The authors state that they have no conflicts of interest.
Increased Muscle Mass With Myostatin Deficiency Improves Gains in Bone Strength With Exercise†
Version of Record online: 5 DEC 2005
Copyright © 2006 ASBMR
Journal of Bone and Mineral Research
Volume 21, Issue 3, pages 477–483, March 2006
How to Cite
Hamrick, M. W., Samaddar, T., Pennington, C. and McCormick, J. (2006), Increased Muscle Mass With Myostatin Deficiency Improves Gains in Bone Strength With Exercise. J Bone Miner Res, 21: 477–483. doi: 10.1359/JBMR.051203
- Issue online: 4 DEC 2009
- Version of Record online: 5 DEC 2005
- Manuscript Accepted: 2 DEC 2005
- Manuscript Revised: 27 OCT 2005
- Manuscript Received: 19 AUG 2005
- mechanical loading;
- bone material properties;
- bone mechanical properties;
We tested the hypothesis that increased muscle mass augments increases in bone strength normally observed with exercise. Myostatin-deficient mice, which show increased muscle mass, were exercised along with wildtype mice. Results indicate that increases in bone strength with exercise are greater in myostatin-deficient mice than in wildtype mice, suggesting that the combination of increased muscle mass and physical activity has a greater effect on bone strength than either increased muscle mass or intense exercise alone.
Introduction: Muscle (lean) mass is known to be a significant predictor of peak BMD in young people, and exercise is also found to increase bone mass in growing humans and laboratory animals. We sought to determine if increased muscle mass resulting from myostatin deficiency would enhance gains in bone strength that usually accompany exercise.
Materials and Methods: Male mice lacking myostatin (GDF-8) were used as an animal model showing increased muscle mass. Wildtype and myostatin-deficient mice (n = 10-12 per genotype) were exercised on a treadmill for 30 minutes/day, 5 days/week, for 4 weeks starting at 12 weeks of age. Caged wildtype and myostatin-deficient mice (n = 10-12 per genotype) were included as sedentary controls. Structural and biomechanical parameters were measured from the radius.
Results: Ultimate force (Fu), displacement (Du), toughness (energy-to-fracture; U), and ultimate strain (εu) increased significantly with exercise in myostatin-deficient mice but not in normal mice. When Fu is normalized by body mass, exercised myostatin-deficient mice show an increase in relative bone strength of 30% compared with caged controls, whereas exercised wildtype mice do not show a significant increase in ultimate force relative to caged controls. Relative to body weight, the radii of exercised myostatin-deficient mice are >25% stronger than those of exercised normal mice.
Conclusions: Increased muscle mass resulting from inhibition of myostatin function improves the positive effects of exercise on bone strength.
THE RISK OF developing osteoporosis in later life is related to bone deposition during puberty and early adulthood.(1) For example, it is estimated that 60% of fracture risk in older adults is explained by the peak bone mass attained at skeletal maturity.(2) This point is underscored by the fact that ∼90% of adult BMC is acquired by the end of the adolescent period,(3,4) and 25% of the total BMC in the skeleton at maturity is acquired in just 2 years of adolescence.(4) Increased physical activity has significant, positive effects on the accretion of bone mineral in young people, indicating that bone cells are highly responsive to changes in their mechanical environment during growth.(5-7) In fact, recent studies indicate that exercise has far greater capacity to add new bone to the immature skeleton than to the skeleton of adults.(8-10) A recent longitudinal study of boys and girls across pubertal development has shown that increases in muscle mass tend to precede increases in bone mass during the pubertal growth spurt.(11)
The idea that bone strength is correlated with muscle mass was articulated clearly by Frost,(12) who stated “bone strength and mass normally adapt to the largest voluntary loads on bones. The loads come from muscles, not body weight.” This relationship is often referred to as the “muscle-bone mechanostat.” Applied strains increase fluid flow in bone(13) and strain magnitude is the principal transducing mechanism to which bone apposition is assigned in the “mechanostat” model.(14,15) Peak strains and loading rates are increased with jogging and running,(16) which would explain the increase in bone mass with exercise in young people.(17) In addition, high-frequency shock waves are generated in the hindlimb when the foot contacts the ground at the end of swing phase and velocity of the leg is suddenly brought to zero.(18,19) Recent studies have shown that high-frequency shock waves promote bone formation(20,21) and also simulate TGF-β1 expression in osteoprogenitor cells.(22) Increased muscle mass increases limb segment mass in humans, which increases landing velocity and impact shocks.(23,24) Increased muscle mass may therefore increase bone formation through a mechanotransduction pathway by its effects on gait kinetics and kinematics.
Here we use an animal model to better define the effects of exercise and muscle mass on the accretion of bone mineral during growth. Specifically, we studied the effects of increased muscle mass on BMC and BMD in the limb skeleton of mice lacking myostatin (GDF-8). Myostatin is a negative regulator of muscle growth, and mice lacking myostatin show a significant (40-100%) increase in muscle mass compared with normal mice.(25) Previous studies on the skeleton of myostatin-deficient mice indicate that these animals show increased BMD(26-28) and increased bone mass.(29) We also studied the effects of exercise on bone mineral acquisition during growth by exposing normal mice and mice lacking myostatin to regular treadmill exercise over a period of 1 month. Our hypothesis is that BMD and bone strength will be greatest in exercised animals having the greatest muscle mass.
MATERIALS AND METHODS
Sample and exercise treatment
Male myostatin-deficient mice on a CD-1 background were randomly assigned to exercise (EX; n = 10-12) and sedentary control (CON; n = 10-12) groups at 12 weeks of age. Male wildtype CD-1 mice of the same age (12 weeks) were also randomly assigned to exercise (n = 10-12) and sedentary control (n = 10-12) groups. Mice assigned to the exercise (EX) groups began an exercise regimen that involved 30 minutes of treadmill exercise 5 days/week for 4 weeks. Mice were exercised on a Columbus Instruments Exer-3/6 treadmill at a rate of 12 m/minute. Exercised mice and sedentary control mice were killed at 16 weeks of age. Animals were killed, and tissue was collected at the age of 4 months because studies on inbred mouse strains indicate that mice reach peak BMD at this time.(30) Mice were weighed, triceps brachii muscles were dissected free and weighed, and the right forelimb was frozen for mechanical testing. The left forelimb was fixed in formalin and stored in 70% ETOH for radiography and histology. Radiographs of the articulated radius and ulna were taken at 30.5 kVP and 2.5 mA for 15 s using a Faxitron X-ray.
The right radius was thawed in cold water at room temperature for 1 h and prepared for mechanical testing in three-point anteroposterior bending using a Vitrodyne V1000 Materials Testing system.(29) The radius was chosen for study because it has been shown to exhibit the least measurement error and variability of all mouse long bones in three-point bending.(31) Radii were mounted on stainless steel fixtures spaced 5 mm apart, 2.5 mm either side of center. Testing was linear displacement control with a displacement rate of 0.10 mm/s using a Transducer Techniques 4.5-kg load cell. Radii were loaded to failure with data points recorded every 0.01 s. Structural, or extrinsic properties, such as ultimate force (Fu; height of curve), stiffness (S; slope of curve), ultimate displacement (du; width of curve), and energy to fracture (U; area under curve) were calculated from load-displacement curves.(32) The cross-sectional moment of inertia (CSMI) was measured from histological sections (described below) using the parallel axis theorem CSMI = ic + Ad2, where for each rectangular element in the section of width w and height h, ic = wh3/12, A = wh, and d is the distance of each rectangular element from the neutral axis.(33) Ultimate stress (σu) was calculated as FuLb/8I, ultimate strain (εu) as du × (6b/L2), and Young's modulus (E) as S × (L3/48I), where L equals length of the test specimen (5 mm), b equals anteroposterior diameter of the specimen, and I is the cross-sectional moment of inertia (CSMI).(33,34)
Histology, histomorphometry, and electron microscopy
The left radius was cut at the midshaft using a diamond wire saw, the distal one-half was decalcified and embedded in paraffin, and cross-sections were cut at 4 μm and stained with H&E. The following structural parameters were measured from radius diaphyseal cross-sections using SigmaScan image analysis software: total section area (T.Ar), marrow area (Ma.Ar), cortical thickness (Ct.Wi), and cortical area (Ct.Ar). The proximal one-half of the radius was embedded in methyl-methacrylate and sectioned using a low-speed diamond saw. Sections were mounted into stubs using colloidal graphite, coated with carbon, and viewed using backscattered electron microscopy and X-ray dispersive spectroscopy (EDAX).
A single-factor ANOVA with group (wildtype CON, wildtype EX, myostatin-deficient CON, myostatin-deficient EX) as the factor was used to test for between-group differences in mechanical and structural parameters, and Fisher's least significant difference (LSD) test was used for pairwise posthoc comparisons. A two-factor ANOVA was used to test for treatment × genotype interactions for all variables measured. Because the mice in different treatment groups were found to differ in body weight, two approaches were used to adjust for body mass differences. First, data were adjusted for differences in body mass using a general linear model of covariance (ANCOVA) with body weight as the covariate. Second, dependent variables were divided by body mass, and the normalized variables were compared between groups using ANOVA. Finally, coefficients of determination (r2) were calculated for dependent variables using linear regression with either body mass or triceps brachii muscle mass as independent variables.
Body composition data (Table 1; Fig. 1) show that body mass declined significantly with exercise in wildtype and myostatin-deficient mice. Body mass of sedentary myostatin-deficient mice was greater than that of sedentary wildtype mice, but body weight of exercised myostatin-deficient mice was not significantly different from that of sedentary wildtype mice (Table 1). Two-factor ANOVA showed significant treatment × genotype interactions for triceps mass (p < 0.05) and forelimb mass (p < 0.001), which declined with exercise in myostatin-deficient mice but increased with exercise in wildtype mice. Biomechanical testing of the radius showed that ultimate force increased significantly with exercise in myostatin-deficient mice (Fig. 1; Table 2) but not in wildtype mice. Likewise, total displacement, energy-to-fracture, and ultimate strain increased significantly with exercise in myostatin-deficient mice but not in wildtype mice. Plots of force-displacement curves (Fig. 2) showed that exercise tends to increase total displacements in animals of both genotypes, which increases the area under the curve (toughness, or energy to fracture), but the increase in displacement was only significant for myostatin-deficient animals. Two-factor ANOVA revealed a significant (p = 0.01) treatment × genotype interaction for displacement, in which exercise yielded a much greater increase in myostatin-deficient mice than in normal mice.
Structural parameters (Table 3) and cross-sections (Fig. 3) showed that the radii of myostatin-deficient mice differed significantly in shape from those of wildtype mice. Specifically, the cross-sections of wildtype mice were relatively circular, whereas those of myostatin-deficient mice became elliptical, expanded in the anteroposterior (A-P) direction. The radii of myostatin-deficient mice were curved longitudinally (Fig. 3), presumably to increase space for muscle packing in the forearm region. Exercise significantly decreased medullary area in the knockout mice but not in wildtype mice (Table 3), and structural and biomechanical parameters normalized by body mass (Table 4) indicated that the exercised myostatin-deficient mice had greater bone diameters, relative to body weight, than exercised wildtype mice. These data suggest that the increased bone strength acquired with exercise in the myostatin-deficient mice was caused primarily by changes in bone structural parameters, which is further indicated by backscattered electron micrographs indicating similar mineralization patterns between sedentary and exercised myostatin-deficient mice (Fig. 4). Coefficients of determination calculated for several biomechanical and structural parameters indicated that muscle mass is a more robust predictor of ultimate breaking force, stiffness, total cross-sectional area, and cross-sectional area moment of inertia than body weight (Table 5).
We tested the prediction that exercise has a greater effect on bone strength in animals with increased muscle mass compared with animals having relatively smaller limb muscles. Our results support this prediction. Myostatin-deficient mice have significantly larger muscles than normal mice, and exercise yielded a greater increase in bone strength, toughness, and ultimate strain in double-muscled mice than in normal mice. The marked increase in bone strength with exercise in myostatin-deficient animals involved significant changes in bone structural parameters, such as medullary area and bone diameter. Relative to body weight, exercised myostatin-deficient mice have radii that are 25% stronger than those of exercised wildtype mice (Table 4). These data suggest that the combination of exercise and increased muscle mass has a much greater effect on bone strength than simply exercise or muscle mass alone. Moreover, body mass of exercised myostatin-deficient mice did not differ significantly from that of sedentary control mice, yet ultimate fracture strength of the radius was 50% greater in the exercised myostatin-deficient mice than in weight-matched sedentary control mice. Muscle mass would therefore seem to be a more important determinant of bone strength than body weight, and regression analyses indeed showed that, within this sample, muscle mass is a stronger predictor of bone strength, stiffness, bone cross-sectional area, and bone cross-sectional area moment of inertia than body mass. It should be noted, however, that the increased curvature of the radius in the myostatin-deficient animals might place this bone under a very different loading regimen than bones that are less curved (e.g., humerus and femur). It is therefore possible that the effects we observed with exercise in the radius may not be observed in other regions of the skeleton.
Notably, exercise yielded a 15% increase in triceps brachii mass in normal mice but a decrease of 10% in triceps brachii mass in the knockout mice. Myostatin-deficiency increases mass of skeletal muscles, but this seems to be caused primarily by a disproportionate increase in type II (fast-twitch) fibers.(35) It is possible that the decrease in triceps mass observed with endurance exercise in the myostatin-deficient mice is caused by a loss of surplus fast-twitch fibers. A more likely explanation is that muscle tissue was broken down in the exercised myostatin-deficient mice to serve the energetic demands of increased physical activity. It is known that muscle proteins are often broken down and converted to energy with strenuous and prolonged aerobic exercise. Myostatin-deficient mice have very little body fat.(36,37) It is likely that, whereas normal mice used fat reserves as energy sources during exercise, the myostatin-deficient mice were instead forced to use muscle because they had very little body fat at the start of the training regimen. Nevertheless, even though some triceps mass was lost with exercise in the myostatin-deficient mice, the triceps muscles of the exercised knockout mice were 70% larger than those of exercised wildtype mice relative to body weight.
The myostatin-deficient mice included in this study have forelimbs that weigh ∼70% more than those of wildtype mice. In a preliminary study,(38) we found that adult mice lacking myostatin choose slower walking gaits than normal mice during free-speed walking trials. It is well known that animals can moderate peak external loads on their limbs by increasing the time over which force is generated (e.g., moving more slowly),(39) which also reduces the number of impact and loading cycles. It is likely that the slower walking gaits of myostatin-deficient animals attenuate the peak vertical forces experienced by their forelimbs during daily cage activity; however, during treadmill exercise, the knockout mice are forced to move at the same speed as the normal mice. This means that the knockout mice must increase muscle volume recruitment to move their heavier limbs at the same rate as the normal mice. This increase in muscle recruitment in the knockout mice will increase contractile forces applied to the limb bones. In addition, given equal limb accelerations between normal and knockout mice, impact force will be higher in the knockouts because of the increased limb segment mass. Increased muscle mass increases limb segment mass in humans, which increases landing velocity and impact shocks.(23,24) The increased bone stresses that accompany muscle recruitment during swing phase, and the greater impact shocks that accompany stance phase are therefore likely to increase bone strength in myostatin deficient mice through a mechanotransduction pathway. An alternative hypothesis is that myostatin deficiency has a direct effect on bone metabolism by regulating osteoblast and/or osteoclast activity. This hypothesis is unlikely because the myostatin receptor, the type IIB activin receptor, is not expressed at significant levels in skeletal tissues.(40)
Increased muscle mass in the myostatin-deficient mice was attained through genetic manipulation and so its relevance for the human condition, in which increased muscle mass is usually achieved through resistance exercise, could be questioned. We believe, however, that our findings are clinically significant for the following two reasons. First, studies of patients with Duchenne muscular dystrophy (DMD) show that these individuals have significantly lower trunk and limb BMD than healthy subjects.(41-43) Analyses of body composition and bone densitometry in these patients also show that the low BMD observed with DMD is highly correlated with a significant decrease in lean body mass.(30,41,43) Myostatin inhibitors have recently received considerable attention as novel agents for treating DMD.(44) Our results suggest that a combination of exercise and treatment with myostatin inhibitors may be useful for increasing bone strength in patients suffering from DMD. Second, the data presented here predict that gains in bone strength among young people will be greatest in cases where regular endurance exercise is integrated with activities that increase muscle mass, such as strength training. Given that the risk for osteoporosis in later life is related to bone deposition during puberty and early adulthood, and bone tissue seems to be most responsive to the anabolic effects of exercise early in life, we believe that these findings may be significant for osteoporosis prevention strategies aimed at maximizing bone mass and bone quality in young people.
This study was supported by National Institutes of Health (NIAMS) Grant AR049717-01A2.
- 151992 Orthopedic principles of skeletal growth, modeling, and remodeling. In: CarlsonD, GoldsteinS (eds.) Bone Biodynamics in Orthodontic and Orthopedic Treatment. University of Michigan, Ann Arbor, MI, USA, pp. 15–51.
- 382005 Locomotor biomechanics and muscle-bone interactions in myostatin-deficient mice. FASEB J 19:199.4., , ,