and tartrate-resistant acid phosphatase-positive cell development in vitro. These observations indicated that OPN deficiency specifically induces PTH activation of periosteal bone formation in the cortical bone envelope.

TREATMENT OF OSTEOPOROSIS requires not only inhibitors for bone resorption but also stimulators for bone formation especially in patients who already have lost a significant amount of bone (1, 2). PTH is a major regulator of calcium homeostasis and plays an important role in both bone formation and resorption. The presence of excess amount of PTH in hyperparathyroidism or continuous administration of PTH results in bone loss. However, intermittent PTH administration leads to an increase in bone mass in patients with postmenopausal osteoporosis as well as in animal models of bone loss (3).

A recent large multicenter study revealed that intermittent PTH treatment increased bone mineral density (BMD) in vertebra, femur, and total body in postmenopausal osteoporosis patients, but it reduced BMD in cortical bone in the forearm (4). Furthermore, expression of constitutively active PTH/PTHrP receptor in transgenic mice increased cancellous bone mass, but it reduced cortical bone mass (5). These data are consistent with the previous observations that PTH signaling has diverse effects in different sites in bone, namely cancellous and cortical bone. However, the mechanisms that are involved in such diverse PTH actions in different bone envelopes are largely unknown.

Osteopontin (OPN) is a noncollagenous bone matrix protein and cytokine expressed in both osteoblasts and osteoclasts, and its expression is under the control of PTH (6, 7). However, the role of OPN in PTH regulation of bone metabolism in vivo is not known, although it is suggested that OPN inhibits mineral crystal growth in vivo (8) and in vitro (9). Therefore, we examined the effects of OPN on PTH actions in regulation of bone metabolism.

Materials and Methods

Animals

OPN-deficient mice were produced as described by Rittling et al. (10). The mice were back-crossed to 129 background and the progenies from original homozygous crosses were maintained as separate colonies. Forty-eight 7-wk-old female mice with either OPN-deficient or wild-type genotype were used. The mice were housed under controlled conditions at 24 C on a 12-h light/12-h dark cycle and fed with standard laboratory chow containing normal calcium and given tap water.

Experimental protocol

Recombinant human PTH (1–34) (Bachem, Torrance, CA) was dissolved in a vehicle of acidified saline containing 0.1% BSA (Sigma-Aldrich, St. Louis, MO). The mice were subjected to daily sc daily injections with PTH (1–34) at 80 μg/kg of body weight or vehicle (saline containing 0.1% BSA), for 5 d/wk for 4 wk. To evaluate mineralization fronts, the mice were injected sc with 4 mg/kg calcein 9 and 2 d, and 100 mg/kg xylenol-orange 4 d before being killed, respectively. Twenty-four hours after the last injection, the mice were anesthetized with Avertin (tribromoethanol) and killed.

Measurement of BMD

Bone mineral density of the whole femora and tibiae was measured based on a dual-energy x-ray absorptiometry (DEXA) using PIXI apparatus (GE Luner, Madison, WI). Ex vivo coefficients of variation were 0.5.

Two-dimenstional (2D) micro-x-ray-computed tomography (μCT) analysis of bone

The femora were subjected to 2D μCT analysis using Musashi (Nittetsu-ELEX, Osaka, Japan). For quantification of cortical bone, 2D images of the cross-section in the middiaphyses of the femora were obtained to determine the cortical bone mass. Cortical bone parameters such as cortical area, cortical thickness, and bone marrow area were analyzed. The image data were subsequently quantified using a Luzex-F automated image analysis system (Nireco, Tokyo, Japan). For cancellous bone, μCT slices were made within the midsagittal planes in the metaphyseal region of the bones to obtain bone volume/tissue volume (BV/TV) values in the 1.47 mm² square area (0.7 × 2.1 mm) located at 0.2 mm away from the growth plate of the distal end of femur.

Histomorphometric analysis of bone

For undecalcified section, the right femora were fixed in 70% ethanol, prestained with Villanueva osteochrome (bone stain), and embedded in methylmethacrylate. Cross-sections (serial 10-μm-thick sections) in middiaphysis and longitudinal sections (serial 3-μm-thick sections) were made using a microtome. For decalcified section, the left femora were fixed in 4% paraformaldehyde in PBS, decalcified in EDTA, embedded in paraffin, and sectioned. Undecalcified sections were used to examine cortical (first and second calcein) bone formation rate (BFR), mineralizing surface per bone surface (MS/BS), and mineral apposition rate (MAR) in periosteal and endosteal regions. Undecalcified sagittal sections were used to examine cancellous (xylenol orange and the second calcein) bone formation (BFR and MAR) in a square area of 1.4 mm², which was 0.2 mm away from to the growth plate. The histomorphometric analysis was carried out at the magnification of ×400.

Osteoclast number was quantified using decalcified 5-μm-thick sagittal sections in the metaphyses of femora. For decalcified section, the left femora were fixed in 4% paraformaldehyde in PBS, decalcified in EDTA, embedded in paraffin, and sectioned. The sections were stained for tartrate-resistant acid phosphatase (TRAP) activity. TRAP-positive multinucleated cells attached to bone were scored as osteoclasts. Measurements were made within a square area of 1.4 mm² (0.7 × 2.0 mm) located 0.2 mm away from the growth plate of the distal ends of femora to obtain osteoclast number per area (N.Oc), osteoclast number per bone surface (N.Oc/BS), and osteoclast surface per bone surface (Oc.S/BS) as defined by Parfitt et al. (11).

Bone marrow cultures

The proximal and distal epiphyseal ends were removed from long bones, and bone marrow was flushed out. The number of total bone marrow cells was counted, and the cells were plated in 12-well plates (3.8 cm² per well) at a density of 5 × 10⁶ cells/well. Osteoclastogenesis was examined in the cultures in which TRAP-positive osteoclast-like multinucleated cells were formed in α-MEM supplemented with 10% fetal bovine serum, 100 μg/ml antibiotics-antimycotics mixture, 10 nm 1,25(OH)₂ vitamin D₃, and 100 nm dexamethasone. The medium was changed every 3–4 d.

For mineralized nodule formation, bone marrow cells from the right tibiae were cultured in α-MEM supplemented with 10% fetal bovine serum, 100 μg/ml antibiotics-antimycotics mixture, 50 μg/ml ascorbic acid, and 10 mm sodium β-glycerophosphate. The medium was changed every 3–4 d. The cultures were rinsed with PBS and fixed in 95% EtOH for 10 min on d 21. The cultures were stained for 10 min in a saturated solution of alizarin red, rinsed with water, and dried in air. The area of mineralized nodules per total dish surface was measured by using the Luzex-F automated image analyzer (Nireco).

Statistical evaluations

The results were presented as mean values ± sd. Statistical analysis was conducted according to Mann-Whitney U test. P < 0.05 was considered to be statistically significant. Data from only one of the two bilateral bones (right or left) were used per each individual mouse.

Results

OPN deficiency induced PTH enhancement in whole-femur BMD

To examine the role of OPN in mediating the effect of intermittent administration of PTH on BMD of whole femur, the bones of mice were subjected to the analyses using a DEXA apparatus. BMD of the whole femur in wild-type mice was not increased by the intermittent treatment with PTH (Fig. 1A). In contrast, BMD of the whole femur was increased significantly by PTH treatment in OPN-deficient mice (Fig. 1A). Similarly, OPN deficiency also induced PTH enhancement in whole-tibia BMD levels (Fig. 1B).

OPN deficiency modestly enhanced PTH-dependent increase in the cancellous bone

To understand the underlying mechanism of the OPN-deficiency effects on the BMD response to PTH, measurements of the cancellous bone volume (BV/TV) in the metaphyseal regions of the femora in each group were conducted using 2D μCT. Intermittent PTH treatment enhanced the levels of cancellous bone in wild-type mice as reported previously (Fig. 2, A, B, and E). Although PTH treatment also increased cancellous bone levels in OPN-deficient mice (Fig. 2, C, D, and E), no major difference was observed between BV/TV values in PTH-treated wild-type and PTH-treated OPN-deficient mice (Fig. 2E). Thus, these data did not, at least fully, explain the OPN deficiency-induced PTH enhancement in the whole femur and tibia BMD.

OPN deficiency-induced PTH enhancement of the cortical bone mass

Beause intermittent PTH treatment enhanced whole-femur and tibia BMD only in OPN-deficient mice but not in wild-type mice without having a major difference in the PTH-dependent enhancement in the levels of cancellous bone volume, we further examined the cortical bone mass in the middiaphysis using cross-sections obtained by 2D μCT (Fig. 3, A–D). Examinations on the cross-sections (in a plane perpendicular to the long axis) in the middiaphyses of femora in wild-type mice indicated that PTH treatment did not alter the total cross-sectional area of the diaphysis of the femora (Fig. 3, A, B, and E). In contrast, PTH treatment increased the levels of total cross-sectional area in the diaphyses of the femoral midshafts in OPN-deficient mice (Fig. 3, C–E). Because cross-sectional area of femora includes both cortical bone and marrow areas, we examined these regions separately. PTH treatment did not alter the levels of cortical bone area in wild-type mice (Fig. 3F). In contrast, PTH treatment increased the levels of cortical bone area in OPN-deficient mice (Fig. 3F). Cortical thickness was estimated on the basis of the averaged values of thickness measured at the eight points at every 45° direction (Fig. 3G). PTH treatment did not alter cortical thickness in wild-type mice, but in OPN-deficient mice, cortical thickness was increased by PTH treatment (Fig. 3H). PTH treatment decreased marrow area in wild-type mice (Fig. 3I). However, PTH treatment did not significantly affect marrow area in OPN-deficient mice (Fig. 3I).

OPN deficiency not only blocked PTH-induced suppression of bone formation parameters but also induced PTH enhancement in periosteal BFR and MAR in cortical bone

Because OPN deficiency induced PTH enhancement of cortical bone mass, we further examined cellular mechanisms for these observations. In wild-type mice, PTH treatment suppressed periosteal BFR and periosteal MAR in the cortex of diaphyseal bone (Fig. 4, A–C). Mineralization surface was not significantly reduced (Fig. 4D). Strikingly, OPN deficiency not only blocked the suppressive effects of PTH on periosteal BFR and MAR but also induced PTH enhancement in periosteal BFR (by about 100%, P < 0.05), MAR (by about 20%, P < 0.05), and MS/BS (by about 50%, P < 0.05) (Fig. 4, B–D). Contrary to periosteum, PTH treatment enhanced BFR and MAR in the endosteum in wild-type, but OPN deficiency blocked such PTH effects in the endosteum (Fig. 4, E–H). Thus, these observations indicated that OPN deficiency blocked PTH suppression of bone formation and induced PTH enhancement of cortical bone formation specifically via the action through the periosteal region.

OPN deficiency also augmented the effect of PTH on bone formation in the cancellous bone

To further understand the role of OPN in PTH regulation of osteoblastic side in the gain in cancellous bone mass, we examined the effects of PTH treatment on bone formation in either wild-type or OPN-deficient mice (Fig. 5A). PTH treatment increased BFR as well as MAR in metaphyseal cancellous bone region in wild-type mice as reported previously (Fig. 5, B and C). In OPN-deficient mice, basal levels of BFR and MAR were similar to the basal levels in wild-type mice. However, OPN deficiency augmented PTH enhancement of MAR and BFR in the cancellous bone, compared with the PTH effects on these parameters observed in wild-type mice (Fig. 5, B and C).

OPN deficiency did not alter the PTH increase in osteoclast number per area

To examine the osteoclast side in terms of the cellular basis for the OPN deficiency effects on the PTH-induced bone gain in cancellous bone in vivo, TRAP-positive cells (osteoclasts) were examined in the metaphyseal regions (Fig. 6). PTH treatment increased N.Oc/BS in wild-type mice but not in OPN-deficient mice (Fig. 6, A–E). Similarly, Oc.S/BS was enhanced by the PTH treatment in wild-type but not in OPN-deficient mice (Fig. 6F). Thus, although the N.Oc/BS was similar because of OPN-deficiency enhancement of PTH increase in cancellous bone, OPN-deficiency did not alter the PTH-increase in total number of osteoclasts per area of bone.

OPN deficiency augmented PTH enhancement of in vitro bone nodule formation in the bone marrow cell cultures

To examine the effects of OPN deficiency on the osteogenic activity at the cellular levels in vitro, we conducted mineralized nodule formation assays in culture. Without PTH treatment in vivo, wild-type and OPN-deficiency bone marrow cell cultures yielded similar levels of nodule formation (Fig. 7, A and B). PTH treatment in vivo slightly but not significantly increased the number of calcified nodules developed in the cultures of bone marrow cells obtained from wild-type mice (Fig. 7, A and B). In contrast, OPN deficiency enhanced mineralized nodule formation in the cultures of bone marrow cells from the mice subjected to intermittent PTH treatment in vivo (Fig. 7, A and B).

With regard to osteoclast development in vitro, wild-type and OPN-deficient bone marrow cells gave rise to similar levels of TRAP-positive cell development in the absence of PTH treatment in vivo. PTH treatment in vivo enhanced TRAP-positive cell development in the marrow cell cultures in wild-type cells (Fig. 8, A–E). Contrary to the observations on the mineralized nodule formation in culture, OPN deficiency did not alter the TRAP-positive cell formation in the bone marrow cell cultures (Fig. 8E). These observations in vitro further indicated that the major target cells of OPN-deficiency modulation of PTH enhancement of bone cell activation are not those in osteoclast lineage but those in osteoblast lineage.

Discussion

Our data demonstrated that OPN deficiency induces PTH enhancement of cortical bone volume. This OPN-deficiency effect on cortical bone volume accounted for the overall BMD increase in the whole femora and tibiae. In cortical bone, OPN deficiency reversed the suppressive action of PTH on periosteal bone formation rate and mineral apposition rate and induced enhancing actions of PTH on the two bone formation parameters. In addition, OPN deficiency still maintained and even further enhanced PTH activation of cancellous bone formation. Thus, OPN deficiency enhanced intermittent PTH treatment-induced bone gain mostly via periosteal osteoblastic activation in cortical bone envelope.

In vitro experiments also indicated that mineralized nodule formation was not significantly induced in the wild-type bone marrow cells obtained from the mice subjected to intermittent PTH treatment. In contrast, OPN-deficient cells demonstrated PTH enhancement in nodule formation. These observations further indicated that the targets of OPN-deficiency effects on PTH actions are the cells in osteoblastic lineage.

Contrary to the observations regarding the effects of OPN deficiency on the PTH actions in the cells in osteoblastic lineage, OPN deficiency did not alter the levels of PTH enhancement in deoxypyridinoline excretion in urine used as a marker of systemic osteoclastic bone resorption, the levels of osteoclast number per area, and osteoclast development in culture, compared with those in wild type. Although OPN deficiency did not allow PTH-induced increase in N.Oc/BS, this likely was due to the increase in the cancellous bone. These data indicate that OPN deficiency enhances intermittent PTH treatment-induced bone gain mostly, if not fully, through the enhancement of PTH increase in osteoblastic activity, especially in the periosteal of the cortex rather than its influence on the PTH increase in osteoclastic activity.

It has been known that osteoblasts expressed OPN mRNA. However, the roles of OPN in bone formation have not yet been understood. Overexpression of α-v β-3 integrin in osteoblasts negatively regulated bone nodule mineralization and osteoblast differentiation in vitro (12). Furthermore, it is suggested that OPN-deficient mice reveal hypermineralization in bone with increased mineral crystallinity and crystal size (8). Thus, OPN may play a role as an inhibitor of enhanced bone formation or crystal growth in vivo as well as in vitro (9). Because negative signals for bone formation via α-v β-3 integrin in osteoblasts could be reduced by OPN deficiency, it is possible that bone formation and mineralization may tend to be increased in OPN-deficient mice. However, the similar levels of BFR, MAR, and MS/BS untreated wild and OPN-deficient mice indicated that OPN specifically functions when bone formation is under the challenge by intermittent administration of PTH. Alternatively, other molecules such as bone sialoprotein may compensate OPN deficiency in the mice without particular challenge against bone metabolism. Because intermittent PTH treatment enhanced OPN gene expression, OPN could act as a negative feedback system for the PTH stimulation, acting as a break against the acceleration of bone formation by PTH.

PTH enhances differentially the cancellous bone mass levels, depending on the treatment period, species, or strains. In mice, intermittent treatment with PTH selectively increased cancellous bone mass in long bones and/or spine in a several strains (13–16). However, PTH does not increase the mass of whole bone and body BMD in any of the mouse strains (13, 14). As mentioned, in humans, a large multicenter study on osteoporosis patients revealed PTH reduction in cortical bone BMD (4). In rats, intermittent PTH treatment has been reported to increase both cortical and cancellous bone mass in some experiments (17–25); however, the reason for the species difference between rat vs. mouse and human is not known. Our data indicated that OPN deficiency reversed the suppressive PTH effects on periosteal in cortical bone BFR, MAR, and MS/BS and induced PTH activation of the levels of periosteal BFR and MAR, which lead to the increase in cortical bone mass in the pure 129-mouse strain. Our data showing PTH suppression of periosteal BFR with no change in cortical area conflict with data from rabbits and monkeys, which show increased periosteal BFR and greater bone (rabbits) and cortical (rabbits and monkeys) areas. The control data in our study are not consistent with what would be found in an adult animal with intracortical remodeling, more similar to the human condition. It is thus possible our mice model has a limitation in this regard. Whether our observations in our mice with respect to OPN-deficiency effects on PTH actions could be extrapolated into human and other species or strains still requires further elucidation.

OPN deficiency induces PTH enhancement in the gain of cortical bone mass via increases in periosteal BFR without any adverse effects on the previously known PTH actions to increase cancellous bone. Moreover, OPN deficiency further potentiated PTH activation of bone formation in the cancellous bone. The only small adverse effect of OPN deficiency could be its suppression of PTH activation of endosteal BFR, MAR, and MS/BS. However, this did not affect overall PTH increase in the cortical bone mass. In many cases, endosteal bone behaves like metaphyseal cancellous bone. Why this was not the case in OPN-deficient mice is still to be elucidated. Although strain background is different, Calvi et al. (5) reported that transgenic mice expressing constitutively active PTH receptor in bone resulted in increase in cancellous bone and decrease in cortical bone levels. These data indicated that OPN deficiency modulates beneficially the effect of PTH treatment on bone formation in cortical as well as cancellous bone. PTH is associated with addition of bone to both endocortical and periosteal surfaces of cortical bone, and the apparent loss of bone intracortically is transient and a function only of the increased turnover, the latter occurring also in the cancellous bone compartment. So there is certain similarity in cortical and cancellous bone, and the differences that do exist are explainable at the tissue level through an understanding of basic multicellular unit level bone remodeling. As mentioned, OPN deficiency does not affect the intermittent PTH treatment-induced bone resorption parameters in vivo as well as in vitro. Thus, OPN plays a critical inhibitory role in PTH treatment-induced bone formation.

Intermittent PTH treatment has been shown to reduce cortical bone mass in the patients with osteoporosis. Development of measures to inhibit the OPN function could lead to more effective PTH treatment for severe osteoporosis with regard to prevention of PTH-induced bone loss and introduction of bone gain by PTH in cortical bone in addition to the further enhancement of PTH-induced cancellous bone gain.

Acknowledgments

This work was supported by the grants-in-aid received from the Japanese Ministry of Education, Sports and Science (14207056, 14034214, 14028022, 12557123, 13045011, and 13216034), grants from National Space Development Agency of Japan, Japan Society for Promotion of Science (Research for the Future Program, Genome Science), and Tokyo Biochemistry Research Foundation.

Abbreviations

 
• BFR

  Bone formation rate;

 
• BMD

  bone mineral density;

 
• BV/TV

  bone volume/tissue volume;

 
• μCT

  two-dimensional micro-x-ray-computed tomography;

 
• DEXA

  dual-energy x-ray absorptiometry;

 
• MAR

  mineral apposition rate;

 
• MS/BS

  mineralizing surface per bone surface;

 
• N.Oc

  osteoclast number per area;

 
• N.Oc/BS

  osteoclast number per bone surface;

 
• Oc.S/BS

  osteoclast surface per bone surface;

 
• OPN

  osteopontin;

 
• TRAP

  tartrate-resistant acid phosphatase;

 
• 2D

  two-dimensional.