SEARCH

SEARCH BY CITATION

Keywords:

  • sex steroid receptors;
  • adaptive response of bone to mechanical loading;
  • bone formation;
  • SOST/sclerostin;
  • pulsating fluid flow

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

In female mice, estrogen receptor-alpha (ERα) mediates the anabolic response of bone to mechanical loading. Whether ERα plays a similar role in the male skeleton and to what extent androgens and androgen receptor (AR) affect this response in males remain unaddressed. Therefore, we studied the adaptive response of in vivo ulna loading in AR-ERα knockout (KO) mice and corresponding male and female single KO and wild-type (WT) littermates using dynamic histomorphometry and immunohistochemistry. Additionally, cultured bone cells from WT and AR KO mice were subjected to mechanical loading by pulsating fluid flow in the presence or absence of testosterone. In contrast with female mice, ERα inactivation in male mice had no effect on the response to loading. Interestingly, loading induced significantly more periosteal bone formation in AR KO (+320%) and AR-ERα KO mice (+256%) compared with male WT mice (+114%) and had a stronger inhibitory effect on SOST/sclerostin expression in AR KO versus WT mice. In accordance, the fluid flow-induced nitric oxide production was higher in the absence of testosterone in bone cells from WT but not AR KO mice. In conclusion, AR but not ERα activation limits the osteogenic response to loading in male mice possibly via an effect on WNT signaling. © 2010 American Society for Bone and Mineral Research


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

In our aging society, osteoporosis is an increasing public health issue in both women and men. Most often, treatment of osteoporosis only reduces bone loss in patients (antiresorptive drugs, e.g., bisphosphonates). However, there is an important medical need for bone-forming therapies capable of restoring bone loss. Despite the importance of periosteal bone formation in establishing an optimal bone structure during growth and maintaining bone quality during aging, the contribution of periosteal bone formation to bone strength often has been neglected.3 It is well known that the periosteal bone surface is very sensitive to loading as well as unloading. In fact, bone can adjust its structure to improve its resistance to the mechanical demands it experiences.4–6 Numerous studies in animals and humans, in various loading models, have evidenced that mechanical loading stimulates bone formation.7–10 Mechanical stimulation of bone therefore might offer an alternative treatment to prevent and restore bone loss, potentially inducing lifelong benefits to bone mass and strength.

Osteocytes are considered as the major bone cell that mediates mechanosensing and a key factor in regulating loading-induced bone formation and unloading-induced bone loss.11–13 Osteocytes are believed to translate mechanical signals to the bone surface, sending signals to osteoblasts and osteoclasts in order to regulate bone (re)modeling. Recently, it was demonstrated that osteocyte-ablated mice present with severe bone loss and are resistant to unloading-induced bone loss, consistent with the essential role of osteocytes in mechanotransduction.14 Likewise, sclerostin, an osteocyte-specific inhibitor of bone formation expressed by the SOST gene, is downregulated after in vivo mechanical stimulation of bone.15 Sclerostin is known to antagonize WNT signaling through binding to low-density lipoprotein (LDL)-receptor-related proteins 5 and 6.16–18 WNT signaling is now recognized as an important regulator of bone mass and bone cell function and has been linked to mechanical stimulation.15, 19, 20

Besides mechanical stimuli, sex steroids and their receptors also play a major role in bone mass acquisition and maintenance in both humans and rodents.21 In addition, androgens are well known to stimulate muscle growth as well and thereby may indirectly stimulate bone growth.22 Since androgens can stimulate both the androgen receptor (AR) and estrogen receptor-alpha (ERα), our group recently showed that both AR and ERα activation are required to optimize bone and body composition in male mice.23 Since androgen signaling interferes so strongly with bone and muscle growth, we hypothesized that androgens may influence the response of bone to mechanical loading as well. Concomitantly, sex steroids and mechanical loading may share a common signaling mechanism. In fact, earlier observations revealed that periosteal bone formation following mechanical stimulation is significantly reduced in female ERα knockout (KO) mice, indicating that ERα is required for the full osteogenic response to mechanical loading in female mice.24–26 However, a number of questions remain to be clarified: (1) Does ERα play a similar role in the loading response in male mice? (2) To what extent do androgens and AR activation affect the male adaptive response to loading?

To this end, we studied the importance of AR and/or ERα signaling in the response to in vivo ulna loading in AR-ERα KO mice and corresponding male and female single KO and wild-type (WT) control littermates using dynamic histomorphometry and immunohistochemistry. We also examined the in vitro response of AR-ablated bone cells to mechanical stimulation and testosterone administration using pulsating fluid flow.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Animals

AR KO mice were generated using Cre/loxP technology, as reported previously.27 The ERα KO animals were purchased from the Jackson Laboratory (B6.129P2-Esr1tm1Ksk/J, Bar Harbor, ME, USA). Male WT (ERα+/+AR+/Y), ERα KO (ERα−/−AR+/Y), AR KO (ERα+/+AR/Y), and double-KO (ERα−/−AR/Y) littermates, as well as female WT (ERα+/+AR+/+) and ERα KO (ERα−/−AR+/+) littermates were generated through breeding, as described previously.23 Genotyping of the AR gene was performed using polymerase chain reaction (PCR) amplification,27 and the ERα gene was analyzed following the genotyping protocol of the Jackson Laboratory. All mice were on a C57BL/6 background, were bred at the local animal housing facility (Proefdierencentrum, Leuven, Belgium), and lived in conventional conditions: 12-hour light/dark cycle, standard diet (1% calcium, 0.76% phosphate), and water ad libitum. All experimental procedures were conducted after obtaining formal approval from the ethical committee of the Katholieke Universiteit Leuven.

In vivo mechanical stimulation

The left ulnae of male double-KO (AR-ERα KO; n = 9) and corresponding ERα KO (n = 7), AR KO (n = 7), and WT (n = 9) control littermates, as well as female WT (n = 7) and ERα KO (n = 8) mice, were loaded in vivo as described elsewhere.25 Briefly, the flexed elbows and carpi of 20- to 22-week-old mice were placed in two opposing cups in a servohydraulic mechanical testing device (5848 MicroTester, Instron, Norwood, MA, USA). In order to determine the relationship between force and strain under axial compression, ex vivo strains were determined prior to in vivo loading during a single loading session using cyclic axial compressive loads. Precision strain gauges were glued to the lateral midshaft of the left ulna of one to three mice of each genotype, which had been killed immediately before the procedure (Vishay Micro-Measurements Strain Gauges, Type EA-06-015LA-120, Akron N.V., Leuven, Belgium). Compressive loading and strain were strongly correlated in all experimental groups (R2 = 0.98). The engendered strains were not different between male WT, ERα KO, AR KO, AR-ERα KO, and female WT mice and only slightly lower in female ERα KO mice. Therefore, the in vivo loading sessions comprised 40 cycles of a trapezoidal waveform with an axial compressive peak load of −2.5 N corresponding with peak strains of 1560 to 1740 microstrain. Following each load cycle, a rest period of 14.9 seconds was inserted. This in vivo loading protocol was applied to the left limbs of the mice for 10 minutes on three alternate days for 2 weeks. The right ulnae of these mice served as the nonloaded paired controls.

Mice were anesthesized for loading by an ip injection of pentobarbitone sodium (Nembutal, 50 mg/kg of body weight). All mice were injected with the fluorochrome calcein on days 3 and 12 of the experimental period. On day 15, mice were sacrificed by cervical dislocation, and the left and right ulnae were harvested.

Bone histomorphometry

Both ulnae were cleaned from surrounding tissue, immersed in Burckhardt's fixative (24 hours, 4°C), kept in ethanol, and embedded in methyl methacrylate. Cross sections of the undecalcified ulnae perpendicularly to the long axis were prepared at 200 µm thickness in the middiaphyseal region using the contact-point precision band saw (Exakt, Norderstedt, Germany). Sections were ground to a final thickness of 25 µm using a grinding system (Exakt), left unstained, and subjected to dynamic histomorphometry. Three sections in the middiaphyseal region were measured by fluorescence microscopy, and the bone formation rate per bone perimeter (BFR/B.Pm., µm2/µm/day) was assessed at both the endocortical and periosteal bone surfaces. The BFR was obtained by the product of mineral apposition rate (MAR, µm/day) and mineralizing perimeter per bone perimeter (Min.Pm./B.Pm., %). The mineralizing perimeter was calculated as follows: Min.Pm. = [dL + (sL/2)]/B.Pm., where dL represents the length of the double labels, and sL is the length of single labels along the entire endocortical or periosteal bone surfaces. The MAR (µm/day) was calculated as the mean width of double labels divided by interlabel time (9 days). All measurements were performed with a Kontron Image Analyzing Computer (KS400 3.00; Kontron Bildanalyze, Munich, Germany) and a Zeiss microscope with a drawing attachment. Specific software was developed in collaboration with the manufacturer. Histomorphometric parameters are reported according to the recommended American Society for Bone and Mineral Research nomenclature.28

SOST/sclerostin expression

Total RNA was isolated from loaded and nonloaded ulnae of male WT (n = 6) and AR KO mice (n = 7) using TRIzol reagent (Invitrogen, Merelbeke, Belgium) to determine SOST mRNA expression. Similar loading conditions were applied to the left ulnae, as described earlier, but the ulnae were harvested 24 hours after the final loading session.

In addition, loaded and nonloaded ulnae of male WT (n = 4) and AR KO mice (n = 4) were analyzed for sclerostin expression. The ulnae were fixed in 2% paraformaldehyde, decalcified in 0.5 M EDTA (pH 7.4)/PBS prior to dehydration, embedded in paraffin, and sectioned at 4 µm. Sections were taken in the middiaphyseal region. Immunohistochemical staining for sclerostin was performed using a 1:200 dilution of a biotinylated antimouse SOST antibody (R&D Systems, Oxon, United Kingdom) followed by incubation with horseradish peroxidase. Harris hematoxylin and eosin (H&E) stain was used for counterstaining. The sections were photographed using a × 40 objective with a Kontron Elektronik image-analyzing system (KS 400 3.00; Kontron Bildanalyze, Munich, Germany). The number of sclerostin-positive osteocytes (brown staining) and sclerostin-negative osteocytes (purple staining) were counted on a standardized area on each section (10 to 12 sections per mouse) using ImageJ software. The percentage sclerostin-positive cells were calculated as the number of sclerostin-positive osteocytes divided by the total number of osteocytes (positive + negative). The results were expressed as the ratio loaded over control section.

Culture of primary mouse bone cells

Cells were obtained from the limbs of adult AR KO (n = 4) and WT mice (n = 5). The long bones were harvested aseptically, the epiphyses were cut off, and the bone marrow was flushed out using a syringe and needle. The diaphyses were placed in sterile phosphate-buffered saline (PBS), chopped into small fragments, and washed extensively with PBS. Bone fragments were incubated with 2 mg/mL collagenase II (Sigma, St. Louis, MO, USA) for 2 hours at 37°C in a shaking water bath to remove all soft tissue and adhering cells from the bone chips' surface, after which the denuded bone fragments were transferred to 25 cm2 flasks (Nunc, Roskilde, Denmark). The fragments were cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco, Paisley, UK) supplemented with 100 U/mL penicillin (Sigma), 50 µg/mL streptomycin sulphate (Sigma), 50 µg/mL gentamicin (Gibco), 1.25 µg/mL fungizone (Gibco), 100 µg/mL ascorbate (Merck, Darmstadt, Germany), and 10% fetal bovine serum (FBS; Hyclone, Logan, UT, USA). Culture medium was replaced three times per week. When the cell monolayer growing from the bone fragments reached confluency, cells were harvested using 0.25% trypsin (Difco Laboratories, Detroit, MI, USA) and 0.1% EDTA (Sigma) in PBS plated at 2 × 105 cells per T75 tissue culture flask in 15 mL medium, as described earlier, until the cell layer reached confluency again. Then the cells were used for pulsating fluid flow experiments.

Pulsating fluid flow

Mouse bone cells were harvested from the T75 flasks 2 days before pulsating fluid flow (PFF) treatment and seeded onto polylysine-coated (50 µg/mL; poly-L-lysine hydrobromide, mol. wt. 15 to 30 × 104; Sigma) glass slides. The cells were plated at 2 × 105 cells per glass slide and cultured overnight in culture medium, as described earlier. The next morning, the culture medium was replaced by DMEM without phenol red (Gibco, Paisley, UK), containing either testosterone (Serva, Heidelberg, Germany) at concentrations of 10−7 or 10−9 M or vehicle and supplemented with 0.2% bovine serum albumin (BSA). The medium also contained antibiotics and ascorbate as described earlier. Similar medium, with or without testosterone, was used the next day during the PFF experiments. PFF was generated by pumping 13 mL of culture medium, using a roller pump, through a parallel-plate flow chamber containing the bone cells. The cells were subjected to PFF with a mean shear stress of 0.6 Pa, a pulse amplitude of 0.3 Pa, and a pulse frequency of 5 Hz. Stationary control cultures were kept in a petri dish under similar conditions as the experimental cultures, i.e., at 37°C in a humidified atmosphere of 5% CO2 in air. After 5, 10, 15, 30, and 60 minutes of PFF or static culture, the medium was collected and assayed for NO production. After 1 hour, PFF treatment was terminated, and cells were harvested directly in TRIzol reagent (Gibco) for isolation of total RNA and DNA and determination of COX-2 mRNA. The cells derived from one donor mouse were subjected to the different treatments (PFF or static culture, vehicle or testosterone treatment) in one experiment. The entire setup subsequently was repeated for each AR KO (n = 4) and WT (n = 5) donor mouse.

Nitric oxide

NO release was measured as nitrite (NOmath image) accumulation in the conditioned medium using Griess reagent29 consisting of 1% sulfanilamide, 0.1% naphtylethelene diamine dihydrochloride, and 2.5 M H3PO4. Serial dilutions of NaNO2 in nonconditioned medium were used as standard curve. Absorbance was measured at 540 nm.

cDNA synthesis and real-time quantitative PCR

cDNA was synthesized from 1 µg total RNA using reverse-transcriptase SuperScript II RT (Invitrogen, Merelbeke, Belgium). Real-time qRT-PCR was performed on an ABI Prism 7500 Fast Real-Time PCR System (Applied Biosystems). SOST and COX-2 mRNA expression was analyzed by TaqMan Gene Expression Assays (ID Mm00478374_m1 and ID Mm00478374_m1, respectively; Applied Biosystems). Expression levels were calculated using the comparative CT method (ΔΔCT) and were normalized for hypoxanthine guanine phosphoribosyl transferase (HPRT) expression.

Statistical analysis

Statistical analysis of the data was performed using NCSS software (NCSS, Kaysville, UT, USA). Log transformation of the data was performed—if appropriate—to obtain a normal distribution. Comparisons between treated groups and their respective controls were made using paired t tests. Differences between treatment groups or between genotypes were assessed following one-way analysis of variance followed by Fisher's least-significant-difference multiple-comparison test. Data are represented as means ± SE, and p < .05 was accepted as significant.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Androgen receptor disruption increases the mechanical sensitivity of the skeleton

At baseline, no significant differences in periosteal bone formation rate (PsBFR) were observed in the nonloaded control ulnae of male WT, AR KO, ERα KO, and AR-ERα KO mice (Fig. 1A). In vivo mechanical loading of the ulnae significantly increased the PsBFR in all male genotypes compared with the respective nonloaded ulnae (see Fig. 1A, B). However, the relative loading-induced increase in PsBFR, expressed as the ratio loaded over control ulna, was significantly higher in AR KO and AR-ERα KO mice compared with WT mice (+320% and +256% versus +114% in WT mice; see Fig. 1B). ERα disruption did not affect the AR-related increase in PsBFR following loading because the increase in PsBFR following loading was similar in AR KO and AR-ERα KO mice (see Fig. 1B). In parallel, the loading-induced increase in AR-ERα KO mice was significantly higher than in ERα KO mice (+78%). Accordingly, the concept that ERα inactivation in male mice had no apparent effect on the response to loading was further strengthened by the finding of a similar increase in PsBFR following loading in male ERα KO compared with WT mice (+100% versus +114% in WT mice; see Fig. 1B). In contrast, in female mice, a significant increase in PsBFR following mechanical stimulation occurred only in WT and not in ERα KO mice, even though the PsBFR at baseline was not different between genotypes (see Fig. 1A, B). Concomitantly, ERα inactivation in female mice prevented the loading-induced increase of the PsBFR (see Fig. 1B). Bone formation on the endocortical surfaces of the ulnae was not significantly affected by mechanical loading in any of the groups (data not shown).

thumbnail image

Figure 1. Load-induced changes in periosteal bone formation. (A) Periosteal bone formation rate per bone perimeter (PsBFR/B.Pm.) in loaded and nonloaded control ulnae of male WT, ERα KO, AR KO, and AR-ERα KO mice and in female WT and ERα KO mice. (B) The relative fold increase in Ps.BFR/B.Pm. following loading was calculated as the ratio loaded over nonloaded control ulna for each animal (n = 7 to 9 mice/group). Values are expressed as means ± SE. ap < .05 represents a significant increase following loading, as assessed by paired t-test; bp < .05 versus male WT; cp < .05 versus female WT.

Download figure to PowerPoint

The increase in PsBFR following loading was accompanied by a significant decrease in SOST expression in male WT mice (−26.1% ± 8.9%) and even more in AR KO mice (−45.5% ± 4.7%, p = .08 versus WT; see Fig. 2A). Likewise, sclerostin expression, as assessed by the percentage of sclerostin-positive osteocytes in the cortex, was significantly lowered following loading in AR KO mice (−11.7% ± 1.5%, see Fig. 2B, C) but not in male WT mice (see Fig. 2B, C).

thumbnail image

Figure 2. Load-induced changes in SOST/sclerostin expression. The relative change in (A) SOST and (B) sclerostin expression following ulna loading in WT (white bars) and AR KO mice (black bars) expressed as the ratio loaded over nonloaded control ulna (%) (n = 4 to 7 mice/group). (C) Sclerostin immunohistochemistry. Representative histologic midshaft sections of WT nonloaded control (upper left) and loaded (lower left) ulnae and AR KO nonloaded control (upper right) and loaded (lower right) ulnae (n = 4 mice/group). The black arrows demonstrate sclerostin-positive osteocytes, whereas the red arrows show sclerostin-negative osteocytes. Values are expressed as means ± SE. ap < .05 represents a significant load-induced change; bp < .05 versus male WT.

Download figure to PowerPoint

Testosterone inhibits the in vitro response of bone cells to mechanical loading

Application of PFF rapidly stimulated NO production by WT cells during 1 hour of stimulation (Fig. 3A, B, for time point 10 minutes; other time points are not shown). The effect of testosterone administration on PFF-induced NO release by WT cells was most pronounced after 5 and 10 minutes of stimulation (see Fig. 3A, B, for time point 10 minutes). For instance, testosterone was able to inhibit the PFF-induced increase in NO production by WT cells after 10 minutes of stimulation in a dose-dependent manner (see Fig. 3A, B). In fact, administration of 10−7 M testosterone abolished the PFF-induced increase in NO at 10 minutes of treatment (see Fig. 3A, B). This inhibitory effect of testosterone appeared transient, however, because the NO levels in PFF-stimulated cells were similar in WT cells treated with testosterone and vehicle after 15 minutes of mechanical stimulation (data not shown). Testosterone had no effect on NO production by WT cells grown under static control conditions (see Fig. 3A, for time point 10 minutes). In AR KO cells, PFF treatment increased NO production to a similar extent as in WT cells (see Fig. 3A, B, for time point 10 minutes; other time points are not shown). As expected, administration of testosterone had no effect on the release of NO by AR KO cells following PFF or on the static control cultures (see Fig 3A, B, for time point 10 minutes). In line with the increased NO production, application of PFF also increased COX-2 expression to a similar extent in WT (+647%) and AR KO cells (+569%, p = .5 versus WT cells treated with vehicle; Table 1) after 1 hour of stimulation. In general, testosterone administration (10−7 M or 10−9 M) did not significantly affect COX-2 expression in static or PFF conditions in either WT or AR KO cells, although administration of 10−9 M testosterone significantly decreased COX-2 expression in static AR KO cells (see Table 1). Along the same line, testosterone did not affect the PFF-induced increase in COX-2 expression in WT or AR KO cells (data not shown).

thumbnail image

Figure 3. (A) Nitric oxide (NO) production by primary bone cells of either WT (n = 5) or AR KO mice (n = 4) treated with vehicle (VEH), 10−9 M testosterone (10−9 M T) or 10−7 M testosterone (10−7 M T) after 10 minutes of pulsating fluid flow (PFF) treatment or static control culture (CONTROL). (B) The relative fold increase in NO production following 10 minutes of PFF was calculated as the ratio PFF over static control culture for each culture condition. Cells from one donor mouse were subjected to the different treatments in one experiment [static or PFF, testosterone (10−7 M, 10−9 M), or vehicle]. This setup was repeated for each donor [AR KO (n = 4) or WT mice (n = 5)]. Values are expressed as means ± SE. ap < .05 versus respective static control culture, as assessed by paired t-test; bp < .05 versus VEH-treated cells.

Download figure to PowerPoint

Table 1. COX-2 Expression
 WTAR KO
VEH10−9 M T10−7 M TVEH10−9 M T10−7 M T
  • Note: Relative COX-2 expression in WT and AR KO primary bone cells following 1 hour of pulsating fluid flow (PFF) treatment or static control culture (Static). WT and AR KO cells were treated with either vehicle (VEH) or testo sterone (T) at concentrations of 10−9 M or 10−7 M. Expression levels were normalized for hypoxanthine guanine phosphoribosyl transferase (HPRT) expression Values are expressed as mean ± SE (n = 3 to 5 mice/group).

  • *

    p < .05 versus respective static condition;

  • **

    p < .05 versus respective VEH-treated condition.

Static0.42 ± 0.170.20 ± 0.060.27 ± 0.130.29 ± 0.060.10 ± 0.01**0.14 ± 0.06
PFF3.14 ± 1.72*1.34 ± 0.25*1.52 ± 0.49*1.94 ± 0.63*1.66 ± 0.66*2.20 ± 0.58*

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

The primary aim of this study was to evaluate the extent to which sex steroid receptors AR and ERα affect the sensitivity of the male skeleton to mechanical loading. Previous studies have shown that ERα activation enhances the osteogenic response to loading in female mice.24, 25 However, the role of androgens and AR activation, as well as the role of ERα, in the response to mechanical loading in male mice remains to be elucidated. According to our data, AR disruption increases the mechanical sensitivity of the male skeleton, with increased periosteal bone formation in response to loading in AR KO and AR-ERα KO mice. In accordance, a recent study in male tennis players indicates that the exercise-induced increase in periosteal expansion is most pronounced in pre- and peripubertal boys and tends to plateau in postpubertal boys,30 consistent with the concept that bones might become less sensitive to loading toward the end of puberty, when androgen concentrations are high. In mice after orchidectomy, mechanical loading also preserves bone mass,31 suggesting that loss of androgens indeed does not limit the anabolic response of bone to mechanical loading in vivo. Yet our finding that AR activation limits the response to mechanical loading is unexpected because androgens and AR activation are known to optimize bone and muscle acquisition in both humans and rodents.23, 32–34 In this regard, our results imply that the well-established anabolic role of AR activation cannot be explained by changes in the mechanical sensitivity of the skeleton. Our data suggest that mechanical loading might offer a low-cost therapeutic perspective to prevent and even restore androgen deficiency related bone loss.

Our study also addressed the role of ERα in the mechanical bone response to loading. In line with earlier observations,24, 25 our data indicate that female mice require a functional ERα to optimize the osteogenic response to loading because female ERα KO mice show no significant increase in periosteal bone formation following loading. Considering the lack of anabolic response to loading in female ERα KO mice, we investigated whether ERα also affects the response of bone to mechanical loading in male mice. In contrast with female mice, ERα appears to have no effect on the response of bone to mechanical loading in male mice, with male ERα KO and male WT mice showing a similar loading-induced increase in periosteal bone formation. At the same time, ERα inactivation has no effect on the AR-related increase in mechanical sensitivity in male AR-ERα KO mice. Such a discrepancy in the role of estrogen receptors between male and female mice has been shown previously for the ERβ, which appears to be a negative modulator of the periosteal response to mechanical loading in female mice but not in male mice.35, 36 In addition, the interaction between estrogens and mechanical stimuli might be species-specific because estradiol appears to enhance the effects of physical activity on bone in male siblings.37 Overall, these findings demonstrate that the role of ERα in mechanotransduction in mice is gender-dependent.

Because it is well known that SOST/sclerostin is downregulated following in vivo mechanical loading15 and in vivo loading has differential effects on bone mass in WT and AR KO mice, we suggested that SOST/sclerostin expression may be differentially affected by mechanical loading in AR KO and WT mice. Consistent with these earlier reports, SOST/sclerostin signaling is downregulated following in vivo stimulation in our study. In line with the increase in periosteal bone formation following in vivo mechanical stimulation, loading significantly reduces SOST expression in male WT mice and even more so in AR KO mice. Accordingly, sclerostin expression is also decreased after mechanical loading in AR KO mice but not in male WT mice. These results indicate that deletion of AR signaling limits the inhibiting effect of SOST/sclerostin expression following loading. The finding of an interaction between androgen signaling and SOST/sclerostin therefore may explain—at least partly—the increase in mechanical sensitivity of the AR KO skeleton. While previous reports have associated ERα signaling to WNT/β-catenin signaling in the adaptive response to loading in female mice,38 our results suggest that androgen signaling in response to in vivo stimulation may affect WNT signaling through modulation of SOST/sclerostin expression in male mice. However, a potential contribution of ERα activation to SOST/sclerostin signaling in males and females cannot be fully excluded by our data because the loading-induced changes in SOST and sclerostin expression levels were not assessed in ERα KO and AR-ERα KO mice.

The AR ablation-related increase in mechanical sensitivity of the male skeleton also may result from an altered response of mechanosensitive bone cells to mechanical loading. Human bones are exposed to high-frequency (30-Hz), low amplitude strains, for example, owing to muscle activity, but also to low-frequency (1- to 9-Hz), high-amplitude strains (up to 3000 microstrain) strains owing to, for example, running.39 The fluid shear stress—likely the mechanical stimulus for osteocytes—resulting from these daily strains have been determined theoretically and range from 0.8 to 3 Pa.40 The in vitro mechanical stimulus for the bone cells thus closely resembled the magnitude of in vivo mechanical loading parameters. In order to establish whether mechanosensitive bone cells respond differently to mechanical loading in the presence or absence of testosterone, we not only investigated the in vivo response to mechanical loading but also the in vitro response to mechanical stimulation. We therefore isolated primary bone cells from the long bones of AR KO and WT mice and subjected these cells to PFF and measured the cellular NO production and COX-2 mRNA expression. Previous studies have highlighted the importance of NO production and COX-2 expression for the in vitro response to fluid flow.41–43 NO is a short-lived, highly reactive free radical involved in bone metabolism and in the early anabolic response of bones to mechanical loading.44–46 In addition, prostaglandins are known to play an essential role in the mechanical response to loading as well.42, 47 In this regard, COX-2—a key enzyme in prostaglandin production48—has been shown to play a crucial role in mechanotransduction in bone cells following fluid flow.49 According to our data, fluid flow stimulates NO and COX-2 expression to a similar extent in male WT and AR KO cells treated with vehicle. Administration of testosterone dose dependently inhibits the early release of NO in WT cells but does not influence the fluid flow-induced changes in COX-2 expression in WT cells. The latter effect would be consistent with the transient inhibitory effect of testosterone on NO secretion. As expected, testosterone has no effect on the loading-induced NO or COX-2 production by AR KO cells. These findings therefore suggest that the testosterone-dependent inhibitory effect on NO production in WT cells is mediated through AR activation. In the presence of testosterone, our in vitro results thus mimic our in vivo results, where the AR KO mice showed a stronger anabolic response to mechanical loading than male WT mice.

In conclusion, this study clarifies the role of AR and ERα signaling in the mechanical sensitivity of the skeleton in male mice. AR activation limits the periosteal bone response to in vivo mechanical loading, whereas testosterone administration and subsequent AR activation block the in vitro fluid flow-induced NO production. In addition, AR signaling following mechanical loading appears to be associated with changes in SOST/sclerostin signaling. In contrast with female mice, ERα activation in male mice does not interfere with the adaptive response to loading. These findings provide important new insights in the interaction between androgen action and mechanical stimulation and confirm the role of sclerostin and mechanical stimulation as important modulators of bone formation.

Disclosures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

Steven Boonen and Dirk Vanderschueren are senior clinical researchers of the Fund for Scientific Research Flanders. Steven Boonen is holder of the Leuven University Chair in Gerontology and Geriatrics, supported by Novartis. All other authors have no conflicts of interest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References

We would like to thank K Moermans, R Van Looveren, E Van Herck, and J de Blieck-Hogervorst for excellent technical assistance. We also thank P Pazevic for the help with the sclerostin immunohistochemistry. This study was supported by Grant OT/05/53 from the Katholieke Universiteit Leuven and Grant G.0323.09 from the Fund for Scientific Research Flanders, Belgium (FWO Vlaanderen). Astrid Bakker was supported by a grant of the research institute MOVE of the VU University.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Disclosures
  8. Acknowledgements
  9. References