INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

MECHANICAL SIGNAL TRANSDUCTION has a significant influence on the biological processes of a variety of cell types, including smooth muscle cells (25), endothelial cells (8, 9, 12, 33), and bone cells (14). This process, also known as mechanotransduction, involves the conversion of a biophysical force into the cellular/molecular response leading to both rapid changes in kinase-mediated gene expression, as well as slower adaptive changes in cytoskeletal arrangement. The mechanisms for the coupling of cell-level mechanical signals into intracellular biochemical signals are currently under intense investigation, and several candidate pathways have been proposed. These include force transduction through stretch-activated cation channels within the plasma membrane, G protein-coupled calcium-dependent pathways mediated by phospholipase C and protein kinase C signaling cascades, direct nuclear matrix interactions with mechanical stress response elements yet to be identified, calcium-independent pathways involving intracellular kinase activation, and finally integrin-mediated transduction through the cytoskeleton (1, 3).

The integrins, a family of transmembrane heterodimeric glycoproteins, are the major cellular receptors for many extracellular matrix proteins and are perhaps the most well studied group of proposed force transducers (19). Because of their physical interconnection with the cytoskeleton, integrins are proposed to facilitate a series of protein-protein interactions extending from the extracellular matrix to the intracellular filamentous cytoskeleton. Ingber (20) proposed that alterations in cell shape and subsequent gene expression are therefore regulated by integrin-extracellular matrix associations that determine the tensional integrity (tensegrity) of the cell. Because of this cellular structural configuration, a load stimulus would be rapidly propagated across the cell, evoking a variety of signaling cascades to elicit a biochemical response. In fact, studies have demonstrated that cellular attachment to the extracellular matrix plays an important role in the regulation of cellular proliferation, differentiation, morphogenesis, and gene expression (2, 21, 24). Thus integrins appear to function as signaling receptors (18) that elicit biochemical signals via close association with intracellular proteins (30), induction of tyrosine phosphorylation (6), and increases in intracellular calcium (36) upon stimulation. A variety of integrin subunits have been demonstrated in bone, including ₂₁ which binds collagen (18), 1 and _v5 in osteoblasts (11, 15, 29) and _v3 in osteoclasts and endothelial cells (11, 17, 13). The  subunit mediates the association with structural intracellular proteins in localized attachment domains or "focal adhesions" (28). These focal adhesion complexes contain actin-associated proteins such as talin, vinculin, paxillin, and -actinin (5) as well as several protein kinases (10). The focal adhesion kinase (FAK) appears to play a central role in integrin-mediated signal transduction. This kinase is tyrosine-phosphorylated, and its tyrosine kinase activity is enhanced upon integrin engagement (30). Once phosphorylated, FAK is able to couple with several SH₂/SH₃ domain-containing proteins and thus trigger a complex cascade of signal transduction events.

Investigations designed to delineate the molecular mechanisms of mechanical signal transduction in bone in vivo are challenging. In contrast to in vitro studies, however, an in vivo model provides an environment with the appropriate osteoprogenitor cell population, normal blood supply, and a mechanical strain environment to tissue with more material cell-matrix interactions. We have developed an in vivo bone chamber model of intramembranous osteogenesis and adaptation (16, 23) for investigations of mechanical signal transduction at the molecular level (23). The unique feature of the bone chamber is that the hydraulic cap can be activated to apply a controlled compressive force on the trabecular bone that has formed within its walls. Utilizing this model, we demonstrate the temporal and spatial activation of FAK in response to specific mechanical load stimuli. We have begun to delineate the downstream events triggered by its activation in a microenvironment characteristic of trabecular bone in vivo (23).

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Surgical procedure. This study utilized skeletally mature, male canines housed at the AAALAC accredited (Association for Assessment and Accreditation of Laboratory Animal Care), animal facility at the University of Michigan. Before surgery, all dogs were premedicated with a cocktail consisting of butorphanol tartrate (100 mg, Torbugesic), acepromazine maleate (25 mg, Vedco, St. Joseph, MO), glycopyrrolate (5 mg, American Regent Laboratories, Shirley, NY), and 0.9% saline (added as a vehicle to bring the total volume to 50 ml). The cocktail was administered intramuscularly at a dosage of 0.1 mg/kg, 20 min before induction with thiopental sodium (Pentothal, 17.5 mg/kg, iv, to effect). Anesthesia was maintained by inhalation with isoflurane (Aerrane). The analgesics/anti-inflammatory medications, buprenorphine (Buprenex, 0.01-0.05 mg/kg iv/im) and carprofen (Rimadyl, 1 mg/lb, per os) were administered perioperatively as needed. All of the surgical procedures and experimental protocols were approved by the University of Michigan Committee on the Care and Use of Animals.

View larger version (36K): [in this window] [in a new window]   Fig. 1.   The titanium bone chamber is designed as a hollow screw with large portals for tissue infiltration. Components include the main body, loading piston, and hydraulic cap. The loading mechanism works by application of fluid pressure against the piston, which then applies direct force to the tissue within the chamber. The cylindrical specimen of bone that is repeatedly extracted from the chamber measures ~6 mm in length and diameter.

Loading experiments. To determine the temporal activation of FAK, a series of specimens was extracted from four dogs immediately, 15 min, or 45 min after a single 30-min (1,800 cycles) load stimulus of 17.8 N. Each dog underwent 5 loading/extraction procedures and was randomly assigned to be biopsied across each of the three postload time points, for a total of 20 loading experiments. Specifically, eight pairs of specimens were evaluated immediately after the 1,800 cycle load stimulus, six pairs of specimens were evaluated at 15 min, and six pairs of specimens were evaluated at the 45-min time point. The experimental flow is depicted in Fig. 2.

View larger version (8K): [in this window] [in a new window]   Fig. 2.   Experimental flow of the bone chamber model to demonstrate the temporal activation of focal adhesion kinase (FAK) in response to a mechanical load stimulus. The load/extraction procedure was performed up to 5 times in each dog. Results of these experiments are presented in Fig. 4.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Experimental flow of the studies designed to determine the threshold mechanical stimulus for FAK activation. In these studies, the duration of the load stimulus was varied as shown, and specimens were extracted within a postload time frame of 30 min. Results of these experiments are presented in Fig. 5.

Antibodies. Monoclonal antibodies to FAK (mouse monoclonal anti-FAK, clones 2a7 and 4.47) and Src (mouse monoclonal anti-Src) were purchased from Upstate Biotechnology (Lake Placid, NY). Monoclonal antibody to phosphotyrosine (PY20) was purchased from Transduction Laboratories (Lexington, KY). Horseradish peroxidase-conjugated secondary antibodies, goat anti-mouse and goat anti-rabbit, were purchased from Upstate Biotechnology and Pierce Chemical (Rockford, IL), respectively.

Immunocytochemistry. Harvested specimens were immersed in 10% polyvinyl alcohol (PVA-MW: 30,000-70,000; Sigma Chemical, St. Louis, MO) for 1 h. They were then slow cooled for 5 min by immersion in n-hexane (Electron Microscopy Sciences, Fort Washington, PA) which has been chilled to 70°C in a dry ice-alcohol slurry. Several 5- to 7-µm cryosections were taken, and the specimens were subsequently snap frozen in liquid nitrogen for immunoprecipitation and Western blotting as described below. Mounted cryosections were fixed in 10% neutral buffered formalin, washed in PBS, then incubated with blocking solution (1% BSA, 0.3% Tween-20, 10% normal horse serum in PBS) at room temperature. Slides were then incubated with monoclonal anti-FAK antibody (clone 2a7; 1 µg/ml at a 1:100 dilution) in a humid chamber at 4°C overnight. Slides were washed in PBS and then incubated with biotinylated anti-mouse IgG (2.5 µg/ml, 1:200 dilution, in 1% BSA, 0.3% Tween-20) at room temperature. FAK immunoreactivity was visualized with an alkaline phosphatase-streptavidin-substrate detection method (Vector Red, Vector Laboratories, Burlingame, CA) and counterstaining with hematoxylin.

Cell lysis, immunoprecipitation and immunoblotting. Harvested specimens were pulverized in a liquid nitrogen-cooled mortar and pestle, followed by homogenization (Polytron, Brinkman Instruments, Westbury, NY), in 2 ml ice-cold cell lysis buffer (50 mM Tris-HCl, pH 7.4; 1% NP-40; 0.25% sodium deoxycholate; 150 mM NaCl; 1 mM EGTA; 1 mM PMSF; 1 µg/ml aprotinin, leupeptin, pepstatin; 1 mM Na₃VO₄; 1 mM NaF). Insoluble material was removed by centrifugation (9,500 rpm, for 15 min at 4°C), and the supernatants were assayed for total protein concentration with a commercially available kit (Bio-Rad Laboratories, Hercules, CA). Paired samples (loaded and unloaded) were adjusted to the same concentration of protein (75 µg) in a total volume of 1 ml for each sample. Monoclonal FAK antibody (clone 2a7; 4 µl at l µg/µl) was added to the protein lysates and incubated on a shaker at 4°C, overnight. Antibodies were collected on protein G agarose beads (Calbiochem, La Jolla, CA). The precipitated protein complexes were washed three times with PBS, then resuspended in 50 µl of 2× Laemmli buffer with 10% 2-mercaptoethanol. The immunoprecipitates were resolved on an 8% SDS-PAGE gel and electrophoretically transferred to a nitrocellulose membrane (at 200 mA for 4 h). The filters were rinsed briefly in TBS without Tween 20, then blocked in Blotto (5% non-fat dry milk, Bio-Rad) at 4°C overnight. The blots were incubated with the appropriate antibody (1 µg/ml, diluted 1:600 of antiphosphotyrosine; 1:100 anti-FAK, clone 4.47; 1:20,000 anti-Src) on a shaker at 4°C overnight. The filters were washed 5 times in TBS-Tween 20, then incubated (in Blotto) with horseradish peroxidase-conjugated goat anti-rabbit IgG (diluted 1:2,000) or goat anti-mouse IgG (diluted 1:8,000) (Pierce Chemical) on a shaker at room temperature for 1 h. The blots were visualized by enhanced chemiluminescent detection (ECL, Amersham Pharmacia Biotech, Piscataway, NJ, or SuperSignal, Pierce Chemical).

Quantitative data analysis. The immunoblots were quantitated by a method of relative densities. A flat-bed scanner with light box attachment was used to make high-resolution digital copies of each gel. Densitometry was performed by use of NIH Image 1.61, and the Gel Plotting macro algorithm was used to calculate spot intensities from digitized gel image files. Background subtraction was performed by using regions adjacent to each spot. Thus spot intensities that were calibrated to the background were used in all further calculations. Each gel was evaluated separately. By use of this method, the loaded vs. unloaded blot on each gel was compared to determine a nondimensional ratio (loaded/unloaded). A ratio >1 indicated increased activation in loaded specimens compared with unloaded specimens in response to the mechanical load stimulus.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tyrosine phosphorylation of FAK in response to mechanical load. These studies were based on the hypothesis that cyclic compressive loading induces the tyrosine phosphorylation of FAK in a time-dependent manner. In six out of eight loading experiments in which pairs of specimens were extracted immediately after a 30-min load stimulus (17.8 N at 89 N/s, 1,800 cycles at 1 Hz), the Western blot analysis demonstrated a significant increase in FAK phosphorylation in loaded specimens compared with the contralateral controls (Fig. 4). There was decreased phosphorylation of FAK in six out of six pairs of specimens evaluated at 15 and 45 min compared with the immediate time point (Fig. 4). The effects of changing the duration of the load stimulus by varying the number of applied load cycles and harvesting either immediately, or within the 30-min time window are shown in Fig. 5. There was increased phosphorylation of FAK in three out of four pairs of loaded specimens harvested immediately after loading for 30 min (Fig. 5A). A 15-min load stimulus (900 cycles) also elicited FAK phosphorylation above baseline levels in 7 pairs of specimens regardless of whether the specimens were harvested immediately (3 out of 4 experiments) (Fig. 5B), or after a 15-min delay (2 out of 3 experiments) (Fig. 5A and B). A load stimulus of 120 cycles (2 min) never resulted in activation of FAK in five pairs of specimens that were harvested immediately (0 out of 3 experiments) (Fig. 5B) or after a 28-min delay (0 out of 2 experiments) (Fig. 5A).

View larger version (37K): [in this window] [in a new window]   Fig. 4.   Effect of mechanical load on the temporal activation of FAK. Dogs were implanted with a bone chamber in each tibia. For each experiment, one chamber was randomly selected to be loaded (L) and the opposite chamber was the unloaded (UL) control. The unloaded chamber was always biopsied before loading to prevent cross talk between chambers. Twenty pairs of specimens were extracted immediately, 15, or 45 min after the mechanical load stimulus (17.8 N at 89 N/s, 1,800 cycles at 1 Hz). Specimens were homogenized, and pp125FAK was immunoprecipitated from 75 µg of cell lysate with anti-FAK monoclonal antibody (clone 2A7). Immunoprecipitates were resolved by SDS-PAGE and immunoblotted with PY20. Tyrosine phosphorylation was visualized by enhanced chemiluminescence. Blots were analyzed by densitometry as described in MATERIALS AND METHODS. A ratio >1 indicates activation in response to mechanical stimulation.

View larger version (69K): [in this window] [in a new window]   Fig. 5.   FAK activation is dependent on a specific load stimulus and not on a response window of 30 min. Dogs were implanted with a bone chamber in each tibia. For each experiment, 1 chamber was randomly selected as L and the opposite chamber was the UL control. The UL chamber was always biopsied before loading to prevent cross talk between chambers. Twelve pairs of specimens were extracted after load durations of 30 min (1,800 cycles) (A), 15 min (900 cycles) (A and B), or 2 min (120 cycles) (A and B). Specimens were extracted either immediately (A and B), 15 (A and B), or 28 (A) minutes after the load stimulus. Specimens were homogenized, and pp125FAK was immunoprecipitated from 75 µg of cell lysate with anti-FAK monoclonal antibody (clone 2A7, Upstate Biotechnology). Immunoprecipitates were resolved by SDS-PAGE and immunoblotted with PY20 (Transduction Laboratories). Tyrosine phosphorylation was visualized by enhanced chemiluminescence. Blots were analyzed by densitometry as described in MATERIALS AND METHODS. A ratio >1 indicates activation in response to mechanical stimulation.

View larger version (45K): [in this window] [in a new window]   Fig. 6.   FAK protein levels did not change in response to mechanical load stimuli of 1,800, 900, or 120 cycles. Matching aliquots of protein from the specimens analyzed in Figs. 4 (A) and 5 (B) were used to immunoprecipitate FAK with an anti-FAK monoclonal antibody (clone 2A7, Upstate Biotechnology). Immunoprecipitates were resolved by SDS-PAGE and immunoblotted with a monoclonal FAK antibody (clone 4.47, Upstate Biotechnology). FAK protein was visualized by enhanced chemiluminescence.

Mechanical stimulation induces the association of Src with FAK. FAK contains several tyrosine residues that are located in amino acid sequences arranged in motifs for binding SH₂ domains. Tyrosine phosphorylation of FAK is proposed to create the formation of signaling complexes via the initial recruitment of the SH₂ domain-containing protein Src to the phosphorylated tyrosine residue. Thus it was hypothesized that the formation of Src-FAK complexes would parallel the time course of FAK activation.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Mechanical load induced an increase in FAK-Src complexes. FAK was immunoprecipitated from matching aliquots of protein from 6 pairs of specimens, which demonstrated increased FAK activation as depicted in Fig. 5. Immunoprecipitates were resolved by SDS-PAGE and blotted with polyclonal anti-Src antibody (Upstate Biotechnology). Blots were analyzed by densitometry as described in MATERIALS AND METHODS. A ratio >1 indicates activation in response to mechanical stimulation.

FAK immunoreactivity is limited to the marrow space. To determine which bone cells might be responding to mechanical stimulation, cryosections were taken from three pairs of specimens before immunoblotting and were stained with a monoclonal antibody to FAK. Cryosections revealed prominent FAK immunoreactivity among marrow fibroblasts and stromal cells (Fig. 8). There was no immunostaining associated with osteoblastic cells on trabecular surfaces or within osteocytic lacunae. Qualitatively, there were no detectable differences in the amount of FAK staining between loaded (1,800 or 900 cycles) and unloaded specimens, which is consistent with the finding that there was no increase in FAK protein with loading.

View larger version (151K): [in this window] [in a new window]   Fig. 8.   FAK immunoreactivity was limited to the marrow space. Serial cryosections were taken from 3 pairs of frozen specimens before homogenization for immunoblotting. Slides were incubated with monoclonal anti-FAK antibody (1 µg/ml at a 1:100 dilution), and FAK immunoreactivity was visualized with an alkaline phosphatase-streptavidin-substrate detection method (Vector Red, Vector Laboratories) as described in MATERIALS AND METHODS. "Red" FAK immunostaining was evident among marrow stromal cells and fibroblasts. ×40 objective.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

This study demonstrated that induction of tyrosine phosphorylation of FAK occurs in response to specific mechanical load stimuli. In this in vivo model, the threshold for mechanical stimulation of FAK phosphorylation appeared to lie between 120 and 900 cycles of an applied load of 17.8 N.

FAK can autophosphorylate at tyrosine residue 397 and can be phosphorylated at a number of other residues such as Y⁴⁰⁷, Y⁵⁷⁷, and Y⁹²⁵ by other protein tyrosine kinases. Thus it is possible that mechanical load could induce tyrosine phosphorylation of FAK either at Y³⁹⁷ or at an as yet unknown tyrosine residue. This would create high-affinity binding sites for SH₂ domain-containing proteins such as Src. In this study, Src coimmunuoprecipitated with FAK, and this association was coincident with FAK activation. In addition, mechanical stimulation induced an increase in FAK-Src coimmunoprecipitates above control levels. The data suggest, therefore, that there was an increase in focal adhesion complex formation in response to mechanical load. The persistent activation of FAK 15 min after a load stimulus of 900 cycles is consistent with the observation that FAK-Src association leads to the recruitment of other signaling molecules to the complex, with subsequent transphosphorylation of FAK at additional tyrosine residues. Specific phosphotyrosine antibodies, such as Y³⁹⁷, Y⁵⁷⁶, Y⁵⁷⁷, and Y⁹²⁵, are now available (Biosource, Camarillo, CA), and these could be utilized in future experiments.

Fibronectin stimulation of FAK in vitro has been shown to decrease over time coincident with the formation of actin stress fibers and focal contact formation in spreading NIH 3T3 fibroblasts (32). FAK incorporation into mature focal contacts may promote c-Src disassociation and downregulation of FAK activity through conformational changes. Thus the decline in FAK phosphorylation 15 and 45 min after an 1,800-cycle load stimulus may correlate with the formation of mature focal contacts, activation of structural proteins, and cytoskeletal changes. During this time course, the data also demonstrated that unloaded specimens have a significant amount of tyrosine-phosphorylated FAK. An assumption of this model is that there is integrin-dependent binding of cells within the bone chamber microenvironment to an extracellular matrix, an association that would induce the autophosphorylation of FAK. However, autophosphorylation has been demonstrated to have little effect on enzymatic activity in vitro (31). Maintenance of a basal level of tyrosine phosphorylation could serve as the threshold at which FAK remains inactive but is "poised" to respond to the appropriate level of mechanical stimulation.

Mechanotransduction is based on the premise that bone cells are able to sense strain or deformation within the extracellular matrix which surrounds them. Osteocytes, osteoblasts, and bone lining cells, all three of which are morphologic derivatives of a common pluripotent stromal precursor, have been proposed to possess mechanosensory functions. Recently, several investigators (26, 34, 35) have postulated that relatively small fluid shear stresses, created by the movement of fluid through bone during loading, could stimulate the surface membranes of such cells and initiate the mechanotransduction process. Similarly, many studies have demonstrated that direct deformation (4, 7, 22, 27) is a means by which bone cells sense strain and which subsequently has led to diverse biological activities. Interestingly, FAK immunoreactivity was limited to the stromal cells and fibroblasts within the marrow space. It is possible that, because of the less connected, highly deformable nature of the woven trabecular bone, the load stimulus engendered high strains through the marrow cavity. Alternatively, the compressive loading regime could have induced a significant amount of interstitial fluid flow through this area. This fluid flow may have subsequently produced a gradient of fluid pressure on the marrow cells. Interestingly, the stromal fibroblasts may have been the first cells to perceive these local mechanical signals and subsequently increased their tyrosine kinase activity to propagate the signal transduction process.

In summary, these studies, which were conducted utilizing a model that simulates a genuine in vivo trabecular bone microenvironment, demonstrated that tyrosine phosphorylation of FAK was induced by specific mechanical stimuli. Load-induced activation of FAK was rapid, and the duration of activation was dependent on an adequate number of applied load cycles. Mechanical stimulation increased the association of FAK with Src, and the time course of complex formation paralleled the temporal activation of FAK. Finally, it appears that the marrow stromal cell may play an important role in the process of mechanotransduction in woven bone in vivo.