Mechanisms for the Enhancement of Fracture Healing in Rats Treated With Intermittent Low-Dose Human Parathyroid Hormone (1–34)

Authors


  • The authors have no conflict of interest.

Abstract

Recent reports have demonstrated that intermittent treatment with parathyroid hormone (1–34) [PTH(1–34)] increases callus formation and mechanical strength in experimental fracture healing. However, little is known about the optimal dose required for enhancement of fracture repair or the molecular mechanisms by which PTH regulates the healing process. In this study, we analyzed the underlying molecular mechanisms by which PTH affects fracture healing and tested the hypothesis that intermittent low-dose treatment with human PTH(1–34) can increase callus formation and mechanical strength. Unilateral femoral fractures were produced and a daily subcutaneous injection of 10 μg/kg of PTH(1–34) was administered during the entire healing period. Control animals were injected with vehicle solution alone. The results showed that on day 28 and day 42 after fracture, bone mineral content (BMC), bone mineral density (BMD), and ultimate load to failure of the calluses were significantly increased in the PTH-treated group compared with controls (day 28, 61, 46, and 32%; day 42, 119, 74, and 55%, respectively). The number of proliferating cell nuclear antigen (PCNA)-positive subperiosteal osteoprogenitor cells was significantly increased in the calluses of the PTH-treated group on day 2, and TRAP+ multinucleated cells were significantly increased in areas of callus cancellous bone on day 7. The levels of expression of type I collagen (COL1A1), osteonectin (ON), ALP, and osteocalcin (OC) mRNA were increased markedly in the PTH-treated group and accompanied by enhanced expression of insulin-like growth factor (IGF)-I mRNA during the early stages of healing (days 4–7). The increased expression of COL1A1, ON, ALP, and OC mRNA continued during the later stages of healing (days 14–21) despite a lack of up-regulation of IGF-I mRNA. These results suggest that treatment of fractures with intermittent low dose PTH(1–34) enhances callus formation by the early stimulation of proliferation and differentiation of osteoprogenitor cells, increases production of bone matrix proteins, and enhances osteoclastogenesis during the phase of callus remodeling. The resultant effect to increase callus mechanical strength supports the concept that clinical investigations on the ability of injectable low-dose PTH(1–34) to enhance fracture healing are indicated.

INTRODUCTION

PARATHYROID HORMONE (PTH) has varying effects on bone depending on the mode of administration.(1, 2) Continuous infusion of PTH decreases bone mass by stimulating a net increase in bone resorption, whereas intermittent administration increases bone mass by stimulating osteoblast differentiation and increases bone formation. In experimental models of osteoporosis in which bone loss was induced by ovariectomy or orchidectomy, intermittent treatment with PTH promoted recovery of bone mass.(3–6)

Recent reports have shown that intermittent treatment with PTH(1–34) increases callus formation and mechanical strength in experimental fracture healing.(7, 8) In a rat model, Andreassen et al. reported that a daily 200 μg/kg dose of PTH(1–34) was more effective in enhancing callus formation and mechanical strength than a dose of 60 μg/kg.(7) This dose is considered within the dose range normally used in rat experiments, and 200 μg/kg per day is the highest dose shown to induce new bone with normal morphology on surfaces of intact bone.(9, 10) Moreover, in a recent clinical study on the use of PTH(1–34) in the treatment of osteoporosis, patients showed increased bone mass after a mean of 21 months when administered doses as low as 20–40 μg/day.(11)

Fracture repair is a specialized form of wound healing in which bone is regenerated rapidly. Because most fractures achieve sufficient healing to support mechanical loads after 2–3 months, the dose of PTH(1–34) needed to enhance fracture healing may differ from that required for the treatment of osteoporosis. To date, however, no studies regarding the optimal dose required for an enhancement of fracture repair have been reported. In addition, few data are available regarding the molecular mechanisms by which intermittent PTH administration enhances callus formation, despite a number of animal studies indicating that this treatment increases both callus formation and mechanical strength. In this study, we hypothesized that intermittent low-dose treatment with human PTH(1–34) can increase callus formation, resulting in an increase in mechanical strength and bone mass of the callus. To test this hypothesis, we used a standardized rat model of fracture healing and analyzed mechanical strength and bone mass of the callus. Furthermore, we investigated the molecular mechanisms by which PTH treatment may enhance the healing process.

MATERIALS AND METHODS

Animals and hormone administration

Two-month-old male Sprague-Dawley rats were divided into two groups. The experimental group (n = 72) was treated with PTH(1–34) at a dose of 10 μg/kg per day and the control group (n = 64) received vehicle solution (saline). Treatments were administered daily by subcutaneous injection commencing on the day fractures were produced until the time of death. On days 2, 4, 7, 14, and 21 postoperatively, eight rats of each group (control rats, n = 4; PTH-treated rats, n = 4) were killed and the samples that were harvested were used for molecular and histological studies. On day 28 and day 42, 10 rats of the each group were killed, and the fractured and contralateral intact femurs were removed for the analyses of mechanical testing and DXA. At the same time points, blood was also collected for analyzing serum biochemistry and bone metabolic markers.

To determine a minimum effective dose of PTH for the enhancement of healing in rat femoral shaft fractures, we performed preliminary experiments examining histomorphology of day-28 calluses at doses of 2, 10, 50, and 100 μg/kg per day and found that callus formation developed in a dose-dependent manner. Because previous studies have already documented the enhancement of fracture repair in rats treated with doses ranging from 60 to 200 μg/kg per day,(7, 8) we then performed preliminary experiments examining biomechanical strength of the calluses in rats administered either 2 μg or 10 μg of PTH. To the best of our knowledge, there are no prior reports showing the enhancement of fracture healing in rats treated with such low doses of PTH. The results indicated that in comparison with rats treated with vehicle solution alone, administration of 10 μg of PTH significantly increased the mechanical strength of fracture calluses, whereas 2 μg of PTH did not have a significant effect. Thus, we chose a dose of 10 μg/kg per day as a minimum effective dose for this study.

Fracture model

A standard closed middiaphyseal fracture was produced in the right femur of each rat using a device and method developed in our laboratory.(12) Briefly, rats were anesthetized with sodium pentobarbital and a median parapatellar incision was made at the knee. Next, a Kirschner wire (1.1 mm in diameter) was introduced into the medullary canal of the right femur. After closure of the incision at the knee joint, a middiaphyseal fracture was created with an apparatus composed of a blunt guillotine driven by a dropped weight. The rats were permitted full weight bearing and unrestricted activity after recovery from anesthesia. During the experiment, all animals were maintained in cages with free access to food and water. These experimental procedures were approved by the Animal Care and Use Committee of Chiba University.

Mechanical testing

Ultimate load to failure of the healing fracture calluses was measured on day 28 and day 42 postoperatively by a three-point bending procedure using a materials testing system (MZ-500D; Maruto, Inc., Tokyo, Japan). The intramedullary nail was removed and the fractured femur was placed on two rounded bars with the fracture line positioned between the bars. The contralateral intact femur was tested using the same procedure at a point corresponding to the fracture line of the right femur.

DXA

All specimens of fractured and intact femurs subjected to mechanical testing showed failure without loss of bone substance. This permitted reassembly such that the two ends of the specimens were realigned in their original position and fixed with a quick-drying glue. They then were stored in 70% ethanol at room temperature for later measurement of bone mineral content (BMC) by DXA. At the time of BMC measurement, the femur was placed in Ringer's solution with the posterior surface facing downward and scanned by DXA using the regional high-resolution analysis program for small animals (QDR-1000; Hologic, Inc., Waltham, MA, USA). BMC was measured in a 10-mm-high diaphyseal segment (5 mm proximal and 5 mm distal to the fracture line). The BMC of the corresponding diaphyseal segment in the opposite femur was also measured. Bone mineral density (BMD) was calculated as BMC/two-dimensional femur surface.

Tissue preparation

The animals were killed by intracardiac infusion of 4% paraformaldehyde under sodium pentobarbital anesthesia on days 2, 4, 7, 14, and 21 after the production of fractures. The fractured and intact femurs were removed with the surrounding soft tissues and fixed with 4% paraformaldehyde and 0.1 M of PBS (pH 7.4) at 4°C for 24 h. The tissues were decalcified at room temperature with 20% EDTA and 0.05 M of Tris-HCl (pH 7.4), bisected sagittally in the median plane, and embedded in paraffin. Six-micrometer midsagittal sections were mounted on silane-coated slides.

Analysis of cell growth activity in the periosteal callus

Sections were immunostained with a monoclonal antibody against proliferating cell nuclear antigen (PCNA; DAKO, Glostrup, Denmark) to evaluate cell growth activity. Immunohistochemical staining was performed as previously described.(13) Four regions (0.9 mm × 1.5 mm in size) containing subperiosteal osteoblastic cells (also including osteoprogenitor cells) were analyzed in each section. The number of PCNA+ cells was counted on days 2, 4, 7, and 14 after fracture, and the ratio of PCNA+ cells to total cells was calculated and expressed as a percentage. The average measurement obtained was used as the PCNA score.(14, 15)

Analysis of osteoclast number in areas of cancellous bone in the callus

The number of osteoclasts (N.Oc) on days 7, 14, and 21 after fracture was counted in demineralized longitudinal histological sections of the fracture site. Eight regions (0.9 mm × 1.5 mm in size) containing cancellous bone were analyzed in each section. The osteoclast index (N.Oc/callus perimeter in mm) was calculated and the average measurement was determined.(16) TRAP staining was carried out according to the Burstone's azo dye method to identify osteoclasts.(17)

Preparation of probes

The following cDNA clones were used as hybridization probes: mouse type 1 collagen (COL1A1) cDNA containing a 0.25-kb fragment, rat osteonectin (ON) cDNA containing a 0.55-kb fragment, and a rat osteocalcin (OC) cDNA containing a 0.13-kb fragment. The specificity of these probes has been confirmed previously.(18, 19) A full-length cDNA of rat ALP (2.4 kb) was kindly provided by Dr. G.A. Rodan (Merck & Co. Sharp and Dohme Research Lab., West Point, PA, USA).(20) An insulin-like growth factor (IGF)-I cDNA containing a 0.38-kb fragment corresponding to rat IGF-I nucleotides 43–426 (Gene Bank D00698) was generated by reverse-transcription-polymerase chain reaction (RT-PCR) using total RNA from normal rat femurs. The cDNA fragment was inserted into pGEM-T easy vector (Promega, Madison, WI, USA) by T/A cloning, and the sequence was verified by sequencing. The specificity of the IGF-I probe was confirmed by in situ hybridization (ISH) using sections containing normal rat femoral growth plates and the signals were properly localized as previously described.(21)

RNA extraction and Northern blot analysis

Rats were killed after anesthesia with sodium pentobarbital on days 2, 4, 7, 14, and 21 postoperatively, with both the intact and the fractured femurs being harvested for RNA extraction. The tissues were frozen immediately in liquid nitrogen and stored at −80°C until RNA isolation was performed. Total cellular RNA was extracted using TRIzol (Gibco BRL, Rockville, MD, USA) according to the manufacturer's instructions. Twenty micrograms of total RNA from each sample was subjected to 1% agarose gel electrophoresis and transferred to a nylon membrane (Hybond-XL; Amersham Pharmacia Biotech, Buckinghamshire, UK). cDNA probes were labeled with32P using a random priming method. Northern blot analysis was carried out as previously described.(15) The density of each band on the autoradiogram was estimated by an image analyzer (Image Gauge ver. 3.1 software; FUJIFILM, Tokyo, Japan).

ISH

Cells expressing IGF-I mRNA in the fracture calluses were identified by means of ISH. Digoxigenin (DIG)-11-uridine triphosphate (UTP)-labeled single-strand RNA probes (antisense and sense probe) for rat IGF-I cDNA were prepared. ISH was carried out as previously described.(18,22,23) The sections were hybridized with the antisense probe at 50°C for 16 h and the signals were detected using the DIG detection kit (Roche Molecular Biochemicals, Indianapolis, IN, USA). After signal detection, the sections were counterstained with methyl green. The sense probe was used to ensure specificity.

Analysis of serum biochemistry and bone metabolic markers

The concentrations of total calcium, phosphorus, and ALP in the serum were analyzed on 28 days and 42 days of treatment. In addition, serum OC and type I collagen C-terminal telopeptide (ICTP) were measured by radioimmunoassay (RIA). Serum TRAP was measured by visible absorption spectrophotometry.

Statistical analysis

Differences between groups were determined by an ANOVA. Where differences existed, the Fisher protected least significant difference test was used to determine significance. A value of p < 0.05 was considered statistically significant.

RESULTS

Radiological analysis

In the control group, periosteal callus formation was hardly detectable on day 7 after fracture (Fig. 1A), whereas it was detected clearly in the PTH-treated group at the same time point (Fig. 1B, arrowheads). By day 14, bony callus was clearly visible in both groups (Figs. 1C and 1D) and was enhanced in the PTH-treated group (Fig. 1D, arrowheads). Osseous bridging over the fracture site was visible in both groups by day 28 (Figs. 1E and 1F). On day 28 and day 42, the callus in the control group showed obvious radiolucency (Figs. 1E and 1G, arrowheads), and the PTH-treated group exhibited continued enhancement of bony callus formation (Figs. 1F and 1H, arrowheads); remodeling of the callus occurred over a period of up to 42 days.

Figure FIG. 1..

Radiographic findings of (A, C, E, and G) the control group and (B, D, F, and H) the PTH-treated group. In the PTH-treated group, enhancement of bony callus formation is clearly detected (B, arrowheads) on day 7 and (D, arrowheads) is more evident on day 14. On day 28 and day 42, (E and G, arrowheads) calluses in the control group show obvious radiolucency, (F and H, arrowheads) whereas the PTH-treated group exhibits continued enhancement of bony callus formation. Arrows indicate the fracture site.

Biomechanical analysis

The results of mechanical testing are shown in Table 1. On 28 days and 42 days of healing, the ultimate load to failure of the calluses in the PTH-treated group was increased by 61% and 119%, respectively, compared with the control group. The ultimate load to failure of the contralateral intact femurs in the PTH-treated group also increased slightly on day 28 and day 42 (11% and 13%, respectively).

Table Table 1.. Mechanical Strength, BMC, and BMD of Fractured and Contralateral Intact Femur
original image

DXA

BMC and BMD of the fractured and contralateral intact femurs on 28 days and 42 days of healing are shown in Table 1. On day 28, the BMC and BMD of the calluses in the PTH-treated group was increased by 46% and 32%, respectively, compared with the control group. On day 42, the increase of BMC and BMD in the PTH-treated group was 74% and 55%, respectively. However, analysis of the contralateral intact femurs revealed no significant difference between groups on either day 28 or day 42.

Histomorphological analysis

The method for producing fractures used in this study produces very little soft tissue damage and leads to a reproducible, organized healing process. On day 7 after fracture, a comparable degree of trabecular bone development was observed in the region between the periosteum and the cortical bone (hard callus) in both groups. At the same time point, abundant cartilaginous tissue was observed adjacent to the newly formed trabecular bone (Figs. 2A and 2B). Beginning on approximately day 14, this cartilaginous tissue was replaced by woven trabecular bone in a process consistent with normal endochondral ossification. This was followed by bone remodeling, which produced mature trabecular bone. At this time, most of the mature trabeculae in the hard callus of the control group were replaced by bone marrow (Fig. 2C, arrowheads). In contrast, the PTH-treated group exhibited more mature trabeculae in the hard callus (Fig. 2D, arrows). On day 21, the bone marrow became more visible in the hard callus in the control group (Fig. 2E, arrowheads) whereas further development of mature trabeculae was evident in the PTH-treated group (Fig. 2F, arrows). There were no apparent morphological differences in chondrogenesis between groups.

Figure FIG. 2..

Histological features of fracture healing of (A, C, and E) the control and (B, D, and F) the PTH-treated groups. (A and B) On day 7, a comparable degree of trabecular bone development is observed in both groups. (C and E, arrowheads) On day 14 and day 21, calluses in the control group show active bone resorption with early development of marrow cavities, (D and F, arrows) whereas the PTH-treated group exhibits developed mature trabeculae. Asterisks show the fracture site (bars = 500 μm).

Quantification of PCNA+ cells in the periosteal callus

In both groups, thickening of the periosteum near the fracture site occurred concurrently with the proliferation of subperiosteal cells on day 2 after fracture (Figs. 3B and 3C). In the control group, 28% of subperiosteal osteoprogenitor cells were PCNA+ and in the PTH-treated group, the PCNA score for the osteoprogenitor cells was significantly increased to 39% at this time point (p = 0.003; Fig. 3A). As fracture healing proceeded, the PCNA scores gradually declined in both groups with no significant differences being evident on days 4, 7, and 14 (Fig. 3A).

Figure FIG. 3..

(A) Quantification of PCNA+ cells in the periosteal callus and (B and C) distribution of PCNA+ cells during intramembranous ossification. (A) In the PTH-treated group, significant increase is detected in the PCNA score on day 2 but not on days 4, 7, and 14. Data are expressed as mean ± SEM (*significantly different from control, p = 0.003). (B and C) On day 2, a considerable amount of subperiosteal osteoprogenitor cells are positive for PCNA in the (C) PTH-treated group compared with the (B) control group. PCNA+ cells are shown by brown staining in the nuclei. po, Periosteum; cb, cortical bone (bars = 30 μm).

Quantification of N.Oc in the hard callus

In the both groups, mature trabeculae became visible by day 7. At this stage, the control group exhibited TRAP+ multinucleated cells scattered among the newly formed trabeculae (Fig. 4B, arrowheads) with the maximal increase in the osteoclast index evident on day 14 (Fig. 4A). In the PTH-treated group, a number of TRAP+ multinucleated cells were observed among the trabeculae on day 7 (Fig. 4C, arrowheads) with the maximal increase in the osteoclast index also being found at this time point. After day 7, the osteoclast index was decreased (Fig. 4A). A significant difference was seen in the osteoclast index between groups on day 7 (p = 0.003) but not on day 14 and day 21 (Fig. 4A).

Figure FIG. 4..

(A) Quantification of N.Oc in the hard callus and (B and C) distribution of TRAP+ multinucleated cells in areas of cancellous bone. (A) In the PTH-treated group, significant increase is detected in the osteoclast index on day 7 but not on day 14 and day 21. Data are expressed as mean ± SEM (*significantly different from control, p = 0.004). (B and C) On day 7, a number of TRAP+ multinucleated cells are observed among the mature trabeculae in the (C, arrowheads) PTH-treated group compared with the (B, arrowheads) control group. tb, Trabecular bone (bars = 30 μm).

Northern blot analysis

mRNA expression for COL1A1, ON, ALP, OC, and IGF-I in fracture calluses was examined at the same time points in both groups (Fig. 5) and the levels of mRNA expression were quantified (Fig. 6). In the control group, the expression of COL1A1 was clearly detected on day 2, reached a maximum between days 4 and 7, and gradually decreased thereafter. The maximum increase in the expression of COL1A1 in the PTH-treated group was observed on day 7 when the expression level was 2.1-fold greater than the control group. Compared with controls, the expression in the PTH-treated group of COL1A1 was also increased by 2.0- and 2.2-fold on day 14 and day 21, respectively. Expression of ON, one of the main bone matrix proteins produced by relatively immature osteoblasts in addition to COL1A1, reached a peak expression between days 4 and 14 in the control group. In the PTH-treated group, ON expression levels were comparable with the control group on day 2 and day 4 but increased by 1.7-fold on day 7. Subsequently, enhanced ON expression was observed on day 14 and day 21. ALP was expressed at a slightly higher level in the PTH-treated group than in the control group on day 2 and day 4, with the expression being enhanced further on day 7 (2.0-fold). ALP expression gradually decreased after day 7 in the control group, but it remained elevated until day 21 in the PTH-treated group. OC, a bone matrix protein produced specifically by mature osteoblasts, was first detected in both groups on day 7, and the expression of this mRNA increased until day 21. On day 14 and day 21, OC expression was enhanced in the PTH-treated group compared with controls (1.5-fold at each time point). IGF-I, a growth factor involved in mediating the anabolic effects of PTH on bone formation, was highly expressed between days 4 and 7 in both groups. In the PTH-treated group, IGF-I expression was increased compared with the control group by 2.1- and 1.3-fold on day 4 and day 7, respectively. Enhanced expression of IGF-I mRNA was not observed in the PTH-treated group between days 14 and 21, despite the enhanced expression of COL1A1, ON, ALP, and OC.

Figure FIG. 5..

Northern blot analysis for fracture calluses of the control and the PTH-treated groups. Twenty micrograms of total RNA from each daily sample was analyzed for quantifying the levels of mRNA expression for COL1A1, ON, ALP, OC, and IGF-I. GAPDH was used as an internal standard for the amount and integrity of RNA preparation. Representative autoradiographic images are shown.

Figure FIG. 6..

Quantification of the levels of mRNA expression for COL1A1, ON, ALP, OC, and IGF-I in fracture calluses of the control and the PTH-treated groups. Four calluses of the each group were analyzed at each time point. Each band density (Fig. 5) was normalized to the ratio of the internal standard GAPDH. Data are expressed as mean ± SEM. a, Significantly different from control, p < 0.05; b, significantly different from control, p < 0.01; c, significantly different from control, p < 0.001.

In contrast to the dynamic changes in mRNA expression in fracture calluses, the contralateral intact femoral shafts in both groups exhibited a constant level of mRNA expression for COL1A1, ON, ALP, and OC throughout the healing process (data not shown). No differences were evident in the expression levels of these mRNAs between groups at any time point. IGF-I mRNA was not detected in the contralateral intact femoral shafts (data not shown).

ISH

On day 4 after fracture, the expression of IGF-I mRNA was widely detectable in various types of cells including osteoblasts, chondrocytes, mesenchymal cells, and cells among muscle layers near the fracture site in both groups. However, the signals were relatively weak and no differences were evident between groups (data not shown). Similarly, the expression was widely detectable in various cell types in both groups on day 7 (Figs. 7A and 7B). In the hard callus, subperiosteal osteoprogenitor cells (Fig. 7C, arrowheads) and osteoblasts lining the mature trabeculae (Fig. 7C, arrows) of the control group were positive for IGF-I mRNA, with a more intense signal evident in osteoprogenitor cells (Fig. 7D, arrowheads) and osteoblasts lining the trabeculae (Fig. 7D, arrows) of the PTH-treated group. As fracture healing proceeded, the signals of IGF-I mRNA became weaker, with no differences detectable between groups. A faint signal for IGF-I mRNA was seen in hypertrophic chondrocytes in the cartilaginous tissue and mature osteoblasts lining the trabeculae in the hard callus on day 14 and day 21 in both groups (data not shown). No signal was detected when using the sense probe.

Figure FIG. 7..

Expression of IGF-I mRNA in the control and the PTH-treated groups on day 7 after fracture. (A and B) Lower magnification views of IGF-I signal in the (A) control and the (B) PTH-treated groups. Expression of IGF-I mRNA is widely detectable in various types of cells during intramembranous and endochondral ossifications. The areas in the boxes are enlarged in panels C and D, respectively. (C and D) Expression of IGF-I mRNA is detected in the subperiosteal osteoprogenitor cells (arrowheads) and osteoblastic cells lining the trabeculae (arrows). The signals are stronger in the (D) PTH-treated group than in the (C) control group. Asterisks show the fracture site. po, Periosteum; tb, trabecular bone (in panels A and B, bars = 100 μm; in panels C and D, bars = 30 μm).

Serum biochemistry and bone metabolic markers

The serum biochemical results are shown in Table 2. After 28 days of treatment, serum OC was significantly higher in the PTH-treated group compared with controls (p < 0.001). After 42 days of treatment, it trended toward being higher in the PTH-treated group but the difference was not significant. Serum calcium was significantly higher in the PTH-treated group compared with controls on day 42 (p < 0.005). No significant differences were detected between groups in the levels of serum phosphate, ALP, ICTP, and TRAP on day 28 or day 42.

Table Table 2.. Serum Biochemistry and Bone Metabolic Markers
original image

DISCUSSION

Previous in vivo studies have reported that the anabolic effects of PTH on bone formation are mediated by IGF-I,(4, 24) and in vitro studies have shown that IGF-I stimulates proliferation and differentiation of osteoblasts and increases collagen synthesis.(1, 2) In addition, functional antibody blocking of IGF-I prevents PTH-stimulated collagen synthesis and ALP activity,(1, 2) and IGF-I knockout mice fail to show increased bone formation in response to PTH.(24) These reports suggest that IGF-I is essential for the anabolic effects of PTH on bone formation. Based on these findings, we focused on the expression of IGF-I during fracture healing because modulation of IGF-I expression may be an important molecular mechanism by which intermittent treatment with low-dose PTH(1–34) might enhance fracture repair.

To evaluate the effects of PTH on the proliferation of subperiosteal osteoblastic cells, we compared the PCNA score between the control and PTH-treated groups. The PCNA score of the subperiosteal osteoprogenitor cells in the PTH-treated group was significantly higher than that of the control group. Previous in vitro studies indicate that PTH treatment stimulates [3H] thymidine incorporation into DNA in bone cultures with anti-IGF-I antibody having no inhibitory effect.(1) This would suggest that PTH stimulation of cell proliferation is independent of IGF-I. These data are supported by our in vivo results because the PCNA score in the PTH-treated group was significantly increased on day 2 in the absence of any significant up-regulation of IGF-I mRNA expression. IGF-I mRNA was up-regulated between days 4 and 7 in the PTH-treated group, but no significant increase in the PCNA score was found at these time points. Therefore, PTH stimulation of the proliferation of subperiosteal osteoprogenitor cells in vivo appears to be independent of IGF-I.

Osteoclastgenesis and osteogenesis are critical coordinated events during fracture healing because bone remodeling in the hard callus is essential in order that the callus achieves mechanical strength sufficient to support loads imposed during normal weight bearing. Therefore, we analyzed the effects of PTH on osteoclastogenesis in the hard callus using a TRAP assay. The results showed that, on day 7, the osteoclast index in the hard callus in the PTH-treated group was significantly higher than that in controls. Recently, we studied osteoclastogenesis in normal tibial fracture repair in mice and found that osteoprotegerin (OPG) mRNA reached a peak in expression on day 7 whereas RANKL mRNA was decreased at this time point.(25) Based on this finding, it is possible that in the early stages of healing, PTH promotes osteoclastogenesis in the hard callus by regulating the expression of OPG and/or RANKL. Further investigations are required to determine the effect of PTH on the expression of OPG and RANKL during fracture healing and the molecular mechanisms by which PTH stimulates osteoclastogenesis in the hard callus.

Northern blot analysis indicated that PTH markedly enhanced the mRNA expression of bone matrix proteins (COL1A1, ON, and OC) and ALP after day 7 postfracture. These findings, in conjunction with those described previously, suggest that PTH enhances both callus formation and remodeling by stimulating the proliferation of osteoprogenitor cells, synthesis of bone matrix proteins, and osteoclastogenesis. Previous reports have shown that the anabolic effects of PTH on bone formation are mediated by IGF-I.(4, 24) Therefore, we investigated whether the up-regulation of mRNAs for bone matrix proteins and ALP in the PTH-treated group was accompanied by the increased expression of IGF-I mRNA. The results show that PTH increased the expression of IGF-I mRNA in the early stages of healing (days 4–7) but not in later stages (days 14–21). These findings were also reproduced in the ISH results. We interpret these findings to suggest that in the early stages of healing, PTH-induced IGF-I plays an important role in the enhancement of osteoblast differentiation and bone matrix synthesis. Presently, the mechanisms by which PTH induces IGF-I in bone have not been fully established, although a recent in vitro study identified a cyclic adenosine monophosphate (cAMP) response element (CRE) within the IGF-I P1 promoter in osteoblasts.(26) Thus, we speculate that in the early stages of healing, PTH stimulation may induce an intracellular increase in cAMP in osteoblastic cells in the callus, which may regulate directly transcriptional activation of IGF-I.

These results show that PTH administration increases the mechanical strength, BMC, and BMD of the calluses in the later stages of healing (days 28–42). Our unpublished data (The 19th Annual Meeting of the Japanese Society for Bone and Mineral Research, 2001) show that such an increase in mechanical strength and bone mass is maintained until day 56 in the calluses of the PTH-treated rats. Because in this model of fracture healing, callus remodeling is almost completed by day 56, we suggest that intermittent administration of PTH(1–34) at a dose of 10 μg/kg per day also enhances fracture healing in the remodeling phase.

In contrast to its observed effects in the early stages of healing, PTH did not increase the expression of IGF-I mRNA in the later stages despite the enhanced expression for bone matrix proteins and ALP, suggesting that PTH stimulation of osteoblast activity and bone matrix synthesis was IGF-I independent. We propose a possibility that may explain these findings: it is well documented that one of the major PTH signaling pathways is protein kinase A (PKA) and PKA activation leads to phosphorylation of the cAMP-response-element-binding protein (CREB), which then induces transcription factors such as AP-1 members, fos, and jun.(27) In addition, it has been reported that AP-1 consensus binding sequences are present in the promoter regions of osteoblastic genes such as COL1A1, OC, and ALP.(28) Therefore, we speculate that in the later stages of fracture healing, PTH maintains the elevated levels of expression of bone matrix proteins and ALP by directly promoting transcriptional activation of these genes.

In addition to the analysis of the fractured femur, we also analyzed the contralateral intact femur of each animal to determine whether the low-dose (10 μg/kg per day) treatment of PTH(1–34) increases mechanical strength and bone mass of normal femurs. The results indicate that PTH did not affect either BMC or BMD in the intact femurs on either day 28 or day 42, although it slightly increased the mechanical strength. Consistent with these data, Northern blot analysis showed that PTH treatment did not modulate the levels of mRNA expression for COL1A1, ON, ALP, OC, and IGF-I throughout the 21 days of treatment. This suggests that the dose of 10 μg of PTH(1–34) was not sufficient to enhance the bone formation of intact femurs over this relatively short time exposure. In contrast, the increase in mechanical strength and bone mass of fractured femurs observed in this study indicates that this low dose was sufficient for the enhancement of callus formation. Thus, PTH may have a more profound effect on the skeleton during the process of bone repair than it does during normal homeostasis.

To determine the influence of the low-dose treatment of PTH on systemic bone metabolism, we analyzed specific serum biochemical and bone markers after 28 days and 42 days of treatment. Contrary to the absence of any significant difference in the bone mass of the contralateral intact femurs between groups, serum OC in the PTH-treated group was significantly increased on day 28 compared with the control group. On day 42, it still was higher in the PTH-treated group although a significant difference was not found. Because a number of studies show the anabolic effects of intermittent PTH administration on cancellous bone formation,(3–6,9,29,30) we speculate that the target cells for PTH are mature osteoblasts lining the trabeculae in the metaphyses of long bones and vertebrae. Therefore, we suggest that the PTH-induced increase in the serum OC without an increase in the mRNA expression for bone matrix proteins in the contralateral intact femurs may be explained by the lack of mature osteoblasts in the diaphyses of long bones.

In summary, intermittent low-dose treatment with human PTH(1–34) increases both the mechanical strength and the bone mass of the calluses during fracture healing, effects that are potentially explained by the molecular mechanisms elucidated in this study. Although previous reports have shown enhanced fracture healing using relatively high doses of PTH,(7, 8) the findings of this study show that doses in the range that may be tolerated in humans are also effective. Thus, intermittent low-dose treatment with human PTH(1–34) may represent an effective strategy for the enhancement of fracture healing and could become the first systemic intervention for the repair of skeletal injuries.

Acknowledgements

We are grateful to Drs. M. Tahara and S. Sano for their technical support and to Drs. F. Nakajima, K. Goto, and A. Ogasawara (Department of Orthopedic Surgery, Chiba University Graduate School of Medicine) for helpful discussion. This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture of Japan.

Ancillary