The Sirtuin1 Activator SRT3025 Down-Regulates Sclerostin and Rescues Ovariectomy-Induced Bone Loss and Biomechanical Deterioration in Female Mice

Abstract

Estrogen deficiency leads to rapid bone loss and skeletal fragility. Sclerostin, encoded by the sost gene, and a product of the osteocyte, is a negative regulator of bone formation. Blocking sclerostin increases bone mass and strength in animals and humans. Sirtuin1 (Sirt1), a player in aging and metabolism, regulates bone mass and inhibits sost expression by deacetylating histone 3 at its promoter. We asked whether a Sirt1-activating compound could rescue ovariectomy (OVX)-induced bone loss and biomechanical deterioration in 9-week-old C57BL/6 mice. OVX resulted in a substantial decrease in skeletal Sirt1 expression accompanied by an increase in sclerostin. Oral administration of SRT3025, a Sirt1 activator, at 50 and 100 mg/kg·d for 6 weeks starting 6 weeks after OVX fully reversed the deleterious effects of OVX on vertebral bone mass, microarchitecture, and femoral biomechanical properties. Treatment with SRT3025 decreased bone sclerostin expression and increased cortical periosteal mineralizing surface and serum propeptide of type I procollagen, a bone formation marker. In vitro, in the murine long bone osteocyte-Y4 osteocyte-like cell line SRT3025 down-regulated sclerostin and inactive β-catenin, whereas a reciprocal effect was observed with EX-527, a Sirt1 inhibitor. Sirt1 activation by Sirt1-activating compounds is a potential novel pathway to down-regulate sclerostin and design anabolic therapies for osteoporosis concurrently ameliorating other metabolic and age-associated conditions.

Estrogen deficiency leads to increased bone turnover, bone loss, defects in microarchitecture, and skeletal fragility (1). Antiresorptive therapies prevent the progression of structural damage but do not restore lost bone. Sclerostin, encoded by the sost gene and a product of the osteocyte, the longest lived cell in bone, inhibits bone formation, and blocking its action increases bone formation, mass, and strength. Genetic deletion of sost in mice or its inactivating mutations in humans results in a high bone mass phenotype (2–4), whereas mice overexpressing sost display low bone mass and strength (5). Moreover, treatment of estrogen-deficient rats and monkeys with antisclerostin antibodies restores bone mass and strength beyond control animals (6, 7). Lately, romosozumab, a humanized monoclonal antisclerostin antibody, was shown to increase bone mineral density (BMD) and bone formation in postmenopausal women with low bone mass (8).

Sclerostin is regulated by factors and hormones that influence bone formation, including mechanical stimuli, parathyroid hormone (PTH), bone morphogenic proteins, and sex hormones (9). A role for estrogens in sclerostin regulation was suggested by the observations that serum sclerostin is significantly higher in postmenopausal compared with premenopausal women, and estrogen treatment to postmenopausal women reduces sclerostin in serum (10), bone marrow plasma (11), and its mRNA expression at the iliac crest (12).

Sirtuin1 (Sirt1), an NAD⁺-dependent deacetylase and a key player in aging and metabolism (13), was shown to regulate bone mass. Its general overexpression protects against age-associated bone loss in male mice (14), whereas targeted Sirt1 deletion in osteoblasts and/or osteoclasts results in low bone mass (15). We have previously reported that adult Sirt1 haplo-insufficient female mice display low bone mass characterized by reduced bone formation. Moreover, we discovered sost as a target of Sirt1 and demonstrated that Sirt1 inhibits sost gene expression by deacetylating lysine 9 on histone 3 at its promoter (16). These findings prompted us to ask whether pharmacologic activation of Sirt1 could down-regulate sclerostin and reverse ovariectomy (OVX)-induced bone loss, the classic mouse model of human postmenopausal osteoporosis.

Synthetic Sirt1-activating compounds (STACs) were generated and were shown to be allosteric activators of Sirt1 (17). Their beneficial effects in mouse models of type 2 diabetes mellitus (18), emphysema (19), asthma (20), myocardial infarction (21), and multiple sclerosis (22) were previously reported, but their role in osteoporosis has not yet been investigated. In this study, we evaluated the skeletal effects of SRT3025, a third-generation STAC and currently under investigation in humans (23), in OVX mice and discovered that SRT3025 down-regulates the expression of sclerostin in bone in vivo and in vitro and rescues OVX induced-bone loss and biomechanical deterioration.

Materials and Methods

Animal experimentation

Eight groups (10 mice/group) of 8-week-old C57BL/6J female mice were purchased from Harlan and were maintained in the animal facility at Harlan, Rehovot, Israel, 1 week for acclimation. Mice were housed in a constant temperature room with a 12-hour light, 12-hour dark cycle and were allowed free access to Harlan Teklad chow with 18.6% protein, 1% calcium, and 2-IU/g vitamin D₃ and water. Nine-week-old C57BL/6J female mice were subjected to OVX or sham operation (SHAM) and were left untreated for 6 weeks to allow for bone loss. A group of SHAM and a group of OVX mice was killed 1 week after surgery, and additional groups of SHAM and OVX mice were killed 6 weeks after operation to determine bone loss before treatment initiation. The remaining OVX mice were treated from week 6 after OVX with either SRT3025 at 50 mg/kg·d, 100 mg/kg·d, or a vehicle (polyethylene glycol 400:Tween 80:deionized water at a ratio of 20:0.5:79.5 [vol/vol], respectively) administered daily by gavage for 6 weeks. One SHAM group received a vehicle alone (Figure 1A). Serum samples were collected on days 7, 42, 63, and 84 after OVX after a 12-hour fast via retroorbital venous sampling under light CO₂ inhalation, immediately separated, and kept frozen at −20°C until analyzed. To label bone-forming surfaces, mice were injected with calcein (Sigma-Aldrich) at 15 mg/kg, 7 and 2 days before killing. Upon killing by carbon dioxide asphyxiation, serum was collected for drug level determination (Supplemental Figure 1A), 1 femur and the fourth vertebrae were removed and kept in 10% neutral-buffered formalin for micro computed tomography (μCT) and histomorphometric analyses. The second femur was frozen in soaked gauze in −20°C for biomechanical testing. After flushing out the marrow, tibiae were kept in −80°C for protein purification. Body and uterine weights were determined to confirm a successful OVX (Supplemental Figure 1, B and C). The study was approved by The National Council for Animal Experimentation (ethic approval number OPR-A01–5011).

Figure 1.

OVX induces an increase in bone turnover, vertebral and femoral bone loss, and a decrease in bone Sirt1 accompanied by elevated sclerostin. A, Experiment time line. B and C, The changes in serum CTX (B), a bone resorption marker, and in P1NP, a bone formation marker (C), 1, 6, and 12 weeks after OVX or SHAM. D–I, μCT analyses in the L4 and in the distal femoral metaphysis in SHAM vs OVX mice. D, Trabecular bone volume/total volume (BV/TV). E, Trabecular number. F, Trabecular thickness. G, Femoral BV/TV. H, Femoral trabecular number. I, Femoral trabecular thickness; n = 9–10 mice/group. Statistical analysis was performed by two-way-ANOVA with time after operation and type of operation as independent variables followed by Sidak's post hoc correction vs SHAM mice. J and K, Sirt1 and sclerostin protein level in whole tibiae extracts obtained from OVX and SHAM mice 6 weeks after surgical intervention. Immunoblot of a representative image (left) and densitometry (right). HSP90 as a control, n = 3–4 bones/group and n = 6 bones/group for Sirt1 and sclerostin, respectively, analyzed by Student's t test. Results are mean ± SEM; *, P < .05; **, P < .005; ***, P < .001.

Compound

The Sirt1 activator SRT3025 was provided by Sirtris/GSK. The compound was dissolved once every 3–4 days according to the information provided by Sirtris/GSK regarding the stability of the vehicle solution, by adding 1 mL of the vehicle solution (polyethylene glycol 400:Tween 80:deionized water at a ratio of 20:0.5:79.5 [vol/vol]), to each 10.6 mg of crude compound to generate a solution concentration of 10.6 mg/mL. The dissolved compound was kept at room temperature for the whole study period.

For in vitro studies, final concentrations of 0.5μM and 1μM SRT3025 dissolved in dimethyl sulfoxide (DMSO) were used in murine long bone osteocyte-Y4 (MLO-Y4) cells. EX-527 (6-Chloro-2,3,4,9-tetrahydro-1H-Carbazole-1-carboxamide) was purchased from Sigma (Sigma-Aldrich) and dissolved in DMSO.

μCT analyses

Femora and fourth lumbar vertebra (L4) were examined by μCT (Desktop μCT 42; Scanco) 1, 6, and 12 weeks after OVX or SHAM. The scanner was operated at 70 kVp and 114 μA, with an integration time of 200 milliseconds and isotropic resolution of 10 μm. Trabecular bone was analyzed in the middle third of L4 and in the secondary spongiosa in the distal femoral metaphysis. Cortical bone was analyzed in the femoral middiaphysis.

Biomechanical testing

Femurs were subjected to the 3-point bending test. Force-displacement data was generated as previously described (24). Stiffness (N/mm) was calculated as the slope of the linear part of the force-displacement curve. Tissue modulus (E) was calculated using the Euler-Bernoulli beam theory:

$$E=\frac{F{L}^3}{48dl}$$

Where E is effective elastic modulus; F is force; L is distance between supports; d is deflection; I is cross-sectional moment of inertia of the femoral mid-diaphysis determined by μCT.

Histomorphometry

For static and dynamic histomorphometric analyses, lumbar vertebrae (L4) and femurs were embedded undecalcified in polymethylmethacrylate; 4-μm sections were prepared and evaluated a DS-Fi camera attached to an Eclipse 80i microscope (Nikon). Mineralizing surface (MS) and mineral apposition rate (MAR) as well as osteoclast number/bone perimeter were measured in the middle third of the vertebrae. The bone formation rate/bone surface (BFR/BS) was calculated: BFR/BS = MAR × MS/BS. MS was assessed as the extent of bone perimeter that exhibits either single or double labels. The MAR was assessed as the distance between the labels divided by the time interval between the 2 labels (5 d). Osteoclasts were identified by tartarte-resistant acid phosphatase staining (Sigma-Aldrich). For analysis of the cortical middiaphysis, cross-sections of the femoral midshaft were cut on a low-speed diamond wheel saw and polished. Images were acquired with the Zeiss LSM710 confocal microscope (Carl Zeiss) equipped with an Argon laser and a fluorescein isothiocyanate filter cube. To determine dynamic parameters, single-labeled calcein perimeter and double-labeled calcein perimeter were measured on both periosteal and endosteal surfaces. The MSs were then calculated for the 2 surfaces. The analysis was performed using IMAGE-PRO EXPRESS 4.0 software (Media Cybernetics).

Biochemical markers of bone turnover

Serum carboxy-terminal collagen cross-links (CTX) and amino-terminal propeptide of type I procollagen (P1NP) were measured with IDS kits (Immunodiagnostic Systems Ltd). Receptor activator of nuclear factor κ-B ligand (RANKL) was measured with a R&D kit (R&D Systems, Inc).

MLO-Y4 cell cultures

Osteocyte-like MLO-Y4 cells were originally provided by Dr Lynda Bonewald and cultured as previously described (25). Cells were plated on collagen I (Sigma-Aldrich)-coated 6-well plates at a density of 5 × 10³ cells/cm² and maintained in α-MEM, 2.5% fetal calf serum, and 2.5% newborn bovine serum. SRT3025 or the vehicle (0.05% DMSO) was added 24 hours after plating. Protein level was determined in cell lysates 3 days after treatment initiation.

Protein analysis

Proteins from whole tibia were extracted by TRIzol (Life Technologies). Proteins from MLO-Y4 cells were extracted in Laemmli buffer (2% sodium dodecyl sulfate, 10% glycerol, 5% 2-mercaptoethanol, 0.01% bromphenol blue, and 60mM Tris HCl).

Antibodies for immunoblotting were Sirt1 (Millipore), heat shock protein (HSP)90 (BD Transduction Laboratories), and Sclerostin (Abcam). Dephosphorylated β-catenin on Ser37/Thr41 (Millipore), phosphorylated on Ser33/37/Thr41 β-catenin (Cell Signaling).

Statistical analysis

Statistical analysis was performed with GraphPad Prism 6.01. Data are presented as mean ± SEM. Unpaired 2-tailed Student's t test was used to compare means of 2 groups. One-way ANOVA was used to compare means of more than 2 groups with Dunnett's post hoc correction for multiple comparisons. For normalized data comparing more than 2 groups, nonparametric one-way ANOVA was used, followed by Dunn's post hoc correction for multiple comparisons. Differences of P < .05 were considered significant.

Results

OVX reduces Sirt1 and increases sclerostin in bone

The study design is presented in Figure 1A. OVX induced a marked increase in bone turnover as indicated by elevated CTX, a marker of bone resorption, and P1NP, a bone formation marker (Figure 1, B and C). CTX peaked already at 1 week after OVX, whereas P1NP peaked at 6 weeks after OVX, and both markers reached SHAM levels 9 weeks after surgery, reflecting an early short-lived phase of accelerated bone turnover. This early increase in bone turnover led to a substantial reduction in vertebral and femoral bone volume/total volume (BV/TV) as a result of lower trabecular number and thickness (Figure 1, D–I). An age-associated decline in bone mass was also noted in SHAM mice. Importantly, a dramatic decrease in Sirt1 protein expression was observed in tibiae obtained from OVX compared with SHAM mice, accompanied by an increase in sclerostin expression 6 weeks after OVX, further supporting our previous findings of a reciprocal relationship between Sirt1 and sclerostin (Figure 1, J and K).

SRT3025 restores bone mass and structure in OVX mice

Strikingly, as evident from representative median images of L4, SRT3025 administration fully restored vertebral BV/TV, trabecular number, thickness, and connectivity, and femoral trabecular thickness lost with OVX to SHAM level (Figure 2, B and E, and Supplemental Figure 2). Moreover, treatment with SRT3025 reduced trabecular spacing, trabecular bone pattern factor, and the structure model index, indices associated with unfavorable bone architecture and mechanical strength (26) (Figure 2, F–H).

Figure 2.

SRT3025 administration restores vertebral bone mass and microarchitecture lost with OVX. A, Images of a representative L4 in the various treatment groups. Scale bar, 0.5 mm. B–H, μCT analyses of L4. B, Trabecular bone volume/total volume (BV/TV). C, Trabecular thickness. D, Trabecular number. E, Trabecular connectivity density. F, Trabecular spacing. G, Trabecular bone pattern factor (TBPf). H, Structure model index (SMI); n = 7–9 mice/group. Analyzed by one-way ANOVA followed by Dunnett's post hoc analysis vs vehicle-treated OVX mice. Results are mean ± SEM. *, P < .05; *, P < .005; ***, P < .001.

SRT3025 administration restores femoral biomechanical properties in OVX mice

Despite no change in cortical thickness (Supplemental Figure 1D), SRT3025 administration completely reversed OVX-induced deterioration in femoral biomechanical properties as evident by restoration of stiffness and the material's modulus of elasticity (Young's modulus) to SHAM level (Figure 3, A and B). These results suggest that femoral biomechanics was improved due to favorable effects on bone material properties and possibly microarchitecture and not due to geometrical changes. Of note, the Young's modulus is maybe underestimated when derived from the beam theory and analyses conducted in whole bones due to unfavorable aspect ratio and imprecise assumptions concerning the geometry (such as assuming the diaphysis to be a perfect tube), but it can be used for comparative purposes (24).

Figure 3.

SRT3025 administration restores femoral biomechanical properties and increases femoral cortical periosteal mineralizing surface and serum P1NP, a bone formation marker. A and B, Femoral biomechanical parameters determined by the 3-point bending test: A, Stiffness (Newton/millimeter); B, Young's elasticity modulus (E) measured in Giga Pascal (GPa); n = 4–5 mice/group. Analyzed by one-way ANOVA followed by Dunnett's post hoc analysis vs vehicle-treated OVX mice. C–F, Cortical midshaft femur fluorochrome-derived BFRs on the periosteal surface. C, Representative images. D, Histomorphometric analysis of MS/BS (% MS/BS). E, MAR. F, BFR/BS; n = 5–6 mice/group. Analyzed by Student's t test in vehicle vs SRT3025-treated OVX mice. Scale bar, 0.5 mm. G, Serum P1NP, a bone formation marker (H) CTX, a bone resorption marker (I) RANKL, a bone resorption marker, in vehicle- vs SRT3025-treated OVX mice 3 weeks after treatment initiation; n = 6–9 mice/group. Analyzed by Student's t test. Results are mean ± SEM. *, P < .05; **, P < .005.

To gain insight into the mechanism by which SRT3025 administration exerts these favorable effects, bone formation and resorption were evaluated upon study termination in vertebral cancellous and femoral cortical bone by static and dynamic histomorphometric analyses. No differences between treated and untreated OVX mice could be detected in cancellous bone at this time point (Supplemental Figure 3), most probably because the effects of OVX on formation and resorption were rapid, transitory, and occurred earlier, as reflected by the changes in bone turnover markers (Figure 1, B and C). However, MS on the periosteal surface in cortical bone was higher in treated compared with untreated OVX mice with a trended increase in MAR and BFR (Figure 3, D–F), suggesting stimulation of periosteal bone formation, and consistent with previous studies showing increased periosteal bone formation with sclerostin blockade (6, 7). Accordingly, the formation marker P1NP, but not the resorption markers CTX or RANKL was elevated in treated mice (Figure 3, G–I), allowing for an anabolic window.

SRT3025 down-regulates sclerostin in vivo and in vitro

Next, we asked whether sclerostin is affected by SRT3025 administration. Strikingly, tibial sclerostin was dramatically reduced in 100-mg/kg·d SRT3025-treated OVX mice as compared with vehicle-treated OVX mice (Figure 4A). A trend for reduction in sclerostin was also noted with the SRT3025 50 mg/kg·d dose (Figure 4A). Sclerostin is a product of the osteocyte, the major cellular component in bone. To test whether the effect of SRT3025 on sclerostin is cell autonomous, experiments were conducted in the murine osteocyte cell line MLO-Y4, an established model to study osteocytes in vitro (27), using SRT3025 concentrations 1.5- to 6-fold higher than detected in the plasma of treated mice (Supplemental Figure 3A). Indeed, SRT3025 significantly reduced sclerostin level (Figure 4B and Supplemental Figure 4). Conversely, treatment with EX-527, a Sirt1 inhibitor (28), elevated sclerostin (Figure 4C). Together, these results demonstrate a direct effect of Sirt1 activation or inhibition on sclerostin production by osteocyte-like cells (Figure 4, A and B). Because sclerostin mediates its deleterious effect on bone formation via inhibition of the canonical WNT signaling pathway (29), we investigated the changes in inactive and active β-catenin in SRT3025- and EX-527-treated MLO-Y4 cells. In agreement with the changes in sclerostin, the ratio between inactive phosphorylated to active dephosphorylated β-catenin was markedly decreased in SRT3025-treated MLO-Y4 cells (Figure 4D), suggesting relieving β-catenin from its constitutive proteosomal degradation and activation of the canonical WNT pathway. A reciprocal effect was observed with Sirt1 inhibition by EX-527 (Figure 4E). Taken together, this data suggests that SRT3025 down-regulates sclerostin and activates the canonical WNT pathway in osteocytes.

Figure 4.

SRT3025 down-regulates sclerostin in vivo and in vitro. A, Sclerostin level in whole tibiae extracts in SRT3025- vs vehicle-treated OVX mice. Immunoblot of a representative image (left) and densitometry (right) are presented with HSP90 as a control; n = 3–4 bones/group. Analyzed by non parametric ANOVA followed by Dunn's post hoc analysis. B, Sclerostin expression in SRT3025- vs vehicle-treated MLO-Y4 murine osteocyte-like cells. Immunoblot of a representative image (left) and densitometry (right) are presented with HSP90 as a control; n = 6 repeats analyzed by nonparametric ANOVA followed by Dunn's post hoc analysis. C, Sclerostin expression in vehicle and EX-527-treated MLO-Y4 murine osteocyte-like cells. n = 4 repeats analyzed by Student's t-test. D–E, The ratio between inactive and active β-catenin in SRT3025- and EX-527- vs vehicle-treated MLO-Y4 cells. Immunoblot of a representative image (left) and densitometry (right) of phosphorylated (Ser33/37/Thr41) β-catenin and dephosphorylated β-catenin (Ser37/Thr41). D, n = 4 repeats analyzed by nonparametric ANOVA followed by Dunn's post hoc analysis. E, n = 5 repeats analyzed by Student's t test. Results are mean ± SEM. *, P < .05 vs vehicle-treated cells.

Discussion

This study demonstrates that treatment with SRT3025 restores bone mass, structure, and biomechanical properties in OVX mice, the classic animal model of postmenopausal osteoporosis. Moreover, we show that SRT3025 down-regulates sclerostin, an inhibitor of bone formation, in vivo and in vitro, thereby activating the canonical WNT-β-catenin pathway in osteocytes. The results of the current study are consistent with our previous report of increased skeletal sclerostin in Sirt1 haplo-insufficient female mice and suggest that Sirt1 activation by SRT3025 is a potentially novel mechanism to decrease sclerostin and generate osteo-anabolic therapies for conditions of skeletal fragility while concurrently ameliorating other metabolic and age-associated pathologies.

Human and mouse data have pointed to blocking sclerostin as a promising approach to design much needed osteo-anabolic therapies. Currently available therapies for the treatment and prevention of osteoporosis are predominantly inhibitors of osteoclast-mediated bone resorption, such as the bisphosphonates and anti-RANKL antibodies. These therapies increase BMD, reduce hip fracture risk but do not build new bone, and are associated with rare significant side effects with long-term use. Intermittent recombinant PTH is the only anabolic treatment available for osteoporosis, but its bone-building effect is limited, because it stimulates bone formation and consequently bone resorption due to increased osteoblast-mediated RANKL secretion (30). In contrast, sclerostin inhibition by antibodies was shown to stimulate bone formation while repressing bone resorption via activation of the canonical WNT signaling pathway (8). Similarly, SRT3025 down-regulated sclerostin, increased P1NP, and did not increase RANKL and CTX, allowing for an anabolic window. Moreover, similar to sclerostin inhibition by antibodies and unlike PTH, which exerts its effect on endosteal surfaces (6, 31), SRT3025 administration increased outer femoral cortical periosteal MSs. Thus, our findings suggest that SRT3025 administration reduces sclerostin and mimics the beneficial effects of blocking its action by antibodies.

Growing evidence demonstrates that type 2 diabetes mellitus is an independent risk factor for skeletal fragility (32). Interestingly, sclerostin appears to play a role in diabetic bone disease, because elevated bone and serum sclerostin was reported in diabetic rats and humans, respectively (33, 34). Furthermore, sclerostin inhibition by a monoclonal antibody increases bone mass in diabetic rats (35). On the other hand, antidiabetic therapy reduced sclerostin and increased BMD (36). Whether SRT3025, currently evaluated for the treatment of metabolic conditions in humans (23), can affect diabetes-induced bone fragility by modulating sclerostin or by other mechanisms remains to be investigated and warrants evaluation of the skeletal effects of STACs in animal models of diabetes and in diabetic patients.

Finally, our finding of reduced bone Sirt1 expression with OVX is consistent with previous studies showing reduced Sirt1 in bone marrow-derived mesenchymal stem cells obtained from OVX mice (37) and in whole tibiae obtained from OVX rats (38). Regulation of Sirt1 by estrogen was suggested by demonstrating recruitment of estradiol-bound estrogen receptor α to the sirt1 promoter to increase its transcription in mammary epithelial cells in vitro (39), whereas systemic estradiol administration to OVX mice increased Sirt1 protein level in bone marrow-derived mesenchymal cells in vivo (37, 38). Whether OVX-induced decrease in Sirt1 occurs in other tissues beyond bone and contributes to the metabolic derangements associated with estrogen deficiency and menopause remains to be investigated and may shed new light on the mechanisms underlying menopause-associated changes in body weight, composition, and cardiovascular risk.

This study has a number of limitations. This was a treatment study in which therapy was initiated in OVX mice when bone turnover was already reduced to SHAM level, and histomorphometric analysis was conducted at the end of the study only. Earlier initiation of treatment and sequential histomorphometric analyses may have revealed changes in trabecular bone and more prominent effects in cortical bone. Future prevention and treatment studies in OVX- and age-induced bone loss with a broader SRT3025 dose range are warranted. In addition, because Sirt1 was shown to regulate diverse cellular pathways (40), it is not unlikely that additional targets beyond sclerostin contribute to the observed beneficial skeletal effects.

In summary, this is the first study to demonstrate the therapeutic potential of SRT3025, a third-generation STAC, currently investigated in humans, as a therapy for bone loss induced by estrogen deficiency, revealing a novel mechanism to down-regulate sclerostin and opening new avenues to explore Sirt1 activation as a means to design osteo-anabolic therapies for conditions of skeletal fragility, such as postmenopausal osteoporosis.

Abbreviations

BFR/BS

bone formation rate/bone surface

BMD

bone mineral density

BV/TV

bone volume/total volume

μCT

micro computed tomography

CTX

carboxy-terminal collagen cross-links

DMSO

dimethyl sulfoxide

EX-527

6-Chloro-2,3,4,9-tetrahydro-1H-Carbazole-1-carboxamide

HSP

heat shock protein

L4

fourth lumbar vertebra

MAR

mineral apposition rate

MLO-Y4

murine long bone osteocyte-Y4

MS

mineralizing surface

OVX

ovariectomy

P1NP

amino-terminal propeptide of type I procollagen

PTH

parathyroid hormone

RANKL

receptor activator of nuclear factor κ-B ligand

Sirt1

Sirtuin1

SHAM

sham operation

STAC

Sirt1-activating compound.