The experimental work characterizing the anabolic effect of parathyroid hormone (PTH) in bone has been performed in nonmurine ovariectomized (OVX) animals, mainly rats. A major drawback of these animal models is their inaccessibility to genetic manipulations such as gene knockout and overexpression. Therefore, this study on PTH anabolic activity was carried out in OVX mice that can be manipulated genetically in future studies. Adult Swiss-Webster mice were OVX, and after the fifth postoperative week were treated intermittently with human PTH(1–34) [hPTH(1–34)] or vehicle for 4 weeks. Femoral bones were evaluated by microcomputed tomography (μCT) followed by histomorphometry. A tight correlation was observed between trabecular density (BV/TV) determinations made by both methods. The BV/TV showed >60% loss in the distal metaphysis in 5-week and 9-week post-OVX, non-PTH-treated animals. PTH induced a ∼35% recovery of this loss and a ∼40% reversal of the associated decreases in trabecular number (Tb.N) and connectivity. PTH also caused a shift from single to double calcein-labeled trabecular surfaces, a significant enhancement in the mineralizing perimeter and a respective 2- and 3-fold stimulation of the mineral appositional rate (MAR) and bone formation rate (BFR). Diaphyseal endosteal cortical MAR and thickness also were increased with a high correlation between these parameters. These data show that OVX osteoporotic mice respond to PTH by increased osteoblast activity and the consequent restoration of trabecular network. The Swiss-Webster mouse model will be useful in future studies investigating molecular mechanisms involved in the pathogenesis and treatment of osteoporosis, including the mechanisms of action of known and future bone antiresorptive and anabolic agents.
OSTEOPOROSIS IS characterized by an imbalance between bone formation and resorption resulting in a net bone (particularly cancellous) loss and increased fracture incidence.1-5) Currently, several antiresorptive agents are approved by the Food and Drug Administration (FDA) for preventing or treating postmenopausal osteoporosis: estrogen (or hormone replacement therapy), selective estrogen receptor modulators (SERMs), calcitonin, and bisphosphonates. These therapies slow bone turnover, generate small increments in bone mass, and reduce fracture incidence.(6) However, a substantial number of fractures still occur in patients taking these medications and in cases of senile osteoporosis for which the efficacy of these drugs is low.7-9) Therefore, bone anabolic agents that restore bone mass, architecture, and mechanical properties would be highly beneficial for these patients.
A large number of studies in experimental animals and humans have indicated that intermittent administration of low doses of parathyroid hormone (PTH) by daily subcutaneous injection effectively stimulates cancellous and often cortical bone formation and reverses the bone loss induced by estrogen deficiency.(6) The initial concern regarding the PTH-induced enhancement in cancellous bone mass at the expense of cortical bone10-12) has not been substantiated by more recent studies.13-15)
Most of the in vivo studies designed to elucidate the mechanism of the PTH skeletal anabolic effect in estrogen-deficient animal models have been examined in femoral and tibial metaphyseal cancellous and diaphyseal cortical bone of ovariectomized (OVX) rats. The OVX rat model also is recommended by the FDA for feasibility studies assessing the antiosteoporotic efficacy of antiresorptive and anabolic agents.(16) These studies have shown that intermittently administered PTH or PTH analogues enhance cancellous bone mass by stimulating the number and activity of trabecular osteoblasts.(17, 18) However, the molecular mechanisms responsible for these effects have not been established, one reason being the inaccessibility of the rat (and other animals) model to potential genetic manipulations such as gene knockout or overexpression.
Different mouse strains, which are readily accessible to such manipulations, exhibit a high variability in their bone mineral density (BMD)(19) as well as cancellous bone content and distribution.(20) Nevertheless, a substantial amount of estrogen-sensitive metaphyseal secondary spongiosa has been shown in the proximal tibial and distal femoral metaphyses of Swiss-Webster mice (Fig. 1).(21, 22) Using this mouse strain, the present correlative microcomputerized tomographic (μCT) and histomorphometric study shows for the first time that intermittently administered human PTH(1-34) [hPTH(1-34)] reverses OVX-induced metaphyseal cancellous bone loss in mice. These findings establish the osteoporotic Swiss-Webster mouse as a model with potential usefulness for the study of molecular mechanisms involved in the pathogenesis and treatment of osteoporosis.
MATERIALS AND METHODS
Sixty-three 11-week-old virgin female Swiss-Webster mice were purchased from Taconic Farms (Germantown, NY, USA) and maintained at the animal research facility at the Beth Israel Deaconess Medical Center. Animals were fed Purina Formulab Diet containing 1% calcium (Purina Mills, Richmond, IN, USA) and water ad libitum throughout the experiment. Mice were killed by CO2 inhalation.
The experimental protocol was approved by the Institutional Animal Care and Use Committee of the Beth Israel Deaconess Medical Center. The study design is shown schematically in Fig. 2. The mice were subdivided randomly into four groups of bilaterally OVX animals and three groups of sham OVX animals, with nine animals in each group. A group of each of OVX and sham animals was killed 1 week postoperatively (T1/OVX and T1/sham, respectively) and served as baseline controls. Four additional weeks were allowed to pass before initiation of treatment in the remaining groups to permit significant bone loss to occur in the OVX animals. At this time, one group of OVX (T5/OVX) and one group of sham OVX mice (T5/sham) were killed to evaluate pretreatment bone loss.
A 4-week daily treatment (5 days a week) consisting of subcutaneous (sc) injections of either 80 μg/kg per day of hPTH(1-34) (Advanced ChemTech, Louisville, KY, USA) or vehicle (VEH) only (saline containing 0.001N HCl and 2% heat-inactivated mouse serum) was then administered to the respective T9/OVX/PTH and T9/OVX/VEH groups. At the end of the treatment period, the animals were killed together with a 9-week postoperative sham OVX group (T9/sham). To label mineralization fronts, all T9 groups were given sc injections of calcein (15 mg/kg) in 2% sodium bicarbonate solution 3 days and 1 day before death. At death, the femoral bones were separated, cleaned of soft tissues, fixed in 10% phosphate-buffered formalin (pH 7.2) for 48 h, and then kept in 70% ethanol until further use.
For a detailed qualitative and quantitative three-dimensional (3D) evaluation, whole femoral bones were examined by a desktop μCT system (μCT 20; Scanco Medical AG, Bassersdorf, Switzerland) equipped with a 10-μm focal spot microfocus X-ray tube as a source.(23) For image acquisition, the specimen was mounted on a turntable shifted automatically in an axial direction over 216° (180° plus half fan angle on either side), taking 600 projections. To scan the entire femoral width (3.4-5.1 mm), including the femoral head, a total of 100-150 μCT slices were acquired at a 34-μm slice increment. CT images were reconstructed in 512 × 512 pixel matrices using a standard convolution-backprojection procedure with a Shepp and Logan filter. Images were stored in 3D arrays with an isotropic voxel size of 34 μm. A constrained 3D Gaussian filter was used to suppress partly the noise in the volumes and the mineralized tissue was segmented from soft tissues by a global thresholding procedure.(24) Morphometric parameters were determined using a direct 3D approach(25) in three different preselected analysis regions: whole bone (including the articular ends), secondary spongiosa in the distal metaphysis, and diaphyseal cortical bone. For the whole bone, we formulated a new parameter, the apparent volume density (AVD), which is the percent mineralized tissue volume over the total volume defined by the external bone envelope. Parameters determined in the metaphyseal trabecular bone included bone volume density (BV/TV), bone surface density (BS/TV), trabecular number (Tb.N), trabecular thickness (Tb.Th), trabecular spacing (Tb.Sp), and connectivity density. Percent cortical volume (%BV), percent marrow volume (%MV), and cortical thickness (Ct.Th) were assessed in a 1-mm-thick slab in the middiaphysis.
After μCT image acquisition, the specimens were dehydrated in progressive concentrations of ethanol, cleared in xylene, and embedded in methylmethacrylate. Longitudinal 5-μM sections from the center of each bone were deplasticized and left unstained for dynamic measurements. To identify osteoclasts, additional sections were deplasticized and stained with modified Masson-Goldner Trichrome with Bierbrich Scarlet (Sigma, St. Louis, MO, USA). Histomorphometric analysis was performed using the BioQuant computerized image analysis system (R & M Biometrics, Nashville, TN, USA). The measurements, terminology, and units used for the histomorphometric analysis were according to the convention of standardized nomenclature.(26)
All analyses were carried out using SigmaStat software (SPSS Science, Chicago, IL, USA). The significance of differences in quantitative μCT and histomorphometric parameters between time/treatment groups was determined based on analysis of variance (ANOVA). When significant differences were indicated by ANOVA, group means were compared using the Tukey test for pairwise comparisons. In addition, correlation coefficients were calculated between certain μCT and histomorphometric parameters.
Effect of PTH on metaphysial distal femoral cancellous bone
μCT and histomorphometric analyses of femoral cancellous and cortical bone showed little change over the duration of this study in any of the structural variables in sham OVX Swiss-Webster mice (groups T1/sham, T5/sham, and T9/sham; Figs. 3, 4, and 5). μCT measurements in OVX mice exhibited significant decrease in BV/TV of 76% and 85% 5 weeks and 9 weeks postoperatively, respectively (groups T5/OVX and T9/OVX/VEH; Fig. 3A). This decrease was accompanied by substantial disruption of the cancellous architecture (Fig. 4). In mice treated for 4 weeks with PTH (T9/OVX/PTH) the trabecular BV/TV was 293% higher than VEH-treated animals (T9/OVX/VEH) and 187% over T5/OVX mice (Fig. 3A). The histomorphometric BV/TV determination tightly correlated with the μCT measurements (r = 0.887; p < 0.001) and showed a 62% loss in T9/OVX/VEH compared with T9/sham. Similar to the μCT measurements, histomorphometry showed that the trabecular BV/TV in the T9/OVX/PTH mice was 183% higher than in the T9/OVX/VEH group (Fig. 3B). In addition, the PTH treatment partially restored the trabecular architecture (Fig. 4).
The μCT analysis showed a respective 53% and 62% reduction in Tb.N in groups T5/OVX and T9/OVX/VEH (Fig. 3E) as well as a 28% and 35% decrease in the thickness (Tb.Th) of the remaining trabeculae (Fig. 3C). As expected, these deficits, which apparently led to the cancellous bone loss, were associated with a significant corresponding increase in Tb.Sp (Fig. 3D) and lower connectivity (Fig. 3F). All of these parameters were significantly reversed by the PTH treatment. Tb.N, Tb.Th, and connectivity in the T9/OVX/PTH showed a respective 35, 45, and 30% reversal compared with T5/OVX and T9/sham as well as 164, 132, and 419% increase over T9/OVX/VEH. Tb.Sp in group T9/OVX/PTH was 32% and 45% smaller than in groups T5/OVX and T9/OVX/VEH, respectively (Fig. 3D).
The OVX-induced bone loss was associated with 38% and 61% increases in osteoclast number (N.Oc/BS), quantitated histomorphometrically, and observed 1 week and 5 weeks post-OVX, respectively (groups T1/OVX and T5/OVX). The N.Oc/BS declined to sham control levels 9 weeks postoperatively regardless of the PTH treatment (groups T9/OVX/VEH and T9/OVX/PTH; Fig. 5). Trabecular osteoblast and osteoclast surfaces (Oc.Ss) are shown in Fig. 6. In agreement with the osteoclast counts shown in Fig. 5, changes in the Oc.S also were insignificant among T9/sham, T9/OVX/VEH, and T9/OVX/PTH. However, osteoblast surface was significantly higher in T9/OVX/PTH than in both T9/sham and T9/OVX/VEH. Moreover, the PTH-treated mice exhibited stacking of plump osteoblasts alongside trabecular surfaces compared with the single sheath of flat osteoblasts observed in the T9/sham and T9/OVX/VEH non-PTH-treated animals (Fig. 6 inset), an observation that is consistent with the static μCT and histomorphometric data presented previously.
The increased trabecular osteoblast surface in the PTH-treated mice as well as the osteoblast morphology in these animals are consistent with increased osteoblastic activity as reflected by the calcein label-based histomorphometric measurements. Although mineral apposition rate (MAR), the mineralizing perimeter, and the bone formation rate (BFR/BS) were similar in groups T9/sham and T9/OVX/VEH, they showed a respective 219, 149, and 295% increase in the T9/OVX/PTH over the T9/OVX/VEH mice (Figs. 7A-3C). In addition, group T9/OVX/PTH showed a significant decrease in single-labeled surfaces (SLS; from 73% in T9/sham to 37% in T9/OVX/PTH) paralleled by an increase in double-labeled surface (DLS; from 3% in T9/sham animals to 37% in T9/OVX/PTH; Fig. 7D). Likewise, fluorescent photomicrographs revealed abundance of double labels with increased interlabel spacing (Fig. 7E). These fluorescent images also are consistent with increased osteoblastic activity. The perimeter of most osteocyte lacunae in the T9/OVX/PTH mice showed a single calcein label (Fig. 8). Such lacunae could not be found in the non-PTH-treated mice.
Effect of PTH on diaphyseal cortical bone
Middiaphyseal measurements using μCT and histomorphometry indicated that the 4-week PTH treatment increased Ct.Th, reduced the medullary cavity volume, and increased cortical MAR in Swiss-Webster mice. Ct.Th significantly increased by ∼11% in T9/OVX/PTH mice over T5/OVX and T9/OVX/VEH animals with a corresponding ∼9% decrease in the %MV (Fig. 9). As expected, the cortical endosteal MAR also was increased in group T9/OVX/PTH, ∼3-fold over both T9/sham and T9/OVX/VEH (Fig. 9). The decrease in %MV and high correlation between Ct.Th and the endosteal MAR (r = 0.713) imply that the increased Ct.Th resulted from an endosteal bone-forming activity.
Effect of PTH on whole femoral bone volume
μCT analysis of whole femoral bones showed a significant OVX-induced decrease in AVD in OVX mice. Most of this decrease was reversed by the intermittent PTH administration. Losses of 24% and 30% were recorded in T5/OVX and T9/OVX/VEH, respectively. The difference in AVD between T9/OVX/PTH and T9/sham was statistically insignificant (Fig. 10).
In this study, we show for the first time the reversal of bone loss in estrogen-depleted mice treated intermittently with PTH. Unlike rats, in which the presence of metaphyseal trabecular bone as well as OVX-induced bone loss are strain independent, different mouse strains vary substantially in their bone mineral and trabecular bone content and skeletal responses to estrogen deficiency.(19, 20, 27-30) Apparently, this strain-related diversity has largely excluded the mouse from general use as a model for testing the efficacy of exogenously administered antiosteoporotic agents. Therefore, we selected the Swiss-Webster strain based on the presence of a substantial amount of trabecular bone in its femoral metaphyses and a quantifiable loss of this bone after OVX (Fig. 1).
As in this study, previous reports indicated that OVX Swiss-Webster mice rapidly lose metaphyseal cancellous bone within 4-5 weeks. This bone loss was associated with increased osteoclastic resorption as suggested by a decrease in [3H]-tetracyclin retention in bone.(21) Likewise, the present bone loss is associated with increased N.Oc's already 1 week post-OVX (T1/OVX). This increase accompanies the initial 5-week, steep bone loss (T5/OVX). The N.Oc declines to sham-OVX control levels at 9 weeks postoperatively, when further bone loss is minimal, and is unaffected by the intermittent PTH treatment (T9/sham, T9/OVX/VEH, and T9/OVX/PTH; Fig. 5). This trend of changes in N.Oc and bone loss is reminiscent of results obtained in post-OVX follow-up in rats.(31, 32) However, it is unclear whether, as in the case of rats, the increase in N.Oc is accompanied by high bone turnover inasmuch as previous studies in 4-week post-OVX Swiss-Webster mice(21) and the present results in the 9-week OVX, VEH-treated group (T9/OVX/VEH) exhibit only mild, statistically insignificant increases in DLS, mineralization perimeter, MAR, and BFR. This issue remains to be elucidated by methods of a lower signal-to-noise ratio than histomorphometry, such as the new osteocalcin and C-telopeptide determinations in mouse serum.(31, 32)
The ∼35% PTH-induced recovery of BV/TV in the OVX Swiss-Webster mice is somewhat lower compared with OVX rats treated with PTH for a similar time period (∼46% reversal).(30) However, OVX mice appear to lose trabecular bone more rapidly than rats (Fig. 3).33-37) Therefore, the anabolic effects of PTH relative to the remaining spongiosa may be considered comparable in both species. Although the present increase in BV/TV in mice was paralleled by increases in Tb.Th and Tb.N, both of a similar magnitude (∼40% reversal; Fig. 3), the rescue of BV/TV in the rat is mainly attributable to increased Tb.Th (∼100% reversal).(30) It remains to be determined whether this difference is species or methodology related (histomorphometry vs. μCT). The increases in Tb.Th and Tb.N suggest that the anabolic effect of PTH is brought about by both thickening of existing and formation of new trabeculae. Judging by the comparable Oc.S/BS in T9/sham and T9/OVX/PTH, the PTH treatment did not affect bone resorption (Fig. 6). The increases in Tb.N and Tb.Th and the resulting amplification of the trabecular density all appear to occur consequent to large increases in osteoblast number and activity. As in rats,(38) after 4 weeks of treatment, the Swiss-Webster mice showed a respective ∼2- and ∼3-fold increase in MAR and BFR/BS (Fig. 8). In the PTH-treated mice, this increase correlates well with the stacking of plump osteoblasts at trabecular surfaces as compared with the single layer of rather flat osteoblasts observed in the sham- and vehicle-treated animals (Fig. 6).
The anabolic effect of PTH not only increased all osteoblastic parameters, but also resulted in uptake of the calcein label into the perimeter of most trabecular osteocytic lacunae. Normally, mineralization in osteocytic lacunae can be detected only by confocal microscopy.(39) Although absent in non-PTH-treated animals, the frequent identification of calcein-labeled lacunae in the PTH-treated mice by conventional fluorescent microscopy suggests enhanced mineralization activity. Because osteocytes express receptors for PTH,(40) the presence of labeled lacunae in the center of trabeculae (Fig. 7) is likely the result of a direct effect of PTH on osteocytes. Because PTH has a key role in osteocytic mechanosignaling,(41) the PTH-induced stimulation of osteocyte activity may be of significance in the restoration of bone biomechanical properties in estrogen-depleted subjects.
The anabolic effect of PTH in the Swiss-Webster mice was not restricted to the metaphyseal cancellous bone. As in a few rat studies,(42, 43) we observed a significant increase in the middiaphyseal Ct.Th after treatment with PTH. This increase resulted mainly from endosteal osteoblastic activity inasmuch as it was accompanied by a corresponding decrease in the relative medullary cavity volume (%MV) and stimulation of endosteal MAR (∼3-fold increase; Fig. 9). In addition, the complete reversal of AVD in these animals suggests that other mineralized femoral components such as the proximal metaphysis and epiphyseal ends also were stimulated by the PTH treatment. The AVD measurement by μCT provides a nondiscriminatory BMD assessment similar to BMD determination by dual-energy X-ray absorptiometry (DXA).(44) Importantly, BMD in PTH-treated animals of other species was shown to correlate positively with improved biomechanical properties.43, 45-47)
Previous studies in species other than mice have shown a close correlation between μCT and histomorphometric measurements of structural metrics.48-50) Although μCT morphometry was used previously for the structural analysis of mouse bone,(51, 52) the present tight correlation between μCT and histomorphometric BV/TV determinations, as well as the high correlation between cortical MAR and μCT Ct.Th confirm its applicability to the study of mouse osteoporotic models. The use of quantitative μCT in mice provides an advantageous alternative to DXA in instances in which high precision is required. In addition, unlike DXA, μCT can readily discriminate the different bone compartments (trabecular vs. cortical).(51, 53) However, DXA is still useful for studying live animals and in cases when speed and cost of the testing are of concern.
So far, the OVX rat has been the “gold standard” for evaluating the antiosteoporotic activity of anabolic and antiresorptive agents. In addition, knockout and transgenic mice have been used extensively for the study of genes involved in the pathogenesis of and potential therapies for osteoporosis.54-58) The OVX Swiss-Webster mouse offers the opportunity to introduce genetic modifications into a valid animal model for osteoporosis, thus enabling study of the molecular mechanisms involved in the effect of PTH and other anabolic and antiresorptive agents on the osteoporotic skeleton. Therefore, in future studies investigators should be able to combine molecular and metabolic manipulations in the same animal model.
This work was supported in part by the National Institutes of Health (NIH) grant R01 DK47940 and a collaborative research agreement between Beth Israel Deaconess Medical Center and Millennium Pharmaceuticals, Inc.