Anabolic and Catabolic Regimens of Human Parathyroid Hormone 1–34 Elicit Bone- and Envelope-Specific Attenuation of Skeletal Effects in Sost-Deficient Mice

Abstract

PTH is a potent calcium-regulating factor that has skeletal anabolic effects when administered intermittently or catabolic effects when maintained at consistently high levels. Bone cells express PTH receptors, but the cellular responses to PTH in bone are incompletely understood. Wnt signaling has recently been implicated in the osteo-anabolic response to the hormone. Specifically, the Sost gene, a major antagonist of Wnt signaling, is down-regulated by PTH exposure. We investigated this mechanism by treating Sost-deficient mice and their wild-type littermates with anabolic and catabolic regimens of PTH and measuring the skeletal responses. Male Sost^+/+ and Sost^−/− mice were injected daily with human PTH 1–34 (0, 30, or 90 μg/kg) for 6 wk. Female Sost^+/+ and Sost^−/− mice were continuously infused with vehicle or high-dose PTH (40 μg/kg · d) for 3 wk. Dual energy x-ray absorptiometry-derived measures of intermittent PTH (iPTH)-induced bone gain were impaired in Sost^−/− mice. Further probing revealed normal or enhanced iPTH-induced cortical bone formation rates but concomitant increases in cortical porosity among Sost^−/− mice. Distal femur trabecular bone was highly responsive to iPTH in Sost^−/− mice. Continuous PTH (cPTH) infusion resulted in equal bone loss in Sost^+/+ and Sost^−/− mice as measured by dual energy x-ray absorptiometry. However, distal femur trabecular bone, but not lumbar spine trabecular bone, was spared the bone-wasting effects of cPTH in Sost^−/− mice. These results suggest that changes in Sost expression are not required for iPTH-induced anabolism. iPTH-induced resorption of cortical bone might be overstimulated in Sost-deficient environments. Furthermore, Sost deletion protects some trabecular compartments, but not cortical compartments, from bone loss induced by high-dose PTH infusion.

Bone mass and geometry are major risk factors contributing to fracture susceptibility. Newer approaches to osteoporosis are focused on harnessing anabolic mechanisms to stimulate new bone production. The first generation of these bone anabolic compounds that gained Food and Drug Administration approval is a truncated form of human PTH—or teriparatide. When administered intermittently, 20 μg/d teriparatide can increase spine bone mineral density (BMD) by approximately 10% after 18–24 months of treatment (1, 2). Higher doses are similarly efficacious after only 3 months of treatment (3). Moreover, 18–24 months of teriparatide treatment can reduce the incidence of new vertebral fractures by 65% among postmenopausal patients (4).

Despite its well-documented therapeutic value in improving bone mass and fracture susceptibility, the mechanisms of action for intermittent PTH (iPTH) treatment remain only partially understood. PTH increases osteoblast proliferation (5–7), differentiation/dedifferentiation (8, 9), and lifespan (10, 11). In addition to its proosteoblastic effects, PTH also stimulates osteoclasts to increase bone resorption. Continuously high levels of PTH, like iPTH treatment, stimulate both resorption and formation (12). Unlike iPTH treatment, however, osteoclast activity outpaces osteoblast activity, and net bone loss results. This phenomenon is most readily seen in patients with moderate to severe primary hyperparathyroidism, where parathyroid tumors cause constant, excessive secretion of PTH into the circulation.

At the cell surface, the mechanism of PTH action in bone cells is mediated by the PTH/PTHrP type-1 receptor (PTH1R), a Gα_s- and Gα_q-linked seven-pass transmembrane receptor. The Gα_s arm of receptor activation is important in initiating the anabolic action of PTH, and a number of crucial downstream nodes in the cascade, including direct targets (e.g. protein kinase A, Creb, Crem, Atf4, Runx2) and paracrine/autocrine loops (e.g. Igf1, Fgf2) have been identified (7, 13–17). One set of downstream targets activated by PTH, which are particularly relevant to bone metabolism, are those that participate in, or are transcriptionally activated by, the Wnt-signaling pathway. PTH signaling in bone alters the expression of numerous Wnt pathway genes, including Lrp5, Lrp6, Fzd1, Dkk1, Sost, Wisp1, Wif1, and β-catenin, among others (12, 18–20).

Whereas downstream nodes of the canonical Wnt pathway itself, and canonical Wnt-responsive transcriptional mechanisms are altered by PTH treatment, it is less clear whether (and to what extent) the Wnt pathway is required for the anabolic action of PTH in bone. For example, we (21) and others (22) have found that the Wnt co-receptor Lrp5 is not required for iPTH effects, as indicated by a full anabolic response to iPTH in Lrp5^−/− mice. Moreover, the cytoplasmic tail of Lrp5 does not undergo phosphorylation in response to PTH treatment, whereas that of Lrp6 does (23). Whether Lrp6 is required for PTH effects in bone is less clear, partly because Lrp6^−/− mice die at birth (24) and thus cannot be treated with iPTH. However, other studies that focus on secreted inhibitors of Lrp5/6 can shed light on the role of Lrp6 in PTH signaling, because their effects on Lrp5 can be largely discounted in iPTH signaling (see above). Secreted Wnt signaling inhibitors that antagonize the interaction, either directly or indirectly, between Wnt ligands and the Lrp5/6 receptors include Dkk1, sclerostin, and sFrp1, among others. Sost is of particular interest in potentially mediating PTH signaling for several reasons: 1) Sost expression at the transcript and protein level (sclerostin) is strongly modulated by PTH signaling (25, 26); 2) serum levels of sclerostin are negatively associated with serum PTH levels in healthy women (27), and postmenopausal women treated with teriparatide have decreased serum sclerostin levels (28); 3) Sost-overexpressing mice and Sost-null mice have reduced anabolic responses to iPTH (29); and 4) the anabolic effects of iPTH require PTHR1 receptor expression in the osteocyte (30), a cell type that normally expresses high levels of Sost in the adult skeleton (31, 32). We sought to understand whether Sost down-regulation is required for the anabolic action of iPTH in the skeleton. Furthermore, we sought to determine whether the bone-building effects of complete Sost inhibition would prevent the bone-wasting effects of high-dose continuous PTH (cPTH).

Materials and Methods

Mice

Genetically engineered mice harboring targeted mutations in the Sost gene have been described previously (33). Briefly, the Sost^+/− mice were engineered by replacing approximately 90% of the Sost coding sequence and all of the single intron, with a Neomycin-resistance cassette, via homologous recombination. Sost^+/− mice, on a mixed genetic background of 129/SvJ and Black Swiss, were bred together to generate Sost^+/+ and Sost^−/− mice for the experiments.

Administration of human parathyroid hormone fragment 1–34 to mice in vivo

Fifty-four male mice (27 Sost^+/+ and 27 Sost^−/−) were used for the intermittent PTH (iPTH) experiments, and 36 female mice (18 Sost^+/+ and 18 Sost^−/−) were used for the cPTH experiments. Mice in the iPTH experiments were assigned to one of three groups: 0 μg/kg iPTH (e.g. vehicle), 30 μg/kg iPTH, or 90 μg/kg iPTH (n = 9/dose group). These mice received single, daily (7 d/wk), sc injections of 100 μl vehicle or human PTH 1–34 (Bachem Inc., Torrance, CA), from 10–16 wk of age. PTH concentrations were adjusted weekly based on weekly body mass measurements.

Mice in the cPTH experiments were assigned to one of two groups: 0 μg/kg/d cPTH (e.g. vehicle) or 40 μg/kg/d cPTH (n = 9/dose group). At 12 wk of age, PTH- or vehicle-filled microosmotic pumps were surgically implanted as previously described (34). The pumps were filled 6 h before implantation with 100 μl of vehicle or approximately 0.11 μg/μl human PTH 1–34 (described above), then primed in 37 C saline, under sterile conditions, until implantation (6 h later). The pumps were engineered to continuously release 0.49 μl/h over 7 d. The mice were reanesthetized 7 d after the initial pump was implanted and surgically prepared, after which the old pump was removed and replaced by a fresh pump. This procedure was repeated a third time another week later, allowing for 3 wk of continuous PTH infusion (with the exception of the brief, 5-min pump changeout periods).

Dual energy x-ray absorptiometry (DEXA)

Mice were anesthetized via isoflurane inhalation and positioned on a GE Lunar PIXImusII dual energy x-ray absorptiometer. Whole-body scans were performed on mice in the iPTH study 2 d before the first PTH injection (9.7 wk of age) and on the day of euthanasia (16 wk of age). Scans were performed on mice in the cPTH study 2 d before pump implantation (11.7 wk of age) and on the day of euthanasia (15 wk of age). From the whole-body scans, areal bone mineral density (aBMD) and bone mineral content (BMC) were calculated for the postcranial skeleton and for the femur and lumbar spine (L3–L5, inclusive) using the ROI tools.

Micro-computed tomography (μCT)

After euthanasia, a 2.6-mm span (∼5 mm³ of medullary space) of the distal femoral metaphysis and a 2.0-mm centrally located span of the fifth lumbar vertebra (∼2.5 mm³ of medullary space) was scanned in 70% ethanol on a desktop μCT (μCT 20; Scanco Medical AG, Bassersdorf, Switzerland), at 13 μm resolution using 50-kVp tube potential and 151-msec integration time, to measure trabecular three-dimensional morphometric properties as previously described (34). Ten μCT slices were collected from the midshaft femur of the cPTH-treated mice, using 9-μm resolution. Midshaft femur slices were imported into ImageJ (National Institutes of Health, Bethesda. MD), in which the total area between the periosteal and endocortical surfaces and the intracortical pore area were measured, using a thresholding algorithm. Pore area was divided by the area between the periosteal and endocortical surfaces (including the pores) to calculate cortical porosity (Ct.Po; %).

Serum collection and biochemical measurements

Among the iPTH experimental mice, blood samples were collected via tail bleeds 3 wk after initiating treatment (midway through the iPTH experiment) to measure resorption markers. To ensure that serum levels of human PTH were elevated in the cPTH experimental mice, blood samples were collected 2 d before exposure to the first pump implantation, then again 5 d after each of the three pump implantations. These samples were used to assay serum mouse and human PTH levels, and the final blood draw (3 wk after initiating infusion) was used to measure serum markers of resorption. Serum was collected, aliquotted, and stored immediately at −80 C. Serum concentration of C-terminal telopeptide (CtX) was measured by a commercially available ELISA (RatLaps; IDS Inc., Scottsdale, AZ). Serum concentration of human PTH 1–34 and mouse intact PTH were measured separately from the same serum samples using two commercially available ELISA that discriminate between endogenous mouse PTH and the human 1–34 fragment of PTH (Immutopics, Inc., San Clemente, CA). Serum samples at each time point were measured in duplicate and averaged.

Fluorochrome administration and bone histomorphometry

Fluorochrome-derived histomorphometric indices of cortical bone formation were measured in the midshaft femur from mice in the anabolic iPTH experiments. Oxytetracycline HCl (40 mg/kg) was injected sc 3 d before the first PTH injection, and alizarin complexone (18 mg/kg) was injected ip 5.5 wk after the first PTH injection (3 d before euthanasia). After euthanasia, left femora were dissected, cleaned, fixed, and processed for undemineralized histomorphometry. Derived histomorphometric parameters included mineralizing surface per unit bone surface (MS/BS, %), mineral apposition rate (MAR, μm/d), and bone formation rate per unit bone surface (BFR/BS, μm³/μm²/yr). Two midshaft femur sections from the iPTH- and vehicle-treated mice were measured for cortical porosity by manually tracing the intracortical pores on digital photomicrographs, using ImagePro Express 6.0 (Media Cybernetics, Inc., Silver Spring, MD). Similar to the μCT protocol described above, pore area was divided by the area between the periosteal and endocortical surfaces (including the pores) to calculate Ct.Po (%).

Statistical analysis

Statistical analysis was computed using JMP (Version 4.0; SAS Institute, Inc., Cary, NC). The endpoints were analyzed using two-way ANOVA, with PTH treatment (0, 30, or 90 μg/kg/d for the iPTH study; 0 or 40 μg/kg/d for the cPTH study) and genotype (Sost^+/+ or Sost^−/−) as main effects. When at least one main effect was significant, interaction terms were calculated and tested for significance. Within each genotype, differences among PTH dose groups were tested for significance using Fisher's protected least-significant difference (PLSD) post hoc test. Statistical significance was taken at P < 0.05. Data are presented as mean ± sem.

Results

Sost deletion results in compromised iPTH anabolic effects on DXA-derived bone mass measurements of the skeleton

To assess the bone-building effects of intermittent PTH (iPTH) in the absence of sclerostin, Sost^+/+ and Sost^−/− mice were treated for 6 wk with 0, 30, or 90 μg/kg/d of iPTH. Whole-body aBMD increased by 6% (3.5 mg/cm²) among Sost^+/+ mice receiving vehicle, whereas Sost^+/+ mice receiving 30 and 90 μg iPTH exhibited a 9% (5 mg/cm²) and an 11% (6 mg/cm²) increase in aBMD (P < 0.05 for both doses, compared with vehicle; Fig. 1, A and B) during the treatment period. However, in the Sost^−/− mice, aBMD increased by 9% (8 mg/cm²) in the vehicle-treated animals and was not further increased as a result of iPTH treatment (ANOVA, P = 0.09 and 0.23 for percent and absolute changes in aBMD, respectively). No differences in body mass gain over the experimental period were detected (Table 1). Whole-body BMC followed similar trends as found for aBMD. (Fig. 1, C and D). DXA-derived aBMD and BMC in femora from Sost^+/+ and Sost^−/− mice showed trends similar to those found for the whole-body measurements (Table 1), although a weak dose response was noted in the femur for Sost^−/− mice. Lumbar spine aBMD and BMC did not exhibit a dose response effect with PTH treatment for either Sost^+/+ or Sost^−/− mice (Table 1).

Fig. 1.

Whole-body DEXA-derived measures of percent change (left side) and absolute change (right side) in aBMD (panels A and B) and BMC (panels C and D) from 10 wk of age (beginning of iPTH treatment) to 16 wk of age (end of iPTH treatment) reveal that iPTH treatment increased aBMD and BMC in a dose-dependent manner among Sost^+/+ mice, but Sost^−/− mice failed to exhibit a significant increase in aBMD or BMC at either iPTH dose. Both percent change and absolute change are presented because the two genotypes have large differences in baseline bone mass. The P values shown at the top of each panel indicate the results of a two-way ANOVA using iPTH treatment and Sost genotype as main effects. Asterisks indicate significant difference (P < 0.05) from the genotype-matched vehicle-treated group, and daggers indicate significant difference from the genotype-matched 30-μg iPTH group, using PLSD post hoc tests.

Table 1.

Body weight, region-specific DEXA measurments, μCT-derived trabecular properties at the distal and spine, and histologically derived bone formation parameters in Sost^+/+ and Sost^−/− mice treated for 6 wk with iPTH 1–34

	Sost+/+			Sost−/−			Main effects (P value)
	0 μg iPTH	30 μg iPTH	90 μg iPTH	0 μg iPTH	30 μg iPTH	90 μg iPTH	Mutation	iPTH	interaction
%Δ Body weight	11.3 ± 2.3	14.3 ± 3.5	12.7 ± 2.0	14.1 ± 2.9	12.7 ± 2.5	14.7 ± 2.3	0.639	0.929	0.688
Distal femur μCT
Tb.N (no./mm2)	2.49 ± 0.4	3.44 ± 0.30a	3.54 ± 0.49a	3.98 ± 0.54	4.27 ± 0.44	4.61 ± 0.76	<0.001	<0.001	0.199
Tb.Th (μm)	65.1 ± 6.8	67.8 ± 7.0	70.0 ± 6.9	109.0 ± 9.7	122.8 ± 17.2	145.7 ± 30.8a,b	<0.001	0.004	0.032
Tb.Sp (μm)	420.4 ± 96.6	283.4 ± 28.3a	275.2 ± 43.2a	216.5 ± 44.9	185.4 ± 38.4	178.7 ± 56.6a	<0.001	<0.001	0.010
Tb.BMC (mg)	0.42 ± 0.1	0.50 ± 0.11	0.59 ± 0.21a	1.51 ± 0.44	1.88 ± 0.43	2.43 ± 0.80a,b	<0.001	0.001	0.024
Whole-femur DEXA
%Δ aBMD	7.7 ± 1.1	18.6 ± 2.0a	20.3 ± 1.7a	16.2 ± 1.2	18.4 ± 1.0	25.4 ± 1.6a,b	<0.001	<0.001	0.015
%Δ BMC	12.1 ± 2.9	23.1 ± 2.7a	35.2 ± 2.5a,b	21.3 ± 2.8	28.5 ± 1.9	35.2 ± 3.4a	0.034	<0.001	0.260
Midshaft femur histology
Ec.MAR (μm/dy)	0.39 ± 0.03	0.52 ± 0.02a	0.54 ± 0.03a	0.45 ± 0.03	0.69 ± 0.03a	0.76 ± 0.03a	<0.001	0.001	0.021
Ec.MS/BS (%)	64.3 ± 3.5	56.1 ± 2.1	64.2 ± 3.6	100.0 ± 0.0	92.9 ± 3.8	98.2 ± 1.1	<0.001	0.053	0.882
Ps.MAR (μm/dy)	0.42 ± 0.05	0.72 ± 0.03a	0.76 ± 0.07a	0.57 ± 0.04	0.82 ± 0.04a	0.93 ± 0.06a,b	0.003	<0.001	0.451
Ps.MS/BS (%)	29.4 ± 3.8	47.7 ± 4.4a	54.0 ± 6.2a	44.2 ± 2.2	70.3 ± 5.8a	69.7 ± 5.6a	<0.001	<0.001	0.837
Ct.BAc (mm2)	1.30 ± 0.18	1.35 ± 0.12	1.52 ± 0.13a,b	2.21 ± 0.22	2.51 ± 0.85	2.26 ± 0.30	<0.001	0.601	0.559
Spine (L3–L5) DEXA
%Δ aBMD	−1.8 ± 2.2	0.7 ± 3.1	−0.9 ± 3.1	9.1 ± 2.3	−4.9 ± 2.2	0.1 ± 2.6	0.327	0.086	0.010
%Δ BMC	−3.1 ± 2.8	6.1 ± 5.0	3.0 ± 4.1	15.0 ± 3.1	−1.5 ± 2.9	5.8 ± 3.3	0.141	0.605	0.003
Spine (L5) μCT
BV/TV (%)	28.8 ± 1.3	23.7 ± 1.3	27.5 ± 3.3	68 ± 5.3	59.2 ± 2.8	66.2 ± 3.7	<0.001	0.092	0.830
Tb.N (no./mm2)	4.19 ± 0.3	3.70 ± 0.16	3.80 ± 0.24	5.60 ± 0.20	6.49 ± 0.22	6.46 ± 0.25	<0.001	0.538	0.006
Tb.Th (μm)	68.3 ± 2.7	76.5 ± 2.1	80.4 ± 4.6	143.2 ± 13.8	108.8 ± 3.4	128.3 ± 9.6	<0.001	0.103	0.013
Tb.Sp (μm)	226.4 ± 7.9	272.8 ± 13.8	268.8 ± 19.7	148.5 ± 12.5	127.7 ± 8.1	129.0 ± 9.2	<0.001	0.462	0.006
Tb.BMC (mg)	0.59 ± 0.03	0.49 ± 0.04	0.53 ± 0.09	1.34 ± 0.10	1.14 ± 0.05	1.24 ± 0.09	<0.001	0.096	0.787

^aSignificantly different (P < 0.05) from vehicle-treated group using Fisher's PLSD post hoc test.

^bSignificantly different (P < 0.05) from 30-μg group using Fisher's PLSD post hoc test.

^cMineralized tissue area between periosteal and endocortical surfaces (excludes area occupied by intracortical pores).

Sost deletion results in enhanced iPTH anabolic effects in the distal femur trabecular bone compartment

Analysis of the distal femur trabecular bone architecture yielded a different result than that obtained by DEXA (Fig. 2, A and B). Bone volume fraction (BV/TV) increased significantly as a result of iPTH treatment in Sost^+/+ mice, from 9.3% in vehicle-treated mice to 13.0% in the 30-μg group, to 16.2% in the 90-μg group (P < 0.05 for both doses, compared with vehicle). BV/TV also increased among iPTH-treated Sost^−/− mice, from 32.9% in vehicle-treated mice to 40.9% in the 30-μg group, to 56.7% in the 90-μg group. Although the 30-μg group failed to reach a statistically significant difference from the vehicle-treated mice (P = 0.10), the 90-μg group was significantly different from both vehicle-treated and 30-μg groups (P < 0.05 for both comparisons).

Fig. 2.

A, Trabecular BV/TV in the distal femoral metaphysis was enhanced in a dose-responsive manner by iPTH treatment in Sost^+/+ and Sost^−/− mice. The P values shown at the top of the panel indicate the results of a two-way ANOVA using iPTH treatment and Sost genotype as main effects. Asterisks indicate significant difference (P < 0.05) from the genotype-matched vehicle-treated group, and the dagger indicates significant difference from the genotype-matched 30-μg iPTH group, using PLSD post hoc tests. B, Representative images of μCT reconstructions (anterior portion removed in silico to reveal trabecular bone) from the distal femur of Sost^+/+ and Sost^−/− mice after 6 wk of 0, 30, or 90 μg/kg iPTH. Note the large difference in baseline bone phenotype between Sost^+/+ and Sost^−/− mice, but also the dose-responsive increase in trabecular mass with increasing iPTH dose, in both genotypes. Other μCT-derived architectural properties are reported in Table 1. C, Serum levels of type I collagen CTx, a marker of bone resorption, after 3 wk of treatment were decreased in the 30-μg iPTH groups and increased in the 90-μg iPTH groups. Although neither treated group was significantly different from the vehicle-treated group, the 90-μg group had significantly higher levels of serum CTx than the 30-μg group for both genotypes. Results from the two-way ANOVA (upper corner of panel) reveal a significant iPTH effect and a significant Sost genotype effect, but no interaction was found, indicating that the two genotypes responded similarly to iPTH. The dagger indicates significant difference from the genotype-matched 30-μg iPTH group, using PLSD post hoc tests.

The iPTH-induced increase in BV/ TV among the 90-μg group, when compared with the vehicle-treated group, was similar between genotypes (74% increase in Sost^+/+ mice vs. 70% increase in Sost^−/− mice), but a significant interaction term was found (P = 0.03) from a two-way ANOVA, using genotype and iPTH group as main effects, suggesting that the Sost^−/− might have responded more anabolically than the Sost^+/+ mice. Because BV/TV is a measurement relative to medullary volume, we conducted a similar analysis on a more absolute measure of bone gain in the trabecular compartment: trabecular BMC. Among the Sost^+/+ mice, the iPTH-induced increase in Tb (trabecular).BMC resulting from 90 μg iPTH treatment was 61% greater than in the vehicle-treated mice, whereas among the Sost^−/− mice group, the iPTH-induced increase was only 40% greater than in the vehicle-treated mice (Table 1). Not surprisingly, a significant interaction term was found (P = 0.024), suggesting that iPTH added more trabecular bone mass in the Sost^−/− mice, compared with the Sost^+/+ mice. Similar results were found for other femoral trabecular bone parameters, including trabecular thickness and spacing, but not trabecular number (Table 1). The fifth lumbar vertebral trabecular bone compartment was analyzed via μCT, but like the DEXA measurements of the same region, iPTH treatment failed to increase trabecular parameters in a dose-response manner in either the Sost^+/+ or Sost^−/− mice (Table 1).

iPTH-induced changes in resorption markers are not altered by Sost deletion

To gain insight into whether the absence of sclerostin alters the bone resorption response to iPTH, we measured CTx levels, a standard serum marker of bone resorption. Untreated Sost^−/− mice had significantly greater (P < 0.05) serum levels of CTx than untreated Sost^+/+ mice (23.3 vs. 32.2 ng/ml; P < 0.05), suggesting more baseline resorption occurred in the Sost^−/− mice (Fig. 2C). In both genotypes, iPTH treatment resulted in a decrease in CTx at the 30-μg PTH dose, and an increase in CTx at the 90-μg PTH dose, although statistical significance was reached only for the 30- vs. 90-μg group comparisons. A two-way ANOVA revealed significant genotype and PTH effects, but no significant interaction was found (P = 0.97), indicating that the mice resorbed bone similarly in response to iPTH treatment, whether Sost was deleted or not.

iPTH-induced cortical bone formation parameters and cortical bone porosity are enhanced by Sost deletion

To probe further into the potential compartmental specificity of the Sost effects on iPTH anabolism, fluorochrome-derived cortical bone formation rates were measured on the periosteal and endocortical surfaces of the midshaft femur. Periosteal (Ps.) bone formation rates (Ps.BFR/BS) were 13 μm³/μm²/yr in vehicle-treated Sost^+/+ mice and were increased significantly, to 34–39 μm³/μm²/yr by iPTH-treatment (P < 0.001 for each treatment vs. vehicle control; Fig 3A). Ps.BFR/BS in vehicle-treated Sost^−/− mice was 25 μm³/μm²/yr and was increased significantly, to 58–68 μm³/μm²/yr by iPTH-treatment (P < 0.004 and P = 0.001 for the 30- and 90-μg groups vs. vehicle control, respectively). No difference between mice in the two iPTH dose groups (30 vs. 90 μg) was found for either genotype (P = 0.20–0.22). In addition, the mutation × iPTH interaction term was not statistically significant, indicating that iPTH enhanced Ps.BFR/BS equally in both genotypes. The iPTH-induced gains in Ps.BFR/BS were fueled increases in both activated surface and apposition rates (Table 1).

Fig. 3.

A, Midshaft femur fluorochrome-derived BFR on the periosteal surface were significantly and dose-responsively increased by iPTH treatment, regardless of genotype. B, BFR/BS on the endocortical surface was not significantly altered in Sost^+/+ mice, but Sost^−/− mice exhibitied a significant, dose-responsive increase in Ec.BFR/BS. D, Photomicrographs of undemineralized midshaft femur sections from Sost^+/+ and Sost^−/− mice that were labeled with tetracycline (golden label indicated by yellow arrows) at the beginning of the experiment, and alizarin complexone (red label near bone edge indicated by pink arrows) at the end of the experiment show similar iPTH-induced increases in new periosteal bone formation (i.e. bone between the gold and red labels). Note the greater baseline bone formation in the Sost^−/− mice. Whole-bone photomicrographs illustrate the iPTH-induced increase in cortical porosity (orange arrows) among Sost^−/− mice, which is quantified in panel C. The P values shown at the top of each panel indicate the results of a two-way ANOVA using iPTH treatment and Sost genotype as main effects. Asterisks indicate significant difference (P < 0.05) from the genotype-matched vehicle-treated group, and daggers indicate significant difference from the genotype-matched 30-μg iPTH group, using PLSD post hoc tests.

On the endocortical (Ec)surface, bone formation rates (Ec.BFR/BS) were unchanged by iPTH treatment in Sost^+/+ mice (25–31 μm³/μm²/yr; P = 0.12), but Sost^−/− mice exhibited a significant, dose-responsive increase in Ec.BFR/BS with iPTH treatment (Fig 3B). Vehicle-treated Sost^−/− mice exhibited an endocortical formation rate of 45 μm³/μm²/yr, which was significantly enhanced to 64 μm³/μm²/yr among the 30-μg iPTH group (P = 0.005), and further enhanced to 75 μm³/μm²/yr among the 90 μg iPTH group (P < 0.001 vs. vehicle control and P = 0.034 vs. 30-μg group). A significant interaction term (mutation × iPTH) was found, indicating that the Sost^−/− mice were more responsive to iPTH treatment than Sost^+/+ mice for Ec.BFR/BS. The iPTH-induced gains in Ec.BFR/BS among Sost^−/− mice (no iPTH effect was found among Sost^+/+ mice) were fueled largely by gains in mineral apposition rates, with no measureable effects on activated surface (Table 1). The lack of effect of iPTH on Ec.MS/BS was likely the result of a saturated surface at baseline; vehicle-treated Sost^−/− mice were at or near 100% Ec.MS/BS.

The serial whole-body and femoral DEXA measurements, which are dominated by cortical bone mass, suggested that Sost^−/− mice responded poorly to iPTH, yet the cortical bone formation parameters derived from histological analysis revealed that cortical bone formation parameters among Sost^−/− mice responded to iPTH as well or better than that of Sost^+/+ mice. Those two observations were difficult to reconcile, so we followed up on an observation made while reading the sections for bone formation rates: a noticeable variability in porosity. We measured Ct.Po within the femoral cortex from the same sections used to derive bone formation rates (Fig 3D). Ct.Po was unaffected by iPTH treatment in Sost^+/+ mice (0.2–1.4%, P = 0.25), whereas Sost^−/− mice exhibited a significant increase in Ct.Po with iPTH treatment, from 4% among vehicle-treated Sost^−/− mice to 12–14% among iPTH-treated Sost^−/− mice (P = 0.012 and P = 0.048 for the 30- and 90-μg groups vs. vehicle control, respectively; Fig 3C). The mutation × iPTH interaction term was borderline significant (P = 0.05) for Ct.Po, indicating that cortical porosity might have been increased by iPTH treatment more dramatically in Sost^−/− mice, compared with Sost^+/+ mice. The antagonistic effects of increased porosity in the presence of increased formation rates were summarized in the measurements made for midshaft femur cortical tissue area (calculated without pore area), which exhibited an iPTH-induced increase among Sost^+/+ mice but remained steady in Sost^−/− mice (Table 1).

Sost deletion has no effect on continuous PTH (cPTH)-induced catabolism in the whole skeleton

We next assessed whether the bone-wasting effects of continuous PTH (cPTH) infusion were altered by the absence of sclerostin. Sost^+/+ and Sost^−/− mice were treated for 3 wk with continuous infusion of vehicle or high-dose (40 μg/kg/d) human PTH 1–34. Analysis of human PTH 1–34 levels in the serum, collected 5 d after implantation of each of the three weekly osmotic pumps, indicated that the pumps were working properly (Fig. 4).

Fig. 4.

A, Serum levels of human PTH (the minipump infusate) were below detection level in vehicle infused Sost^+/+ mice (solid triangles) at baseline and each time point thereafter, indicating that the ELISA was reliably measuring human and not mouse endogenous mouse PTH. Sost^+/+ (open triangle) and Sost^−/− (open circle) mice infused with 40 μg/kg/d PTH had significantly elevated human PTH levels 5 d after each of the three sequential minipumps was implanted and/or replaced (indicated by P arrows along x-axis), indicating that the minipumps were reliably achieving elevated serum human PTH levels. B, Serum levels of endogenous intact mouse PTH were significantly reduced in human PTH-infused Sost^+/+ (open triangle) and Sost^−/− mice (open circle), but remained steady in vehicle-infused Sost^+/+ (solid triangle) and Sost^−/− (solid circle) mice. Asterisks indicate significant difference (P < 0.05) from the genotype-matched vehicle-treated group at each time point. veh, Vehicle.

During the 3-wk cPTH infusion period, whole-body areal bone mineral density (aBMD) increased nonsignificantly by 2% (1.2 mg/cm²) among Sost^+/+ mice receiving vehicle, whereas Sost^+/+ mice receiving 40 μg cPTH infusion exhibited a 5% (2.8 mg/cm²) decrease in aBMD (P < 0.01; Fig. 5, A and B). In the Sost^−/− mice, aBMD increased by 5.9% (4.1 mg/cm²) in the vehicle-infused animals, whereas Sost^−/− mice receiving 40 μg cPTH infusion exhibited a 4.1% (3.4 mg/cm²) decrease in aBMD (P < 0.01). A two-way ANOVA, using genotype and infusion group (cPTH or vehicle) as main effects, resulted in a significant infusion effect (P < 0.01), with no significant genotype effect or interaction term. These data suggest that the absence of sclerostin did not alter the degree of bone loss (measured at the whole-body level) resulting from cPTH infusion.

Fig. 5.

Whole-body DEXA-derived measures of percent change (left side) and absolute change (right side) in aBMD (panels A and B) and BMC (panels C and D) from 12 wk of age (beginning of cPTH infusion) to 15 wk of age (end of cPTH infusion) reveal that cPTH infusion decreased BMC and aBMD equally in Sost^+/+ and Sost^−/− mice. Both percent change and absolute change are presented because the two genotypes have large differences in baseline bone mass. The P values shown at the bottom of each panel indicate the results of a two-way ANOVA using cPTH treatment and Sost genotype as main effects. Asterisks indicate significant (P < 0.05) change from baseline measurement to the final measurement, using a paired t test. The significant interaction term yielded for BMC indicates that the Sost mutation significantly affected the way cPTH infusion altered bone, but closer inspection of the data indicates that the interaction was fueled largely by increases in BMC among the vehicle-treated Sost^−/− mice, compared with the vehicle-treatedd Sost^+/+ mice, rather than by genotype-related differences in cPTH effects on bone loss. Veh, Vehicle.

Whole-body BMC followed similar trends as found for aBMD, for both relative (% change) and absolute changes during the experimental period (Fig. 5, C and D). A two-way ANOVA resulted in a significant infusion effect (P < 0.01), a nonsignificant genotype effect, and a significant interaction effect (P < 0.05). However, follow-up tests revealed that the statistical significance of the interaction term was driven not by genotype-related differences in the degree of cPTH-induced bone loss, which was similar between genotypes, but rather by the bone gain in the vehicle-infused mice, which was significantly greater in the Sost^−/− mice. Similar results were found for serial DEXA measurements collected from the femur and lumbar spine (Table 2). These data suggest that the absence of sclerostin did not alter the degree of bone loss, as measured by DEXA, resulting from continuous PTH infusion.

Table 2.

Body weight, region-specific DEXA measurements, μCT-derived trabecular properties at the distal femur and spine, and μCT-derived cortical porosity in Sost^+/+ and Sost^−/− mice treated for 3 wk with cPTH 1–34 infusion

	Sost+/+		Sost−/−		Main effects (P value)
	0 μg cPTH	40 μg cPTH	0 μg cPTH	40 μg cPTH	Mutation	cPTH	Interaction
%Δ Body weight	5.11 ± 1.03a	−2.8 ± 1.59a	4.3 ± 1.34a	0.87 ± 1.4	0.329	0.002	0.182
Distal femur μCT
Tb.N (no./mm2)	3.51 ± 0.12	2.43 ± 0.24a	4.05 ± 0.20	3.95 ± 0.20	<0.001	0.007	0.020
Tb.Th (μm)	62.5 ± 2.0	54.9 ± 2.00a	99.9 ± 2.9	105.0 ± 3.4	<0.001	0.653	0.026
Tb.Sp (μm)	285.9 ± 11.3	442.7 ± 45.0a	216.2 ± 16.6	219.5 ± 18.3	<0.001	0.008	0.010
Tb.BMC (mg)	0.56 ± 0.05	0.21 ± 0.03a	1.40 ± 0.15	1.32 ± 0.15	<0.001	0.062	0.223
Whole femur DEXA
%Δ aBMD	4.3 ± 1.5	−2.8 ± 1.4a	6.8 ± 1.3	−1.8 ± 1.1a	0.207	<0.001	0.584
%Δ BMC	6.3 ± 2.4	2.7 ± 1.9a	12.1 ± 1.5	2.7 ± 1.3a	0.129	0.001	0.131
Midshaft femur cortical porosity
Ct.Po (%)	0.07 ± 0.03	0.72 ± 0.24a	1.44 ± 0.48	3.52 ± 0.70a	<0.001	0.005	0.133
Ct.BAb (mm2)	1.00 ± 0.02	0.91 ± 0.03a	1.66 ± 0.05	1.62 ± 0.04	<0.001	0.227	0.210
Spine (L3–L5) DEXA
%Δ aBMD	−3.9 ± 3.2	−12.8 ± 2.1a	1.9 ± 2.0	−12.8 ± 1.9a	0.229	<0.001	0.226
%Δ BMC	−7.5 ± 3.3	−19.0 ± 2.5a	6.5 ± 2.8	−16.3 ± 3.5a	0.011	<0.001	0.077
Spine (L5) μCT
Tb.N (no./mm2)	4.71 ± 0.2	3.48 ± 0.1a	5.71 ± 0.3	5.47 ± 0.2	<0.001	0.005	0.012
Tb.Th (μm)	65.3 ± 1.4	61.1 ± 2.2a	149.7 ± 13.5	112.6 ± 5.4a	<0.001	0.001	0.010
Tb.Sp (μm)	207.7 ± 7.4	292.3 ± 6.9a	143.0 ± 13.1	154.8 ± 8.8	<0.001	<0.001	<0.001
Tb.BMC (mg)	0.44 ± 0.02	0.26 ± 0.02a	1.14 ± 0.08	0.89 ± 0.08a	<0.001	<0.001	0.479

^aSignificantly different (P < 0.05) from vehicle-treated group using Fisher's PLSD post hoc test.

^bMineralized tissue area between periosteal and endocortical surfaces (excludes area occupied by intracortical pores).

Sost deletion protects trabecular bone in the distal femur, but not in the spine, from the catabolic effects of cPTH

Similar to the iPTH results, analysis of the distal femur trabecular bone architecture in the cPTH experiments yielded a different result than that obtained by whole-body DEXA (Fig. 6, A and B). Femoral BV/TV was significantly lower in cPTH-infused Sost^+/+ mice, dropping from 12.6% in the vehicle-infused mice to 5.8% in the PTH-infused mice (P < 0.01). However, among Sost^−/− mice, PTH infusion had no effect on BV/TV (29.5% vs. 31.6%; P = 0.64). Similar deficiencies in other distal femur trabecular bone parameters were found for Sost^+/+ mice, but not for Sost^−/− mice (Table 2).

Fig. 6.

A, Trabecular bone volume fraction (BV/TV) in the distal femoral metaphysis was reduced by approximately 50% among cPTH-infused Sost^+/+ mice, whereas Sost^−/− mice exhibited no detectable loss of BV/TV after cPTH infusion. The P values shown at the top of the panel indicate the results of a two-way ANOVA using cPTH treatment and Sost genotype as main effects. The asterisk indicates a significant difference (P < 0.05) from the genotype-matched vehicle-infused group. B, Representative images of μCT reconstructions (anterior third and posterior third removed in silico to reveal trabecular bone) from the distal femur of Sost^+/+ and Sost^−/− mice after 3 wk of vehicle or 40 μg/kg cPTH. Note the clear loss of trabecular bone in the Sost^+/+ but not Sost^−/− mice. Other μCT-derived architectural properties are reported in Table 2. C, Fifth lumbar vertebral BV/TV was reduced by approximately 20–35% in response to cPTH treatment, regardless of genotype. Results from the two-way ANOVA (upper corner of panel) reveal significant cPTH and genotype effect, but the Sost mutation had no effect on the degree of bone volume loss induced by cPTH (i.e. nonsignificant interaction term). Other μCT-derived architectural properties are reported in Table 2. D, Representative μCT reconstructions from the central 25% of the fifth lumbar vertebra from Sost^+/+ and Sost^−/− mice after 3 wk of vehicle or 40 μg/kg cPTH. Note the clear loss of trabecular bone in both Sost^+/+ and Sost^−/− mice. E, Serum levels of type I collagen CTx, a marker of bone resorption, were increased after 3 wk of cPTH treatment in Sost^+/+ and Sost^−/− mice, compared with genotype-matched vehicle-infused mice, but statistical significance (asterisk) was reached only for the increase among Sost^+/+ mice. Results from the two-way ANOVA (upper corner of panel) reveal a significant cPTH effect, but the Sost mutation had no effect on CTx, nor did it change the response to cPTH (nonsignificant interaction term). F, Midshaft femur μCT tomographs from vehicle-infused and cPTH-infused Sost^+/+ and Sost^−/− mice. Note the relative increase in cortical porosity (quantified in Table 2) induced by cPTH infusion in both genotypes.

Among Sost^−/− mice, the fifth lumbar trabecular bone compartment was not protected from cPTH-induced loss, as was observed in the distal femur (Fig 6, B and F). Vertebral BV/TV was reduced in both Sost^+/+ and Sost^−/− mice to approximately the same degree. A significant interaction term was not found (P = 0.29), indicating that the Sost mutation did not reduce the cPTH effect on bone loss in the spinal trabecular bone. Other μCT-derived parameters in the spine yielded similar outcomes (e.g. Tb.BMC), more loss (Tb.Th), or less loss (Tb.N, Tb.Sp) among the Sost^−/− mice (Table 2).

cPTH-induced changes in resorption markers are not altered by Sost deletion, but cortical bone is susceptible to cPTH-induced porosity in Sost^−/− mice

To assess the bone resorption response to cPTH, we measured CTx levels in serum samples collected at euthanasia. Untreated Sost^−/− mice had slightly greater levels of CTx than untreated Sost^+/+ mice, but this difference was not significant (Fig. 6C). cPTH treatment resulted in a significant increase in CTx among Sost^+/+ mice. We observed a similar increase in CTx among PTH-infused Sost^−/− mice, compared with vehicle-infused Sost^−/− mice, but there was a large degree of variation among the cPTH-treated serum samples that precluded a statistically significant result. A two-way ANOVA revealed only a significant infusion effect, with no significant genotype or interaction term.

Ct.Po measurements from the midshaft femur in cPTH- and vehicle-infused mice revealed that Ct.Po was increased significantly by cPTH infusion in both Sost^+/+ and Sost^−/− mice (P = 0.021 and P = 0.047, respectively; Fig. 6E and Table 2). The Sost^−/− mice had a greater degree of baseline porosity than the Sost^+/+ mice, but both genotypes exhibited similar relative increases in porosity as a result of cPTH infusion (mutation × cPTH interaction term was P = 0.13).

Discussion

One of our main objectives in the current study was to understand whether modulation of sclerostin levels plays a role in the anabolic action of iPTH. The Sost^−/− mouse model allowed us to assess the effects disregulated Sost expression, in which modulation of (a significant drop in) sclerostin levels in response to PTH treatment, could be prevented via genetic deletion of the gene. Two other studies, one genetic and one pharmacological, have shown that when sclerostin levels are not permitted to decrease normally in response to iPTH treatment, the anabolic effects of iPTH are abrogated (29, 35). We found that in the absence of Sost, iPTH treatment was either ineffective or significantly reduced in its capacity to improve BMD and BMC (a measurement dominated by the cortical bone envelope), whereas those parameters increased significantly when the same iPTH regimen was applied to Sost^+/+ mice. These findings are consistent with a previous report of iPTH treatment in young Sost^−/− mice, in which 8 wk of 100 μg/kg/d iPTH treatment increased tibia and spine BMD significantly in Sost^+/+ mice but not in Sost^−/− mice (29). However, our data suggest that the inability of Sost^−/− cortical bone to respond anabolically to iPTH was not related to impaired cortical bone formation rates. To the contrary, our histomorphometric data from the midshaft femur suggest that Sost^−/− mice responded to iPTH treatment equally well on the periosteal surface, and more robustly on the endocortical surface, than Sost^+/+ mice, but cortical porosity was concomitantly increased. In other words, iPTH-induced gains in cortical bone mass, via increased periosteal and endocortical bone formation parameters, occurred in both genotypes, but those gains were nullified on a DEXA scan by increased intracortical resorption in the Sost^−/− mice. The amplification of iPTH-stimulated resorption in Sost-deficient mice was unexpected and highlights a potential interaction between the resorptive arm of the iPTH response and Wnt signaling.

Whereas cortical bone exhibited impaired iPTH-induced net gain among Sost^−/− mice, the trabecular compartment (distal femur) exhibited significant iPTH-induced gains in mass and structural properties, and for some measurements, exceeded those found in the iPTH-treated Sost^+/+ mice. Thus, it appears that iPTH treatment is fully anabolic in the femoral trabecular envelope, whether Sost is absent or present. Basal trabecular bone mass was already very great in the Sost^−/− mice, yet iPTH treatment was able to further enhance bone accrual in this envelope. These trabecular data, in conjunction with the enhanced cortical bone formation rates discussed earlier, suggest that Sost deletion does not by itself automatically bring the skeleton to its maximal osteogenic potential. Other anabolic pathways, either related to or independent of Wnt signaling, appear to be accessible when sclerostin-mediated Wnt signaling is highly stimulated or otherwise altered.

In light of the significantly enhanced cortical porosity among iPTH-treated Sost^−/− mice, it is unclear why serum measures of resorption (i.e. CTx) were not elevated beyond those observed in the Sost^+/+ mice. Serum CTx was slightly, but significantly, greater in vehicle-treated Sost^−/− mice, compared with vehicle-treated Sost^+/+ mice. It is possible that the timing of our blood draw during the experiment was improperly scheduled to adequately capture the increase in resorption that presumably occurred during the iPTH-induced enlargement of intracortical pores.

Sclerostin is not the only Lrp5/6 inhibitor, and so direct assessment of the role of Wnt signaling in PTH action cannot be assessed solely on the basis of Sost manipulation. Other Wnt inhibitors might also play a role in PTH action. For example, Dkk1 expression is down-regulated after PTH exposure (36, 37), and it is unclear whether the expression level of other Wnt-Lrp5/6 inhibitors (e.g. Dkk1, sFrp1, Wise, Wif1) is altered as part of a compensatory mechanism when Sost is deleted. Mice overexpressing Dkk1 driven by the Col2.3 promoter fragment exhibit normal anabolic responsiveness to iPTH, whereas mice overexpressing sFrp1 driven by the endogenous promoter (i.e. BAC) (38) or mice with sFrp1 deletion (39) exhibit reduced anabolic responsiveness to iPTH.

Recently, Lrp6 has been reported to act as a facilitator of PTH signaling by directly interacting with PTH1R in the presence of PTH (23). More recently, PTH-activated PTH1R was shown to interact directly with Disheveled (Dsh) to induce canonical β-catenin signaling, and the interaction and subsequent signal transduction were independent of Lrp5 or Lrp6 (40). Moreover, Dkk1 overexpressing mice have impaired cPTH responsiveness, but the PTH-induced β-catenin signal (TopGal) in these mice is unaffected by Dkk1 overexpression (36). Thus, while the role of Lrp5 in iPTH signaling appears to be minimal, the role of Lrp6 is controversial. However, the amplification of iPTH's resorptive effects in Sost^−/− mice suggests that there is, at a minimum, significant cross-talk between Wnt- and PTH-activated pathways. Alternatively, the list of known sclerostin-binding partners is currently expanding and now includes additional proteins both in and out of the Wnt pathway [e.g. Bmps (41, 42) Lrp5/6 (43–46), Lrp4 (47, 48), Cyr61 (49), and ErbB3 (50)]. Thus, potential Sost-mediated iPTH effects on pathways other than Wnt cannot be ruled out.

Whereas our observations of reduced iPTH-induced bone gain among Sost^−/− mice are consistent with a previous report of blunted iPTH responsiveness in an independently engineered Sost^−/− mouse model, our data invoke a different mechanism for the deficiency. Kramer et al. (29) reported that reduced bone formation rates, largely driven by reduced mineralizing surface, were responsible for the blunted iPTH anabolism in Sost^−/− mice, compared with Sost^+/+ mice. Conversely, we found cortical bone formation rates to be fully responsive, perhaps even more responsive (e.g. in the case of endocortical measurements) to iPTH in Sost^−/− mice, compared with Sost^+/+ mice. The deficiency in radiographically measured bone mass in our experiments was a result of increased iPTH-induced cortical porosity, which detracted from the gains in cortical bone mass made by iPTH-induced increases in bone formation rates. A further discrepancy revolves around the responsiveness of the trabecular envelope, which was equally or more responsive to iPTH among Sost^−/− mice, compared with Sost^+/+ mice, in our experiment. Kramer et al. (29) reported no iPTH induced changes in μCT-derived Tb.BV/TV in the distal femur for either genotype (we found a similar lack of responsiveness in the spine), but other architectural properties were equally altered by iPTH treatment regardless of genotype. Several differences in experimental design might account for the disparate conclusions reported in their communication and ours.

First, we measured fluorochrome-derived bone formation rates in the cortical bone, whereas their analyses were carried out on trabecular bone. Our initial μCT observations revealing fully anabolic iPTH effects in the trabecular bone of Sost^−/− mice provided little grounds for following up with histomorphometric measurements at that site. Conversely, the failure to gain cortical bone as revealed by DEXA scans prompted us to probe the cortical envelope histologically. Second, we used near-skeletally mature mice (10–16 wk of age) for our experiment, which have significantly different genotype-related baseline phenotypes at that age, whereas Kramer et al. (29) began their experiments at 6 wk of age, 2 wk after the time point at which bone mass was nearly identical between Sost^−/− and Sost^+/+ mice. It is perhaps relevant to point out that the Sost^−/− mouse model we used had significantly greater bone mass at 4 wk of age, compared with Sost^+/+ (data not shown).

Another objective of our study was to discern whether Sost deletion would protect mice from the bone-wasting effects of high-dose continuous PTH (cPTH) infusion, a model for patients with moderate to severe hyperparathyroidism. We found that both Sost^+/+ and Sost^−/− mice exhibited equal amounts of bone loss in response to 3 wk cPTH infusion, as indicated by significant reductions in DEXA-derived BMD and BMC. However, the distal femur trabecular bone-wasting effects of cPTH infusion were absent in Sost^−/− mice, but not in Sost^+/+ mice. It is interesting to note that the protective effect of Sost deletion on cPTH-induced trabecular bone loss was present in the distal femur spongiosa but not in the lumbar vertebral spongiosa. Although the reason for differential effects at these two sites is elusive, it is relevant to point out that the synergistic effect of PTH with mechanical loading often produces enhanced effects in the limbs compared with the spine (51), and Sost plays a major role in the regulation of loading effects in the skeleton (52). As was seen with iPTH treatment, cPTH infusion induced a significant increase in cortical porosity in Sost^−/− mice. Unlike the iPTH results, however, we were able to detect a significant increase in cortical porosity among Sost^+/+ mice as well. cPTH-induced changes in femoral midshaft porosity in both genotypes were consistent with serum CTx levels measured after 3 wk cPTH infusion, which were elevated in both genotypes (although statistical significance was reached only for Sost^+/+ mice; see Fig. 6D).

It is noteworthy that deletion of Sost (53) or pharmacological inhibition of sclerostin (54) can protect the skeleton from the cortical and trabecular bone-wasting effects of mechanical signaling deficits (i.e. disuse). Our data suggest that the bone-wasting effects of hormonal challenge (i.e. cPTH) are rescued by Sost deletion in an envelope- and site-specific manner, i.e. no protection from cPTH-induced cortical bone loss is associated with Sost deletion, whereas the femoral but not spinal trabecular bone compartment of Sost^−/− mice was unaffected by the otherwise catabolic effects of cPTH infusion. Estrogen deprivation, another model of hormonal challenge to the skeleton, was found to induce more severe cortical and cancellous bone loss in ovariectomized Sost^−/− mice than was observed in ovariectomized Sost^+/+ mice (55).

In summary, we found that mice with homozygous deletion of the Sost gene exhibited diminished bone-building effects of iPTH treatment in the cortical but not trabecular bone envelopes. The cortical attenuation of bone gain in Sost^−/− mice was the result of increased iPTH-induced bone resorption despite normal, or perhaps even enhanced, cortical bone formation rates. iPTH-induced femoral trabecular bone gain was not compromised in Sost-deficient mice. Continuous PTH infusion was equally catabolic in Sost^+/+ and Sost^−/− mice, but the femoral trabecular compartment was spared the bone-wasting effects of cPTH infusion in Sost^−/− but not Sost^+/+ mice. PTH is a major calcium-regulating hormone that can activate multiple pathways, and sclerostin has clear effects on PTH signaling in the skeleton. But the envelope-specific nature of the effects, and the distinctly altered roles of resorption vs. formation, suggest that sclerostin's role in PTH signaling is highly complex.