The authors have no conflict of interest.
Noninvasive Loading of the Murine Tibia: An In Vivo Model for the Study of Mechanotransduction†
Version of Record online: 1 MAR 2002
Copyright © 2002 ASBMR
Journal of Bone and Mineral Research
Volume 17, Issue 3, pages 493–501, March 2002
How to Cite
Gross, T. S., Srinivasan, S., Liu, C. C., Clemens, T. L. and Bain, S. D. (2002), Noninvasive Loading of the Murine Tibia: An In Vivo Model for the Study of Mechanotransduction. J Bone Miner Res, 17: 493–501. doi: 10.1359/jbmr.2002.17.3.493
- Issue online: 2 DEC 2009
- Version of Record online: 1 MAR 2002
- Manuscript Accepted: 29 OCT 2001
- Manuscript Revised: 20 AUG 2001
- Manuscript Received: 21 MAR 2001
- mechanical loading;
- bone formation;
- insulin-like growth factor 1
Transgenic and knockout mice present a unique opportunity to study mechanotransduction pathways in vivo, but the difficulty inherent with applying externally controlled loads to the small mouse skeleton has hampered this approach. We have developed a novel device that enables the noninvasive application of controlled mechanical loads to the murine tibia. Calibration of tissue strains induced by the device indicated that the normal strain environment was repeatable across loading bouts. Two in vivo studies were performed to show the usefulness of the device. Using C57Bl/6J mice, we found that dynamic but not static loading increased cortical bone area. This result is consistent with previous models of bone adaptation, and the lack of adaptation induced by static loading serves as a negative control for the device. In a preliminary study, transgenic mice selectively overexpressing insulin-like growth factor 1 (IGF-1) in osteoblasts underwent a low-magnitude loading regimen. Periosteal bone formation was elevated 5-fold in the IGF-1-overexpressing mice but was not elevated in wild-type littermates, showing the potential for synergism between mechanical loading and selected factors. Based on these data, we anticipate that the murine tibia-loading device will enhance assessment of mechanotransduction pathways in vivo and, as a result, has the potential to facilitate novel gene discovery and optimization of synergies between drug therapies and mechanical loading.
CORTICAL AND trabecular bone begin to degrade once peak bone mass is attained during the third decade of human life.(1) This degradation results from a combination of increased bone resorption and diminished bone formation. Ultimately, if the tissue is degraded sufficiently, structural failure can occur in response to what under normal circumstances might only be considered mild trauma (e.g., a slip and fall(2)). In recent years, inhibitors of elevated bone resorption have become effective pharmaceutical interventions for the bone loss associated with postmenopausal estrogen deficiency.(3) However, although these interventions successfully block bone resorption, therapies that reverse skeletal osteopenia by stimulating bone formation are limited. The need for such alternatives has become the catalyst for an extensive research effort in both the public and the private sectors.(4)
Mechanical loading of the skeleton is a known anabolic stimulus for bone. In vivo studies have shown that bone is responsive to mechanical parameters such as strain magnitude,(5,6) strain rate,(7,8) and strain gradients.(9,10) Recent in vitro data are consistent with these tissue level studies, and confirm that bone cells are highly responsive to mechanical stimuli such as substrate deformation(11) and fluid flow.(12,13) However, despite its anabolic potential, exercise has proven only mildly successful in enhancing bone mass.(14,15) The lack of mechanistic understanding of how bone as a tissue perceives and responds to mechanical loading undoubtedly has contributed to this lack of efficacy.
Transgenic and knockout mice hold the rare potential for detailed exploration of mechanotransduction pathways in vivo. Mice have become increasingly used as a model in connective tissue biology, and the murine skeleton responds to sex steroids and pharmaceutical agents in a manner similar to other mammalian models of human bone physiology.(16,17) Their small size and relatively short life span greatly enhance cost-effective screening of developmental compounds as compared with larger animals. However, although the mouse has proven its usefulness as a model for drug discovery and development, the small size of the murine skeleton has hampered efforts to develop devices capable of applying externally controlled, noninvasive mechanical loading to specific skeletal sites. This obstacle has been overcome only recently.(18,19)
Here, we present a novel, noninvasive device that applies controlled external loads to the murine tibia, thereby facilitating use of the mouse as a tool for studying mechanotransduction signaling pathways. To show the usefulness of this model, we calibrated the tissue deformations induced in the tibia diaphysis by the device and then performed two initial in vivo studies. In the first, C57Bl/6J mice were loaded with a brief physiological magnitude regimen to assess whether bone area was altered in a manner consistent with previous in vivo studies of bone adaptation. In the second loading study, we used a transgenic mouse selectively overexpressing insulin-like growth factor 1 (IGF-1) in osteoblasts to examine tissue-specific synergism between mechanical loading and IGF-1, two potential anabolic stimulants of bone formation.
MATERIALS AND METHODS
Noninvasive murine tibia loading device
The device places the tibia in cantilever-like bending by fixing the proximal tibia and applying loads to the distal tibia just proximal to the ankle (Fig. 1). Loads were applied via a shielded linear electromagnetic actuator (Motran Industries, Inc., Valencia, CA, USA) such that the tibia was loaded in cantilever bending about the anterior-posterior (A-P) axis with the medial surface in compression and the lateral surface in tension. To load the tibia, the mouse is mask-anesthetized (2% isoflurane, chosen for rapid induction and recovery) and the thoracic region just distal to the forelimbs is strapped onto a brass plate using a Velcro strap to prevent rollover and motion. The right tibia is positioned against the fixed lateral support, and the medial gripping cup is moved along a horizontal positioning slot and locked in place such that the proximal tibia is fixed against the lateral support. The actuator (with the distal loading bracket attached) is then moved into position against the distal tibia. Under computer control, the actuator applies a load waveform via a digital analog (D/A) interface that enables programmable waveforms (including frequency, magnitude, and cycle number; Microstar Laboratories). The actuator and op-amp electronics were designed to generate a maximum 9N force in response to a 10-V driving signal. For a 100-cycle loading protocol, the entire procedure from induction to recovery requires approximately 5 minutes.
Calibration of induced strains
Because the induced normal strain environment is nonuniform due to cantilever-like bending and the nonuniform geometry of the tibia, calibration of the induced strain environment was essential. Similar to calibration of other in vivo models of bone adaptation, we used a combined experimental (in situ strain gaging) and numerical (finite element [FE] analysis) approach to characterize the induced normal strain environment.(9) A small gage width (0.25 mm) single element gage was attached just distal to the murine tibia midshaft on the lateral surface and strain data were recorded for two mice (killed immediately before application of the gage). Both mice were inserted and completely removed from the device five times over 2 days. Data collection was repeated for applied forces ranging between 0.1 and 0.9N in increments of 0.1N. At each increment, strain data from four sequential loading cycles were recorded and averaged (Fig. 1C). Under these conditions, peak strains measured at the gage attachment site ranged between 150 and 1600 μϵ. Within a given loading bout and across episodes of loading, the CV (CV = SD/mean) of the measured peak strains was assessed. To extend strain data to a full-field mapping of normal strains induced in the tibia by the loading regimen, we also developed an anatomically based FE model of the murine tibia (Fig. 2). Each calibration tibia was embedded in acrylic and serially sectioned at 1-mm intervals, and each section was ground to 80 μm. The model consisted of approximately 1650 20-node isoparametric brick elements and was based on scanned images of the 80-μm thick cross-sections taken from the site of distal load application to the proximal gripper cup. The proximal end of the model was fixed and the end force used in the study (0.3N) was applied at the distal end of the model (with a moment arm identical to that used in the loading device). Using known end-loading boundary conditions and previously published bone material properties for 10-week-old C57Bl/6J mice,(20) FE analysis was performed and results were contrasted with the experimental strain gage data. At the location of strain gage attachment, predicted strains were within 10% of those measured experimentally. For each experimental group, maximal normal strains induced at the tibia midshaft were estimated via beam theory.(21) Briefly, force and moment boundary conditions at the midshaft were established using the validated FE models, and these boundary conditions were applied to the midshaft cross-section of each animal's intact contralateral tibia to estimate normal strains induced by the device at the initiation of the loading protocol.
C57Bl/6J in vivo loading study
Fifteen female C57Bl/6J mice (10 weeks) underwent a 4-week protocol. Mice were randomly assigned to one of three groups: (1) dynamic loading, (2) static loading, or (3) age-matched controls. Before the loading procedure, the mice were anesthetized (2% isoflurane) and then positioned in the loading device. Once a day for 5 consecutive days, the right tibia of mice in the dynamic loading group received a 1-Hz (0.01/s loading rate sawtooth waveform), 0.3N peak load magnitude, 100-load cycle regimen. The right tibia of mice in the static group were exposed to a single, pseudostatic, load cycle each day of identical load magnitude as the dynamic group for an identical total loading period (100 s), but two orders of magnitude lower loading rate (0.0001/s). Animals in the aged-matched control group were anesthetized as with the dynamic and static groups but did not receive any exogenous loading. The animals in all groups were permitted normal cage activity in addition to the brief daily external loading procedure. The mice were then left unloaded for an additional 3 weeks to allow consolidation of new bone before death (i.e., permitted normal cage activity but received no loading via the device). At the conclusion of the 4-week study, the mice were killed and serial peripheral quantitative computed tomography (pQCT) scans were obtained along the diaphysis of both tibias to visualize potential areal adaptation caused by loading. Static histomorphometry was used to determine whether dynamic loading altered cortical bone area. The experimental (right) and contralateral (left) tibia were fixed in 10% neutral buffered formalin for 24 h followed by dehydration in a series of increasing ethyl alcohol (EtOH) concentrations. After the final 100% EtOH, the samples were cleared in xylene and then processed undecalcified for embedding in a methylmethacrylate-dibutyl phthalate plastic composite.(22) Transverse sections from the midshaft were machine ground on a lapping machine to a specimen thickness of approximately 30 μm (±10 μm), mounted on glass slides, and stained with toluidine blue to assess cross-sectional areal properties. Periosteal and endocortical areas were determined using a software program (OsteoMeasure; Osteometrics, Inc., Decatur, GA, USA) interfaced with a light/epifluorescent microscope and video subsystem.
IGF-1 overexpression loading study
Wild-type (FVB/N background, n = 3) and IGF-1-overexpressing mice (n = 3) were loaded as described previously. Identities of the mice (all female) were blinded throughout the experiment and analysis and all were 18 weeks of age at the initiation of the study. By this age, bone formation parameters are not distinguishable between wild-type and IGF-1-overexpressing mice, although the transgene is still expressed.(23) With the exception of the magnitude of peak-induced normal strain (less than the C57Bl/6J study because of the larger cross-sectional area), the loading protocol and applied end loads were identical to that described previously. To assess osteoblast dynamics over the course of the experiment, mice were injected with 10 mg/kg calcein 1 day before the initiation of loading and 1 day before death. After they were killed, ground sections were obtained from the midshaft of both the right and the left tibia and ground to 80 μm. Sections were imaged using an Olympus BH-2 light microscope (Olympus America Inc., Melville, NY, USA) fitted with a Polaroid digital camera (Polaroid Corp., Cambridge, MA, USA). Single-labeled surface (sLS), double-labeled surface (dLS), and interlabel thickness (Ir.L.Th) were measured to determine mineralizing surface [MS = (dLS + sLS/2)/BS ∗ 100, where BS is the bone surface], mineral apposition rate (MAR = Ir.L.Th/Ir.L.t, where Ir.L.t is the interlabel time period), and surface referent bone formation rate (BFR = MAR ∗ MS/BS) on both the periosteal and the endocortical surfaces.(24) All mice in both in vivo studies were group-housed, received standard diet and water ad libitum, and were kept on an alternating 12-h light/dark cycle. The protocol was approved by the University of Cincinnati Institutional Animal Care and USA Committee (IACUC) committee.
Characterization of induced normal strains
The murine tibia device loads the right tibia in cantilever bending without contacting the diaphysis of the tibia (Fig. 1). The induced strain environment was highly repeatable in magnitude, both within a given loading episode (CV, 1.2%) and across episodes of loading (CV, 8.3%). FE analysis confirmed that externally induced normal strains were nonuniform along the diaphysis and within a given cross-section (Fig. 2).
Effect of loading on normal mice
The end loads applied in the study (0.3N) induced peak longitudinal normal strains of 850 μϵ at the midshaft of young C57Bl/6J mice (10 weeks). This low-magnitude loading did not induce soft tissue damage at sites of load application in preliminary studies (Fig. 3). Dynamic but not static loading significantly increased midshaft cortical bone area in the right tibia. Increased bone area versus the nonloaded left tibia was the result of significant expansion of both the periosteal (i.e., enhanced bone formation) and the endocortical surfaces (i.e., enhanced bone resorption; Fig. 4). Cortical width was increased 33.7% because of dynamic loading. In contrast, cortical bone properties of the right tibia of mice undergoing static loading were not significantly altered. Likewise, no differences were observed in areal properties between the left and the right tibia of nonloaded control mice.
Effect of low-magnitude loading on IGF-1 mice
The tibia of both the wild-type and the IGF-1-overexpressing mice were larger in cross-sectional area than those of the C57Bl/6J mice. Consequently, maximal normal strains induced at the midshaft were lower in magnitude (wild-type mice, 720 ± 90 μϵ; IGF-1-overexpressing mice, 600 ± 75 μϵ). The low-magnitude loading regimen did not alter endocortical bone formation parameters in either group (Fig. 5). Periosteal MS was elevated by loading in both wild-type and IGF-1-overexpressing mice (Fig. 6). Both wild-type and IGF-1-overexpressing mice showed negligible periosteal bone formation at the midshaft of the contralateral left tibia. Mechanical loading of the right tibia of wild-type mice did not induce periosteal bone formation. In contrast, IGF-1-overexpressing mice showed a sustained adaptive response, and BFR was elevated over 5-fold versus contralateral bones.
The noninvasive murine loading device described here allows the application of reproducible, externally controlled loading to the murine tibia. The proximal metaphysis is fixed and loads are applied to the distal metaphyses without contacting the diaphysis. The first in vivo study with the device showed that the model responds to external loading in a manner consistent with previous models of bone adaptation. The second study, in which a transgenic mouse was loaded noninvasively for the first time, suggests that there is substantial potential for physiological magnitude mechanical loading to synergistically enhance the anabolic response of bone to specific factors.
The murine tibia loading device possesses a number of advantageous design features. During pilot studies, toluidine blue-stained frontal sections of the entire tibia revealed no pathological responses at the locations of load application (Fig. 3A). Thus, the device will enable examination of tissue adaptation at either the periosteal or the endocortical surfaces at sites along the diaphysis. Because periosteal apposition represents the most efficient and physiological manner by which tissue strength can be augmented, it is of some importance that any model used to explore how bone mechanotransduction might be used to augment bone properties possess the ability to assess adaptation on this surface. Because of the manner in which the tibia is loaded by the device, extremely small end loads are required to induce physiological normal strain magnitudes at the midshaft (0.1-0.6N). We anticipate that this feature will reduce potential for tissue injury associated with externally loading the tibia and will thereby facilitate extended loading protocols, potentially over the life span of the mouse. Additionally, our design holds the potential for application of loading about multiple axes (by rotating the linear actuator with respect to the fixed tibia). The flexibility to load the tibia about multiple axes may prove useful in future studies designed to identify factors mediating the site specificity of bone adaptation to mechanical loading. Currently, two other devices have been developed to apply external loads to the murine tibia.(18,19,25) The ability to apply quantifiable but distinct loads to the murine skeleton should speed our understanding of how mechanical loading is perceived by bone cells and tissue.
Calibration studies confirmed that the device places the tibia in cantilever bending and induces a nonuniform longitudinal normal strain distribution along the tibia diaphysis. Peak normal strains were highly reproducible across loading bouts. The nonuniformity of the induced strain environments necessitates care with histomorphometric analysis and, in likelihood, FE analysis to quantify the varied strain environment along the diaphysis. Initially, we anticipate that it will be necessary to independently calibrate different experimental conditions (e.g., different age or strain of mice). However, the complex strain environments also may prove advantageous in future studies. For example, the nonuniform strain environment should elicit unique patterns of gene expression at different sites along the diaphysis, thereby facilitating the ability to identify the genes and/or genetic pathways that dominate the tissue's response to mechanical load. Moreover, with genetic information in hand, it should be possible to use multiaxis loading to target mechanical stimuli to specific bone surfaces in order to optimize peak bone mass and/or tissue strength.
The data from the first in vivo study indicate that a physiological magnitude loading regimen, superimposed over the animal's normal daily loading, is capable of substantially augmenting cortical bone mass in the C57Bl/6J mouse (a strain noted for low bone density(26)). Dynamic loading elicited significant augmentation of the midshaft bone area. The absence of response to static loading and the lack of bone accretion at a site 3 mm distal to the midshaft in the dynamically loaded tibia (50% lower peak strain magnitude, data not shown) combine to provide a strong negative control for potential confounding influences with the device. Therefore, we conclude that the device noninvasively induces an anabolic response to mechanical load consistent with other models of bone adaptation.(27,28) Elucidation of the interaction between low-magnitude loading and cortical drift during growth, as evidenced by the expanded endocortical surface observed in the dynamic load group, will require further study.
IGF-1 has been associated with both heritable peak bone mass and the response of bone to mechanical loading.(29–31) In the IGF-1-overexpressing mice used in this preliminary study, young mice (6 weeks old) show elevated bone formation and bone mass compared with wild-type littermates.(23) Wild-type and IGF-1-overexpressing bone formation measures are equilibrated by 16 weeks of age, although the transgene is still expressed (and the protein remains active as suggested by this study). An advantage of this approach over one in which loading is superimposed on systemically administered factors (e.g., parathyroid hormone [PTH]) is that the synergism we sought to induce was confined to the tissue under examination and was not confounded by systemic influences. To show synergism, we used a low-magnitude loading regimen that we anticipated would not affect periosteal bone formation in wild-type mice (based on pilot studies from same-age normal mice). As expected, the low-magnitude regimen did not alter periosteal or endocortical osteoblast activity in wild-type mice. Overexpression of IGF-1 alone, without external loading, also was insufficient to induce substantial bone formation in the contralateral left tibia. An anabolic response was observed only when the two stimuli were combined. Further, the 5-fold elevation in periosteal bone formation was achieved despite peak strain magnitudes that were nearly 20% lower than those induced in the wild-type mice (because of the larger cross-sectional area before the initiation of the protocol). These initial results, although encouraging, are tempered by the small group size of this preliminary study. Given the observed differences, a power analysis indicates that doubling group size to six would be sufficient to detect statistically significant bone formation differences with a mean power of 0.93 (p = 0.05). Within the context that mechanical strain stimulates osteoblast proliferation through the estrogen receptor(11) and PTH enhances the response of osteoblasts to mechanical loading,(32) our data further highlight the potential for physiological magnitude loading to synergistically enhance the anabolic effect of selected factors.
In summary, we have developed a device that enables the noninvasive application of controlled mechanical loads to the murine tibia. In addition to use as a unique tool to study mechanotransduction in bone (i.e., via transgenic and knockout mice), we envision two further applications for the model. First, the device provides an in vivo model with which to screen and optimize pharmaceutical agents within the context of the tissue's normal functional environment. This approach will serve both to optimize therapies for patients (e.g., by reducing the dosage required to achieve protection against fracture and thus limiting potential side effects) and to maximize development opportunities by optimizing the drug selection process. Second, our device will facilitate generation of tissue that can be screened for regulatory genes and proteins in which expression is altered in response to mechanical loading. Elucidation of these pathways holds potential for development of novel targets for pharmaceutical compounds that will capitalize on the tissue's own regenerative ability as a means of protecting the skeleton against bone loss pathologies.
The authors gratefully acknowledge the contributions of Craig Bailey and Matthew Heggem in performing the C57 loading experiments and pQCT measurements and Simon Smith for technical assistance with histology. This work was funded by the Whitaker Foundation for Bioengineering (to T.S.G.), the NIH AR48102 (to T.S.G.), the University of Cincinnati Orthopaedic Research and Education Foundation (OREF) (to T.S.G. and S.S.), and the Merit Review Grant from the Veterans Administration (to T.L.C.)
- 11986 Involutional osteoporosis. N Engl J Med 314:1676–1686.,
- 21995 Risk factors for hip fracture in white women. Study of Osteoporotic Fractures Research Group. N Engl J Med 332:767–773., , , , , , , ,
- 31995 Effect of oral alendronate on bone mineral density and the incidence of fractures in postmenopausal osteoporosis. The Alendronate Phase III osteoporosis treatment study group. N Engl J Med 333:1437–1443., , , , , , , , ,
- 4NIAMS 2000 Bone biology and bone diseases research: Strategic plan (FY 2000–2004). www.nih.gov/niams/strategicplan/bone.htm.
- 51985 Regulation of bone mass by mechanical strain magnitude. Calcif Tissue Int 37:411–417.,
- 61994 Mechanical loading thresholds for lamellar and woven bone formation. J Bone Miner Res 9:87–97., , ,
- 71982 The influence of strain rate on adaptive bone remodelling. J Biomech 15:767–781., ,
- 82001 Jumping improves hip and lumbar spine bone mass in prepubescent children: A randomized controlled trial. J Bone Miner Res 16:148–156., ,
- 91997 Strain gradients correlate with sites of periosteal bone formation. J Bone Miner Res 12:982–988., , ,
- 101997 Strain gradients correlate with sites of exercise-induced bone-forming surfaces in the adult skeleton. J Bone Miner Res 12:1737–1745., ,
- 112000 Mechanical strain stimulates osteoblast proliferation through the estrogen receptor in males as well as females. J Bone Miner Res 15:2169–2177., ,
- 121996 Calcium waves in fluid flow stimulated osteoblasts are G protein mediated. Arch Biochem Biophys 326:31–38., , , ,
- 132000 Substrate deformation levels associated with routine physical activity are less stimulatory to bone cells relative to loading-induced oscillatory fluid flow. J Biomech Eng 122:387–393., , , , ,
- 141991 Prevention of postmenopausal osteoporosis. A comparative study of exercise, calcium supplementation, and hormone-replacement therapy. N Engl J Med 325:1189–1195., , , , , ,
- 151995 Effects of a one-year high-intensity versus low-intensity resistance training program on bone mineral density in older women. J Bone Miner Res 10:1788–1795., ,
- 162000 Comparative effects of estrogen and raloxifene on B lymphopoiesis and bone loss induced by sex steroid deficiency in mice. J Bone Miner Res 15:541–549., , , , ,
- 171999 Prevention of osteocyte and osteoblast apoptosis by bisphosphonates and calcitonin. J Clin Invest 104:1363–1374., , , , ,
- 181998 Bone response to in vivo mechanical loading in two breeds of mice. Calcif Tissue Int 63:442–449., , , ,
- 192001 Validation of a loading apparatus: Characterization of murine tibial surface strains in vivo. Trans 47th Orthop Res Soc 26:535., , , ,
- 201999 Growing C57Bl/6 mice increase whole bone mechanical properties by increasing geometric and material properties. J Bone Miner Res 14:2159–2166., ,
- 211977 In vivo and analytical studies of forces and moments in equine long bones. J Biomech 10:701–705.,
- 221990 Cement line staining in undecalcified thin sections of cortical bone. Stain Technol 65:159–163., ,
- 232000 Targeted overexpression of insulin-like growth factor I to osteoblasts of transgenic mice: Increased trabecular bone volume without increased osteoblast proliferation. Endocrinology 141:2674–2682., , , , , , , , , , ,
- 241987 Bone histomorphometry: Standardization of nomenclature, symbols, and units. Report of the ASBMR Histomorphometry Nomenclature Committee. J Bone Miner Res 2:595–610., , , , , , ,
- 251999 Bone response to in vivo mechanical loading in C3H/HeJ mice. Calcif Tissue Int 65:41–46., , , ,
- 261999 Histomorphometric studies show that bone formation and bone mineral apposition rates are greater in C3H/HeJ (high-density) than C57BL/6J (low-density) mice during growth. Bone 25:421–429., , , , , ,
- 271984 Regulation of bone formation by applied dynamic loads. J Bone Joint Surg Am 66:397–402.,
- 281995 Inhibition of bone resorption and stimulation of formation by mechanical loading of the modeling rat ulna in vivo. J Bone Miner Res 10:683–689.,
- 292000 Mapping quantitative trait loci for serum insulin-like growth factor-1 levels in mice. Bone 27:521–528., , , , , ,
- 301995 Increased insulin-like growth factor I mRNA expression in rat osteocytes in response to mechanical stimulation. Am J Physiol 268:E318–E327., , ,
- 312000 Serum IGF-I is higher in gymnasts than runners and predicts bone and lean mass Med Sci Sports Exerc 32:1902–1907., ,
- 321998 Role for parathyroid hormone in mechanical responsiveness of rat bone. Am J Physiol 274:E146–E154., , ,