Exercise Improves Load Bearing Bone Structural Properties in Female SPARC^(−/−) Mice but not in Males

Abstract

Secreted protein acidic and rich in cysteine (SPARC) is the most abundant glycoprotein in bone and is thought to play a critical role in bone remodeling and homeostasis. However, the effect of SPARC in relation to gender and exercise on bone quality is not well understood. The purpose of this study was to quantify differences in the structural and biomechanical properties between calvarial and femoral bone from male and female wild-type (WT) and SPARC null (SPARC^(−/−)) mice as well as the ability of exercise to rescue bone health. Male and female WT and transgenic SPARC^(−/−) mice were given either a fixed or rotating running wheel for exercise. Bone structural, biomechanical, and morphological parameters were quantified using micro computed tomography, push out testing for the calvaria, three-point flexural testing for the femurs, and histological staining.

Similar reductions in structural and biomechanical strength were observed in both male and female SPARC^(−/−) calvaria, most of which were not significantly affected by exercise. In femurs, SPARC^(−/−) had a significant effect on structural parameters in both sexes, but was more pronounced in females with some properties being rescued with running. Interestingly, the effect of SPARC^(−/−) on bone mineral density was only detected in female SPARC^(−/−) mice, not males, and was subsequently rescued with exercise.

This study emphasizes the differences between sexes in WT and SPARC^(−/−) mice in regard to structural parameters and biomechanical properties. Research into gender differences can help inform and personalize treatment options in order to more accurately meet patient needs.

INTRODUCTION

Secreted protein acidic and rich in cysteine (SPARC), also known as Osteonectin, is the most abundant glycoprotein in bone and is expressed in a variety of other tissues undergoing active remodeling. SPARC is considered a “matricellular protein” as it mediates interactions between cells and the extracellular matrix (ECM). In bone, SPARC is expressed by osteoblasts and is thought to play a critical role in bone remodeling and homeostasis, although the exact mechanism is not clear. In vitro studies demonstrate that SPARC binds to collagen and hydroxyapatite and can regulate cell proliferation.¹

In vivo analyses have primarily been facilitated through mice with a homozygous null (−/−) mutation in the SPARC gene. SPARC^(−/−)- mice develop osteopenia, have reduced numbers of osteoblasts and osteoclasts, and a corresponding decreased rate of bone formation.^1–3 The role of SPARC in non-skeletal tissues has also been evaluated using the SPARC^(−/−) murine model. For instance, SPARC^(−/−) mice demonstrate early onset cataracts,⁴ reduced fibrillar collagen in connective tissues of the heart,⁵ and altered tumor organization and growth.^{6; 7} While the SPARC^(−/−) murine model has been used for over 30 years, the majority (if not all) of the analyses on SPARC^(−/−) mice did not control for sex or even specify which sex was used in the analysis. While this is likely a consequence of grouping sexes to increase effect size and power, more and more research demonstrate the importance of sex as a variable that can affect health, illness, and treatment efficacy.⁸

Estrogens are known to intervene in many non-reproductive processes, including mineral homeostasis.^{9; 10} Estrogen (E2) protects against bone loss by stimulating osteoclast apoptosis and suppressing osteoblast apoptosis. Estrogen deficiency is therefore associated with an increase in the lifespan of osteoclasts and a concomitant decrease in osteoblast lifespan.¹⁰ SPARC^(−/−) mice and ovariectomized (OVX) mice both demonstrate bone loss, deterioration of bone microarchitecture, and fragility similar to postmenopausal women.^{1; 9; 10}

Previous studies demonstrate that physical exercise has the potential to promote bone formation, thereby improving bone mass and strength.^11–13 Exercise programs have also been shown to effectively reduce postmenopausal bone loss.¹⁴ While the molecular mechanism of exercise on bone is not fully understood, previous studies suggest it is likely a combination of mechanical stimuli and the induction of cell signaling pathways, hormones, and cytokines that produce these beneficial outcomes.^15–18 SPARC has been characterized as an exercise-induced gene, although most of the research surrounding its expression have been on muscle tissue.¹⁹ Thus, further research is required to examine the effect of exercise on bone in the absence of SPARC and whether there are sex-related differences in these effects.

The objective of this research was to quantify differences in the structural and biomechanical properties between calvaria and femoral bone from male (M) and female (F) wild-type (WT) and SPARC^(−/−) mice. We also examined whether voluntary wheel running can rescue bone health in these bones and if there were significant differences in bone quality between the two sexes.

METHODS

Animal Model

All in vivo animal protocols were approved by our Institutional Animal Care Committee (McGill University, AUP #2007–5405) and adhered to the ethical guidelines of the Canadian Council on Animal Care. All approvals were obtained prior to starting any experiments for this study. 40 wild type mice (WT) and 43 SPARC^(−/−) 7–9 month old transgenic mice (21.0 – 49.0g) were used in this study. Animals were bred in-house at McGill University as previously described²⁰. In brief, SPARC-null mice (gift of Dr. Helene Sage) were developed on a mixed C57BL/6 × 129SVJ background²¹ and backcrossed onto a standard C57BL/6 background for >12 generations; they are therefore considered to be fully congenic. The animals were housed with corncob bedding (7097, Teklad Corncob Bedding, Envigo, UK) and cotton nesting squares for enrichment in same-sex groups of 3–5 per pathogen-free ventilated polycarbonate cage (Allentown, Inc, Allentown, NJ, USA). They had unrestricted access to food (2092X Global Soy Protein-Free Extruded Rodent Diet, Irradiated) and water and were on a 12-hour light-dark cycle. Male and female mice were randomly assigned to either an exercise (RUN) or sedentary (SED) group (n>=12). Animals assigned to the RUN group were placed into a cage with a circular plastic running wheel and igloo hut, Fig. 1A (Bio-Serv, Flemington, NJ). Animals assigned to the SED group received an identical running wheel and igloo, however, the wheel was secured in place to prevent rotation. Increased voluntary running was confirmed via counting wheels with all RUN groups demonstrating maintained or increased voluntary running except F-SPARC^(−/−)-RUN.²² After 6 months, mice underwent biobehavioral testing to assess for signs of disability^22–24, euthanized by decapitation under isoflurane anesthesia, and the calvaria and femurs explanted. Tissue was wrapped in saline soaked gauze and sent to Cedars-Sinai Medical Center for μCT, biomechanical, and histological analyzes, Fig. 1B–1C. Samples which were broken were excluded from analysis.

Fig. 1: Experimental design and methods authentication.

A) Experimental design and groups tested. B) Quantification of voluntary running in both mice strains in response to availability of the running wheel; *p<0.05, ***p<0.001. C) Correlation between the biomechanical test (ultimate load and μCT structural parameter (trabecular thickness).

Micro Computed Tomography Analysis.

Ex-vivo micro computed tomography (μCT) scans were used to quantify the bone mineral density (BMD) and structural parameters of the calvaria (Viva CT 40, Scanco Medical, Wangen-Brüttisellen, Switzerland) and femurs (Quantum GX2, PerkinElmer, Waltham, Massachusetts, United States) of each specimen.²⁵ The scanners were set at 55kVp and 145mA, with an integration time of 200ms and a voxel size of 35μm. The entire femur and a 5mm circular region of interest on the calvaria was scanned. Imaging analyses were performed using Scanco uCT analysis software for the calvaria and Analyze14.0 evaluation software (AnalyzeDirect, Overland Park, KS, USA) for the femurs to contour the region of interest and quantify the average architectural parameters including cortical thickness, cross-sectional area, bone volume, total volume, bone volume fraction (BV/TV), and mean density for both the calvaria and the femur.

Biomechanical Testing

Three-point flexural bending was performed to determine strength and stiffness of the left and right femur for each specimen. The femurs were explanted with extraneous soft tissue and muscle dissected from the bone. The femur was mounted onto a three-point bending apparatus attached to the frame of a servo-hydraulic actuator (MTS Bionix 370.02, MTS Corp., Eden Prairie, MN) equipped with a mini load cell (MINI45 Transducer, API Corp., Apex, NC). The posterior side of the femur rested on two 5mm diameter steel roller contact points spanning 20mm. The actuator loaded the mid-shaft of the femur via displacement control at a constant rate of 3mm/min until failure.

Calvaria specimens were also mechanically tested to failure. Each calvaria was placed onto customize-made, calvaria-shaped, 3D-printed plates with a 6mm diameter clearance hole for biomechanical push-out testing. Plates were aligned on the MTS machine and a 4mm diameter pin was lowered at a rate of 0.03mm/sec while load and displacement data were continuously recorded.

The load-displacement curves of each femur and calvarium was used to calculate the mechanical properties of each calvaria, including ultimate load, stiffness, and toughness. Stiffness was calculated as the slope of the elastic region of the measured load-displacement curves. Strength was measured as the maximum load applied before failure. Young’s modulus (E) of the femur was calculated using the beam bending equation: E = (P/S)(L³/48 x I), where P is the applied load, S is the deflection, L is the length of the femur, and I is the moment of inertia.

Histology

After sacrifice, the calvaria was explanted, fixed in a 4% formaldehyde solution, passed through a graded series of ethanol solutions, and embedded in paraffin. Five-micron-thick sections were cut from the paraffin blocks and stained with hematoxylin and eosin (H&E) to evaluate the morphological features of the tissue. Tartrate-resistant acid phosphatase (TRAP) staining was also performed according to manufacturer protocol to quantify osteoblast activity in the calvaria (Sigma Aldrich, St. Louis, MO). Images were captured using a Leica Aperio AT2 slide scanner (Leica Biosystems, Deer Park, IL).

For immunofluorescent staining, tissues were deparaffinized, and the antigens were retrieved by incubation in Proteinase K (Agilent, Carpinteria, CA) for 20 minutes at room temperature. Nonspecific antigens were blocked by applying protein blocking solution (Agilent). Slides were stained with a 10ug/mL concentration of goat anti-mouse SPARC antibody in PBS (R&D Systems, Minneapolis, MN). The primary antibodies were applied to the slides, after which the slides were incubated at 4°C overnight and washed using PBS; the slides were then incubated with secondary antibodies for one hour at room temperature. Finally, the slides were stained with DAPI for five minutes in the dark. ProLongTM Gold Antifade mounting medium (Invitrogen) was applied to the tissue. Images were captured using an Axio Imager Z1 fluorescent microscope (Carl Zeiss, Oberkochen, Germany) equipped with ApoTome and AxioCam HRc cameras. Negative controls were processed using identical protocols while omitting the primary antibody to exclude nonspecific staining. Images were captured with 4×4 tile scans at 20x objective.

Statistical Analysis

Pearson’s correlation test was used to determine the relationship between structural characteristics and biomechanical properties. One-way ANOVA with Tukey’s multiple comparison was used to determine differences between groups with significance set at p<0.05.

RESULTS

Effect of SPARC, Sex and Exercise on Bone Structure Measured using Micro Computed Tomography

Calvaria.

M-SPARC^(−/−) calvaria had significantly greater BMD compared to male M-WT, p<0.01, but there were no significant differences between F-SPARC^(−/−) and F-WT calvaria, Fig 2A. Exercise did not significantly affect the bone mineral density (BMD, HA/cm³) of male or female calvaria. BV was significant greater in WT and bone surface to bone volume ratio (BS/BV) was significantly lower in WT mice compared to SPARC^(−/−) in both sexes, Fig 2B–2C, p<0.0001. Exercise did not significantly affect the connective density of male calvaria, and M-WT calvaria had significantly greater connective density compared to M-SPARC^(−/−) calvaria, Fig. 2D. Conversely, the connective density of female calvaria were significantly greater in the F-SPARC^(−/−)-SED group compared to both F-WT-RUN and F-WT-SED, but F-SPARC^(−/−)-RUN were not significantly different than either of the two WT groups. Trabecular thickness of F-WT and M-WT calvaria were significantly greater compared to both SPARC^(−/−)-RUN and SPARC^(−/−)-SED groups, Fig. 2E.

Fig. 2. μCT analysis of calvarial bone demonstrating effect of SPARC deficiency on structural properties of male and female mice.

(A) BMD, (B) Bone Volume, (C) BS/BV, (D) Connective Density, and (E) Trabecular Thickness. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Femur.

There were no significant differences in BMD between femurs extracted from M-WT and M- SPARC^(−/−) mice, regardless of exercise status. Femurs extracted from F-WT-RUN and F-WT-SED mice had significantly greater BMD compared to SPARC^(−/−)-SED mice, Fig. 3A - 3B, p<0.01. Exercise effectively rescued female BMD as there were no significant differences between F-SPARC^(−/−)-RUN, F-WT-SED, and F-WT-RUN groups, p<0.01. F-SPARC^(−/−)-SED demonstrated significantly reduced BMD compared to F-SPARC^(−/−)-RUN femurs, p<0.01. Bone volume (BV, mm³) was significantly greater in M-WT and M-SED femurs compared to M-SPARC^(−/−)-SED femurs, Fig. 3C, p<0.01. M-SPARC^(−/−)-RUN femurs had significantly reduced BV compared to M-WT-SED mice, p<0.01. The only significant difference in BV among female femurs was between SPARC^(−/−)-SED and WT-RUN groups, p<0.001. Bone surface area to bone volume ratio (BS/BV) was significantly greater in M-SPARC^(−/−)-SED femurs compared to M-WT-RUN femurs, p<0.01, with no other significant differences observed among male femurs, Fig. 3D. F-SPARC^(−/−)-SED femurs had significantly greater BS/BV compared to WT-RUN, WT-SED, and SPARC^(−/−)-RUN, p<0.01.

Fig. 3. μCT analysis shows strong effect of SPARC deficiency on femoral bone density in females, but not in males, and these deficiencies can be rescued by exercise.

(A) Cortex Mean BMD, (B) Trabecular Mean BMD, (C) Bone Volume, (D) BS/BV, and (E) Connective Density. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Effect of SPARC, Sex and Exercise on Calvaria and Femur Bones Biomechanical Properties Measured with Load to Failure Test

Calvaria.

Push out load to failure and toughness were significantly reduced in both male and female SPARC^(−/−) calvaria compared to WT calvaria (p<0.01), with no significant differences between the SED and RUN exercise groups, Fig. 4A – 4B. Stiffness was also significantly reduced in both M-SPARC^(−/−)-SED and F-SPARC^(−/−)-SED calvaria, but there were no significant differences observed between SPARC^(−/−)-RUN and WT calvaria in either of the two sexes, Fig. 4C.

Fig. 4. Biomechanical tests of calvarial bones harvested from mice in different groups show significant effect of SPARC deficiency on bone quality. Exercise rescues bone stiffness in both sexes.

(A) Max load, (B) Toughness, and (C) Stiffness. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Femur.

Ultimate load at failure of F-SPARC^(−/−)-SED femurs was significantly lower than femurs in the F-SPARC^(−/−)-RUN females and marginally lower than F-WT-SED femurs, p<0.001 and p=0.053 respectively, Fig. 5A. There were no differences in ultimate load at failure between femurs in any of the male groups. Running significantly increased the toughness of F-SPARC^(−/−) femurs, p<0.05, but did not affect the toughness of M-SPARC^(−/−) femurs, Fig. 5B. The stiffness and Young’s modulus of female femurs were not significantly different between any of the groups, Fig. 5C. Conversely, M-SPARC^(−/−)-SED femurs were significantly stiffer and had a greater Young’s modulus than M-WT-SED femurs, p<0.05 and p<0.001 respectively. Running further increased the stiffness of M-SPARC^(−/−) femurs, p<0.01, but did not significantly affect the Young’s modulus.

Fig. 5: Biomechanics testing of the femoral bone demonstrates a greater effect of SPARC on female compared to males, with some parameters being rescued with exercise.

(A) Max load, (B) Toughness, and (C) Stiffness. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

Effect of SPARC, Sex and Exercise on Morphological Features on Calvaria and Femur

Histological staining with H&E shows no significant structural and morphological differences between sexes and between experimental groups Fig. 6. TRAP staining for osteoclasts also shows no significant differences between sexes and between experimental groups (Sup. Fig. 1).

Fig. 6.

Histological analysis of the calvarial bone show no significant structural or morphological differences.

IF staining for SPARC confirmed there was significantly less SPARC expression in SPARC^(−/−) calvaria. No obvious differences between sex or physical activity status was observed.

The mouse calvarial bone is relatively thin, constructed mostly from two cortical layers with minimal trabeculae and bone marrow compartments in between. As a result, the thickness of both compartments varies depending on the section and location and therefore morphological differences are best recapitulated using μCT imaging.

DISCUSSION

Data presented in this study demonstrate the effect of exercise and sex on calvaria and femoral bone structure and function in SPARC^(−/−) and wild type C57BL mice. Calvaria from both male and female SPARC^(−/−) mice had similar reductions in μCT measures of bone quality and biomechanical strength besides BMD that was surprisingly elevated in the SPARC^(−/−) mice, most of which were not significantly affected by exercise. Although, exercise did increase the compromised stiffness observed in male and female calvaria from SPARC^(−/−) mice. This suggests the effect of SPARC^(−/−) was greater than voluntary running on calvaria bone structure and function.

In load bearing femurs, the effect of SPARC^(−/−) was detected in the structural parameters of both sexes, but was more noticeable in females with some of these properties rescued with running in F-SPARC^(−/−) mice. Interestingly, the effect of SPARC^(−/−) on bone mineral density was only detected in female SPARC^(−/−) mice, not males, and was subsequently rescued with exercise. Further research is required to better understand the mechanism behind these differences.

Calvaria and femoral bone samples use in the present study were harvest from in vivo studies on SPARC^(−/−) and WT mice that evaluated whether exercise effectively attenuates low back pain.^{22; 23; 26} Biobehavioral tests and analyses of intervertebral disc (IVD) tissue were used in these previous studies to demonstrate that SPARC(−/−) mice develop back pain that closely resembles the degenerative and physiological characteristics of discogenic back pain in humans.^{8; 27; 28} Once validated, the authors used this SPARC(−/−) mouse model to evaluate pharmaceutical and physical treatments to reduce chronic low back pain.^29–31 Their recent work on running exercise concluded that it was an effective means to reduce behavioral measures of pain. Notably, the group also found mechanistic and epigenetic differences between male and female mice, suggesting sex-specific strategies to maintain IVD extracellular matrix homeostasis. Similarly, our work suggests subtle yet distinct sex differences in the regulation of bone homeostasis.

In addition to back pain, SPARC^(−/−) mice exhibit other age-induced changes, including osteopenia, decreased muscle mass and strength, and early onset cataractogenesis.^{1; 4; 32; 33} Dalaney et al. first characterized the osteopenic phenotype of SPARC^(−/−) mice using histomorphometric, radiographic (x-ray), and biomechanical analyses.¹ They revealed that SPARC^(−/−)- mice have decreased osteoblasts and osteoclasts, decreased trabecular bone volume and mass, and reduced mechanical strength and stiffness, which aligns with the present study. While Dalaney et al. also evaluated both male and female mice, they did not evaluate these groups separately and therefore did not report any sex-related variations in their outcomes. Clinically, decreased synthesis of SPARC have been reported in osteoblasts from patients with osteogenesis imperfecta.^34–36 Animal models of osteogenesis imperfecta and bone fragility also report decreases in SPARC, which supports the idea that genetic mutations or polymorphisms may play a role in inherited bone diseases.^{37; 38}

Though bone maintenance and turnover have traditionally functioned on a “use it or lose it” principle, studies have found that non-weight bearing bones like the calvaria respond differently to unloading than weight bearing bones. In a study on the effects of microgravity on bone, Zhang et al. observed that though mice experienced a loss of bone and muscle mass in their legs, their calvaria showed statistically significant increases in measures of bone quality including bone volume, cortical thickness and tissue mineral density.³⁹ Macauley et al, did not observe these changes when similarly exposing mice to microgravity, however, they were found to have 16-fold lower expression of Sclerostin, which is a glycoprotein known for its bone growth inhibitory properties.⁴⁰ Both studies postulate that “head-ward shift of interstitial fluid” may be responsible for observed bone measurement changes. Exercise also been found to cause interstitial fluid movement within bones thus activating mechanoreceptors and induce bone architecture modifications.^{41; 42} However in this study, only F-WT bone volume and F-SPARC^(−/−) BS/BV were statistically significant making it difficult to conclude whether exercise may be having an effect.

Osteonectin, or SPARC, has also been characterized as an exercise-responsive gene expressed in muscle, both in vivo and in vitro, and is downregulated with aging.^43–46 Since exercise improves some of the health-related conditions associated with aging, SPARC has (figuratively and literally) sparked interest as a therapeutic target towards a healthy phenotype. Indeed, Ghanemi et al. recently demonstrated that mice overexpressing SPARC had increased mitochondrial biogenesis, reduced glycemia, and increased myoblast differentiation.⁴⁷ Interestingly, they also noted sex differences in muscle and adipose characteristics but attributed it to differences in body weight. Thus, in addition to the mechanical loading, which is known to promote bone formation during exercise, increased SPARC expression may also play a role in exercise induced bone growth. Our research and the work by Ghanemi et al suggest that these potential benefits may vary in a sex-dependent manner.

While the data presented in this study suggest mechanistic differences in bone mineralization and remodeling between males and females, it is not without limitations. First, sample size was limited to available tissue from a previous study and some of the outcome measures may have been under powered. Second, due to a change in the available μCT systems, the calvaria and femurs were imaged and analyzed using different devices, thus making it difficult to compare between the two tissues, however our conclusions between the different experimental groups are still valid. Finally, no data on bone gene transcriptional status and expression, or on hormone levels were collected on the animals during the study. This information would have helped to provide important links between SPARC, sex, and exercise that we hope to evaluate in future studies. Lastly, while mouse models allow us to analyze and control specific variables, limitations on our ability to translate our results to humans should be noted.

CONCLUSION

SPARC/Osteonectin is a protein known to modulate cellular and extracellular interactions in many tissues throughout the body and has been linked to aging and osteoporotic phenotypes. We observed differences in the effect of SPARC^(−/−) and exercise on the structural and mechanical properties of craniofacial and load bearing long bone between male and female mice. Further research is necessary to investigate the underlying mechanisms between these sex-specific differences in bone.