Osteoblasts subjected to fluid shear increase the expression of the early response gene, c-fos, and the inducible isoform of cyclooxygenase, COX-2, two proteins linked to the anabolic response of bone to mechanical stimulation, in vivo. These increases in gene expression are dependent on shear-induced actin stress fiber formation. Here, we demonstrate that MC3T3-E1 osteoblast-like cells respond to shear with a rapid increase in intracellular Ca2+ concentration ([Ca2+]i) that we postulate is important to subsequent cellular responses to shear. To test this hypothesis, MC3T3-E1 cells were grown on glass slides coated with fibronectin and subjected to laminar fluid flow (12 dyn/cm2). Before application of shear, cells were treated with two Ca2+ channel inhibitors or various blockers of intracellular Ca2+ release for 0.5–1 h. Although gadolinium, a mechanosensitive channel blocker, significantly reduced the [Ca2+]i response, neither gadolinium nor nifedipine, an L-type channel Ca2+ channel blocker, were able to block shear-induced stress fiber formation and increase in c-fos and COX-2 in MC3T3-E1 cells. However, 1,2-bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid-AM, an intracellular Ca2+ chelator, or thapsigargin, which empties intracellular Ca2+ stores, completely inhibited stress fiber formation and c-fos/COX-2 production in sheared osteoblasts. Neomycin or U-73122 inhibition of phospholipase C, which mediates d-myo-inositol 1,4,5-trisphosphate (IP3)-induced intracellular Ca2+ release, also completely suppressed actin reorganization and c-fos/COX-2 production. Pretreatment of MC3T3-E1 cells with U-73343, the inactive isoform of U-73122, did not inhibit these shear-induced responses. These results suggest that IP3-mediated intracellular Ca2+release is required for modulating flow-induced responses in MC3T3-E1 cells.
- actin cytoskeleton
- intracellular calcium, c-fos
- phospholipase C
the mechanical environment of the skeleton is crucial to bone development and architecture. Removal of physical stimulation reduces bone formation rate and matrix protein production (34, 55). Conversely, in vivo application of exogenous mechanical loading stimulates bone formation, exhibiting threshold kinetics for both strain magnitude and strain rate (43, 48). Primary and clonal osteoblast-like cells respond to mechanical stimulation in vitro with increased proliferation and expression of anabolic markers associated with osteogenesis (for review, see Ref. 14). Two factors that respond acutely to mechanical loading in vivo are c-fos and cyclooxygenase-2 (COX-2) (8, 49). Expression of c-fos, an AP-1 subunit, is induced in osteogenic cells within minutes of stimulation. Transgenic mice with altered expression of this protein present severe pathologies, indicating the importance of c-fos in normal bone physiology (17, 53). COX-2, the inducible isoform of cyclooxygenase, is a key enzyme in the production of prostaglandins and is associated with the release of PGE2 and PGI2 in bone cells in response to mechanical stimuli in vitro (26). In rats subjected to tibial four-point bending, periosteal bone formation is inhibited by NS-398, a specific blocker of COX-2 (15). These observations indicate that both c-fos and COX-2 respond quickly in mechanically loaded bone and osteogenic cells and that expression of these factors may be essential to the mechanically induced response of bone.
Although the effects of mechanical stimulation on bone formation are well documented, the mechanisms through which the osteoblast translates a biophysical stimulus into a cellular response are poorly understood. Several candidates for this conversion exist, including mechanosensitive ion channels, G protein-dependent pathways, and the integrin-cytoskeletal axis (10, 14, 27). One of the earliest responses of osteoblasts to fluid shear and strain is a rapid increase in intracellular Ca2+ concentration ([Ca2+]i) (21) that is dependent on both extracellular Ca2+ entry and intracellular Ca2+ release (20). This rise in [Ca2+]i has been linked to nitric oxide release and PG production in endothelial cells and osteoblasts in response to shear (31, 51). [Ca2+]ican also modify the actin cytoskeleton through activation of various actin-binding proteins (11, 29) and is important in establishing the linkage between the cytoskeleton and integrins (3). We have recently shown that actin reorganization into stress fibers is required for the flow-induced increase in c-fos and COX-2 expression and production (37). In this study, we investigated the influence of [Ca2+]i in response to fluid shear on actin reorganization and gene expression in MC3T3-E1 osteoblast-like cells. With the use of inhibitors of membrane ion channels and intracellular Ca2+ release, we found that modification of the actin cytoskeleton and gene expression was dependent on phospholipase C (PLC)-mediated release of Ca2+ from intracellular stores and that blocking this pathway completely abolished these responses.
MATERIALS AND METHODS
MC3T3-E1 cells, a mouse osteoblastic cell line (gift from Dr. Joseph P. Bidwell, Indiana University) were cultured in α-MEM media (GIBCO, Grand Island, NY) containing 10% fetal bovine serum and 100 U/ml penicillin G (Sigma, St. Louis, MO) and maintained in a 95% air-5% CO2 humidified environment at 37°C. Cells were grown to confluence on glass slides coated with 10 μg/ml fibronectin (Sigma). To determine the role of [Ca2+]i on mechanically induced cytoskeletal organization and gene expression, cells were treated 0.5–1 h before the application of flow with various agents that alter release or entry of Ca2+ in the cell. 1,2-Bis(2-aminophenoxy)ethane-N,N,N′,N′-tetraacetic acid (BAPTA)-AM was used to chelate intracellular Ca2+. GdCl3 (10 μM) and nifedipine (5 μM) were used to block mechanosensitive and L-type Ca2+ channels, respectively. Thapsigargin (1 μM) was used as an intracellular Ca2+release inhibitor. Neomycin (10 μM) and U-73122 (10 μM) were used to inhibit PLC. U-73343 (10 μM), an isoform of U-73122 that binds to PLC but fails to inhibit, was used to determine the specificity of PLC inhibition on the response of MC3T3-E1 cells to shear. All agents except BAPTA-AM (Calbiochem, San Diego, CA), U-73122, and U-73343 (Biomol, Plymouth Meeting, PA) were purchased from Sigma Chemical.
Fluid flow was applied to cells in a parallel plate flow chamber using a closed flow loop (16) (Cytodyne, San Diego, CA). This system uses a constant hydrostatic pressure head to drive media through the channel of the flow chamber, subjecting the cell monolayer to steady laminar flow and producing a well-defined fluid shear stress. The apparatus was maintained at 37°C, and the medium was aerated with 95% air-5% CO2 during the experiment. A flow sensor (SWF-5 flowmeter; Zepeda Instruments, Seattle, WA) was incorporated into the flow loop to monitor flow rate. The time course of fluid shear-induced gene expression of c-fos and COX-2 was established by subjecting cells to 12 dyn/cm2 flow for 5, 15, 30, 60, and 180 min. Unless otherwise specified, MC3T3-E1 cells were subjected to shears of 12 dyn/cm2 for 1 h in all other experiments.
To determine the effects of fluid shear on [Ca2+]i, MC3T3-E1 cells were grown on fibronectin-coated quartz slides under the same conditions as described above. Before the experiment, cells were loaded with 3 μM fura 2-AM (Molecular Probes), a fluorescent intracellular Ca2+ chelator, in Hanks' balanced salt solution (HBSS) for 30 min at 37°C. Cells were rinsed and incubated for an additional 15 min with HBSS alone to allow for complete deesterification of the fluorescent probe. The quartz slide was then placed in a modified parallel plate flow chamber, and baseline levels of [Ca2+]i were obtained for 5 min. Flow was then introduced to the chamber through a syringe mounted on a Harvard syringe pump (Harvard Apparatus) to produce 12 dyn/cm2 shear stress for 3 min. A ratiometric video-image analysis apparatus (Intracellular Imaging, Cincinnati, OH) was used to determine changes in [Ca2+]i. This apparatus utilizes a xenon lamp equipped with quartz collector lenses that illuminate the cells through a computer-controlled shutter and filter changer containing two interference filters (340 and 380 nm). After excitation, emitted light passes through a 430-nm dichroic mirror, is filtered at 510 nm, and imaged by an integrating charge-coupled device video camera. Consecutive frames obtained at 340 and 380 nm are ratioed, and the [Ca2+]i in each cell is calculated from this ratio by comparison to a fura 2-free acid standard curve.
Mean peak [Ca2+]i response was calculated by determining the difference between peak Ca2+response to flow and baseline Ca2+ values. This difference was then divided by baseline Ca2+ values and expressed as a percentage increase over baseline. The percentage of cells responding was determined by dividing the number of cells that responded with a >50% increase in [Ca2+]i by the total number of cells.
Immunocytochemistry and fluorescence microscopy.
After fluid flow, cells were washed in PBS and fixed in 4% paraformaldehyde in PBS for 15 min. Cells were then permeabilized with 0.2% Triton X-100 in PBS for 5 min followed by two washes of PBS. To reduce the nonspecific binding, cells were treated with 3% donkey serum in PBS for 20 min. For F-actin staining, cells were incubated with FITC-phalloidin (Molecular Probes, Eugene, OR) for 30 min then washed three times with PBS. For COX-2 or c-fos production, cells were incubated with 1:100 primary antibodies against either COX-2 or c-fos (Santa Cruz Biotechnology, Santa Cruz, CA) for 1 h in a humidified chamber at 37°C, followed by incubation with tetramethylrhodamine isothiocyanate-conjugated secondary antibodies (1:50; Jackson ImmunoResearch Laboratories, West Grove, PA) for 30 min. Images were recorded on Kodak Tmax 400 film using a Nikon Optiophot II microscope through either ×60 or ×100 objectives (1.4 numerical aperture).
Northern blot analysis.
After shear application, total RNA was extracted from cells using TRIzol reagent (GIBCO). Ten micrograms of total RNA were run on a 1% agarose/0.44 M formaldehyde gel, transferred to a Hybond nylon membrane (Amersham, Arlington Heights, IL) by capillary blotting, and fixed to the membrane by ultraviolet irradiation. Filters were hybridized for 2 h at 65°C in a rapid hybridization buffer (Amersham) containing radiolabeled cDNA probes. Rat c-fos, COX-2, and glyceraldehyde-3-phosphate dehydrogenase cDNA probes (gifts from Dr. Joseph P. Bidwell) were labeled with [32P]dCTP (specific activity > 3,000 Ci/mmol, New England Nuclear) using a random prime labeling kit (Boehringer Mannheim). Filters were washed and exposed to Fuji Rx film at −80°C. Band intensity was analyzed by scanning densitometry. Statistical significance of differences in expression was performed by one-way ANOVA. Comparisons were made between treated cells and simultaneously sheared controls.
Western blot analysis.
After flow, cells were washed with cold PBS and incubated with ice-cold lysis buffer containing 5 mM HEPES (pH 7.9), 150 mM NaCl, 26% glycerol (vol/vol), 1.5 mM MgCl2, 0.2 mM EDTA, 0.5 mM dithiothreitol, and 0.5 mM phenylmethylsulfonyl fluoride. Whole cell lysates (30 μg) were mixed with equal volume of 2× Laemmli sample buffer. The mixture and prestained molecular weight marker were boiled for 5 min and separated by 7.5% SDS-PAGE. Proteins were electrophoretically transferred to a polyvinylidene difluoride membrane. The membrane was blocked in Tris-buffered saline containing 5% nonfat dry milk and 0.05% Tween 20 (TBST) for 1 h, then incubated with 1 μg/ml rabbit antibodies to α-actinin (Sigma), rabbit antibodies to c-fos (Upstate Biotechnology, Lake Placid, NY), or goat antibodies to COX-2 (Santa Cruz Biotechnology) overnight at 4°C. The membrane was washed with BST buffer, then incubated with donkey anti-rabbit or goat IgG peroxidase conjugate (1:5,000 dilution). The immunodetection was accomplished using the enhanced chemiluminescence kit (New England Nuclear Life Science Products, Boston, MA).
Time course of fluid shear-induced upregulation of c-fos, COX-2, and actin stress fiber formation.
Application of 12 dyn/cm2 fluid shear to MC3T3-E1 cells produced a rapid increase in expression of c-fos and COX-2 mRNA. Expression of c-fos mRNA was increased 2.3-fold within 15 min of flow application and peaked at 30 min with a 3.6-fold increase (Fig.1 A, top). Expression of c-fos was significantly reduced after 1 h of flow and had returned to baseline by 3 h. COX-2 expression was increased fourfold after 30 min of flow (Fig. 1 A, middle) and increased for the duration of flow to a 15-fold increase in expression at 3 h. Consistent with our previous findings, application of fluid shear to MC3T3-E1 cells resulted in reorganization of the actin cytoskeleton. As shown in Fig. 1 B, actin filaments were poorly organized in MC3T3-E1 cells before application of shear, with only a few thin stress fibers detectable. After 1 h of shear at 12 dyn/cm2, actin filaments became organized into thick, abundant stress fibers that aligned in parallel to the long axis of the cell. Figure 1 B(right panels) illustrate that reorganization of actin corresponded to increases in c-fos and COX-2 protein levels, indicating that fluid shear stress can rapidly elicit responses in osteoblasts to produce factors involved in bone formation.
[Ca2+]imodulates shear-induced changes in stress fiber formation and gene expression.
To determine if [Ca2+]i levels changed in response to shear, fura 2-loaded MC3T3-E1 cells were imaged while subjecting the cells to 12 dyn/cm2 shear stress. Baseline values of [Ca2+]i before shear ranged from 50 to 100 nM. Application of shear induced a rapid increase in [Ca2+]i within 20 s of onset. The mean peak [Ca2+]iresponse to shear was 92 ± 12% (SE) over baseline (Fig. 2), with 54% (41/76) of the cells responding (Table 1).
To explore the role of [Ca2+]i in shear-induced actin stress fiber formation and production of c-fos and COX-2, cells were pretreated with BAPTA-AM (30 μM), a membrane-permeable Ca2+ chelator, for 30 min before application of 12 dyn/cm2 shear for 1 h. BAPTA-AM completely abolished the fluid shear-induced development of actin stress fibers in MC3T3-E1 cells (Fig. 3,right). Overlay of immunofluorescent labeling of c-fos demonstrates that BAPTA-AM also prevented the increase in c-fos (Fig.3, right) and COX-2 (data not shown) in response to shear. These data indicate that [Ca2+]i is rapidly elevated in response to fluid shear stress and suggest that an increase in [Ca2+]i is required for actin cytoskeletal reorganization and the increases in c-fos and COX-2 production that accompany this response.
Intracellular Ca2+ release, not extracellular Ca2+ entry, is responsible for shear-induced responses in MC3T3-E1 cells.
To determine if extracellular Ca2+ entry via Ca2+ channels is required for MC3T3-E1 cells to respond to fluid flow, cells were treated for 1 h before application of fluid flow with either GdCl3 (10 μM) or nifedipine (5 μM). Gadolinium inhibits mechanosensitive, cation-selective channels (56) that have been characterized in osteoblast-like osteosarcoma cells (9,12) and reported in MC3T3-E1 cells (13). Nifedipine, a dihydropyridine compound, inhibits the L-type, voltage-sensitive Ca2+channel found in osteoblasts (7, 12). As previously reported (21), addition of 10 μM Gd3+ before application of flow significantly decreased the mean peak [Ca2+]i response and reduced the number of cells responding to shear stress (Table 1). Nifedipine did not significantly alter the [Ca2+]iresponse to shear (data not shown). Overlays of phalloidin-stained F-actin with immunofluorescent labeling of c-fos indicate that neither channel blocker altered the fluid shear-induced response of actin reorganization or c-fos production in MC3T3-E1 cells (Fig.4, B and C). These inhibitors also had no effect on the production of COX-2 in MC3T3-E1 cells in response to shear (see Fig. 6 A). When these inhibitors were used in combination, c-fos and COX-2 production was not significantly different from sheared controls (data not shown).
To determine the role of intracellular Ca2+ release on these shear-induced responses, MC3T3-E1 cells were pretreated for 30 min with thapsigargin (1 μM) before shear. Thapsigargin induces release of Ca2+ from intracellular stores within minutes of application and prevents the refilling of these stores (46). Thapsigargin significantly reduced the mean peak [Ca2+]i response (38 ± 9%) and decreased the number of cells responding to fluid shear stress (Table1). Pretreatment with thapsigargin also completely abrogated the reorganization of the actin cytoskeleton and prevented the upregulation of c-fos (Fig. 4 D) and COX-2 after 1 h of flow. These data indicate that intracellular Ca2+ release, but not mechanosensitive or dihydropyridine-sensitive Ca2+channels, is required for the shear-induced responses of MC3T3-E1 cells.
Role of the PLC pathway in the shear-induced responses of MC3T3-E1 cells.
In osteoblasts, as well as other cell types, intracellular Ca2+ release is activated byd-myo-inositol 1,4,5-trisphosphate (IP3), which is cleaved from phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P 2] by PLC. Pretreatment of MC3T3-E1 cells with neomycin (10 mM), an inhibitor of PLC, for 30 min completely blocked actin stress fiber formation and production of c-fos (Fig. 5 B) and COX-2 in response to shear. However, neomycin has multiple effects in the cell, such as inhibition of Ca2+ channels (19). Therefore, we used U-73122 (10 μM), a specific inhibitor of PLC, and U-73343 (10 μM), an isoform of U-73122 that binds to PLC but does not inhibit. Addition of U-73122 significantly reduced the mean peak [Ca2+]i response and reduced the number of cells responding to shear (Table 1). Like neomycin, pretreatment of MC3T3-E1 cells with U-73122 completely abolished stress fiber formation and increases in production of c-fos (Fig. 5 C) and COX-2, whereas U-73343 had no effect on these responses (Fig.5 D). These data suggest that IP3-mediated Ca2+ release from intracellular stores is responsible for the shear-induced changes in the actin cytoskeleton and gene expression in MC3T3-E1 cells.
To confirm the data obtained from immunocytochemistry, Western blotting of cell extracts from control and blocker-treated cells after application of fluid shear was performed. In nontreated MC3T3-E1 cells, fluid shear induced 2.8 ± 0.5- and 6.5 ± 0.2-fold increases in c-fos and COX-2 protein production, respectively (Fig.6). Figure 6 A illustrates that BAPTA completely abolished the response of c-fos and COX-2, whereas gadolinium (2.0 ± 0.4 and 4.5 ± 0.7, respectively) and nifedipine (1.9 ± 0.5 and 4.1 ± 0.6, respectively) failed to alter c-fos and COX-2 production. Figure 6 B shows that inhibition of intracellular Ca2+ release by thapsigargain completely suppressed shear-induced increases in c-fos and COX-2 [0.1 ± 0.2 and 0.1 ± 0.4 (means ± SE), respectively]. Block of PLC with U-73122 (0.1 ± 0.4 and 0.1 ± 0.5, respectively) mimicked the response of MC3T3-E1 cells to thapsigargin, whereas U-73343 (2.6 ± 0.8 and 3.5 ± 0.9, respectively) failed to inhibit the production of these factors to flow.
Although the anabolic effects of mechanical loading on bone have been well documented, the type of mechanical stimulus that osteogenic cells of bone perceive to produce this response has yet to be determined. Bending of bone undoubtedly subjects bone cells to strain, but also produces fluid movement within the lacunae and canaliculi that has been proposed as a significant mechanical signal (54). It is unclear from both in vivo and in vitro studies which mechanical force is the dominant signal in bone mechanotransduction. In vitro studies demonstrate that both strain and fluid shear create similar responses in osteoblastic cells, activating second messenger and signaling pathways that may be important to the osteogenic response of bone to load (14). Recently, we have shown that fluid forces, not physiological levels of strain, increase the expression of osteopontin, an anabolic marker, in cultured osteoblast-like cells (36). These studies suggest that fluid shear is a significant mechanical stimulus to osteogenic cells.
We have previously found that fluid shear induces expression of c-fos and COX-2 in MC3T3-E1 cells that is dependent on the reorganization of the actin cytoskeleton into stress fibers (37). Here, we demonstrate that IP3-mediated intracellular Ca2+ release is required for these shear-induced responses. Fluid shear activates many second messenger pathways in osteoblasts including cAMP (6, 41), PLC-induced release of IP3 (39), nitric oxide (23), and G proteins (42). One of the earliest responses of primary calvarial osteoblasts is a rapid increase in [Ca2+]i that occurs within seconds of application of shear (20). Subsequent studies found that shear failed to produce this response in MC3T3-E1 cells (2). However, in this study, we were able to elicit an [Ca2+]i response to shear. The reason for the discrepancy between these observations may be that the [Ca2+]i response to shear stress is dependent on the time in culture. MC3T3-E1 cells cultured for 3–4 days respond to shear in a manner similar to that of cultured primary calvarial osteoblasts, corresponding to translocation of the α1C-subunit of the L-type Ca2+ channel to the plasma membrane (Brubaker and Duncan, unpublished observations). The increase in [Ca2+]i is dependent on both extracellular Ca2+ entry and release of Ca2+ from intracellular stores (20). Gadolinium, a mechanosensitive channel blocker, significantly reduced the mean peak [Ca2+]i response, but had no effect on gene expression or stress fiber formation. Nifedipine, an inhibitor of voltage-sensitive L-type Ca2+ channels, also did not affect this response. MC3T3-E1 cells treated with both inhibitors continued to respond to shear stress. Thus, even though the [Ca2+]i increase was dependent on membrane bound channels, we found no significant role for these channels in the shear-induced responses in MC3T3-E1 cells.
These observations agree with similar studies in endothelial cells (30), yet they contrast reports suggesting the involvement of Ca2+ channels in the osteoblastic response to mechanical stimuli. Hung et al. (21) found that removal of extracellular Ca2+ completely abolished the increase in [Ca2+]i in response to shear, and inhibition of Ca2+ channels with verapamil decreased strain-induced 45Ca2+ incorporation in ROS 17/2.8 cells (50). In vitro studies show that Gd3+significantly suppressed shear-induced PGE2 and nitric oxide release in osteocytes and osteoblasts, respectively (1, 42), and nitric oxide production in bone organ cultures subjected to mechanical loading was markedly reduced with Gd3+ and nifedipine treatment (38). The discrepancy between the data presented here and these previous studies may lie in the osteoblast-like cell types used for in vitro studies. For example, mechanosensitive channels are highly expressed and functionally active in UMR-106.01 cells and ROS 17/2.8 cells (12, 33) but not in MC3T3-E1 cells that have not been mechanically stimulated (13, 33). Another possible explanation lies in the sites that would be affected by these sources of Ca2+. Ca2+ entry would rapidly raise [Ca2+]i levels at the membrane, whereas Ca2+ release from the endoplasmic reticulum would increase concentrations toward the cellular interior. These regional differences could play a role in different responses associated with Ca2+ in response to mechanical stimulation.
When cells were pretreated with the intracellular Ca2+release blocker thapsigargin, the [Ca2+]i response was significantly reduced, and actin stress fiber formation together with the increase in gene expression resulting from shear stress were inhibited. Thapsigargin induces a rapid release of Ca2+ from intracellular stores but prevents reuptake of Ca2+ by inhibiting the Ca2+-ATPase of the endoplasmic reticulum (46). [Ca2+]i is transiently elevated in the cytosol, returning to baseline concentrations within 15 min of thapsigargin application (32). [Ca2+]i imaging indicated that baseline levels of [Ca2+]i were normal in thapsigargin-treated cells; however, significant reductions were seen in the mean peak [Ca2+]iresponse and the number of cells responding to shear (Table 1). The failure of thapsigargin-treated MC3T3-E1 cells to alter cytoskeletal organization or increase gene expression in response to shear indicates the necessity of intracellular Ca2+ release. Although thapsigargin-induced Ca2+ release from intracellular stores may produce signaling events within the osteoblast, no changes in the actin cytoskeleton or gene expression were observed with thapsigargin treatment alone. Measurement of [Ca2+]i with fura 2 during flow indicates that [Ca2+]i levels are increased for the duration of flow (Ryder and Duncan, unpublished observations). This prolonged increase in [Ca2+]i may be required for the changes in actin cytoskeleton and gene expression we observe in the osteoblast in response to flow. Intracellular Ca2+ release can be mediated via activation of PLC, which hydrolyzes PtdIns(4,5)P 2 to generate diacylglycerol (DAG) and IP3. IP3 binds to a receptor linked to a channel on the endoplasmic reticulum, inducing release of Ca2+. Fluid shear rapidly activates PLC-induced IP3 production in a number of cell types, including osteoblasts (39) and endothelial cells (4, 5), producing peak levels within minutes of stimulation. This elevation of IP3has been linked to PG and nitric oxide secretion in sheared osteoblasts (39). Initially, neomycin was used to inhibit the PLC pathway; however, this agent has several effects on cells, including inhibiting certain types of Ca2+ channels (19). Another, more specific inhibitor of PLC, is U-73122 (45). Both PLC blockers completely abolished actin stress fiber formation and production of c-fos and COX-2 in response to shear. U-73122 also significantly reduced the [Ca2+]i response in MC3T3-E1 cells (Table 1). Addition of U-73343, an isoform of U-73122 that binds to PLC but does not inhibit (45), did not prevent the shear-induced changes in cytoskeletal organization or increases in gene expression. DAG can also activate several messenger pathways, and DAG and Ca2+ are important in the activation of protein kinase C (35). Protein kinase C has been shown to mediate various flow-induced responses including gene expression and cytoskeletal organization (1, 4, 40). Because Ca2+ is required for protein kinase C activation, we cannot rule out that protein kinase C secondarily mediates this response, rather than the increase in [Ca2+]idirectly effecting these responses alone.
The actin cytoskeleton plays a critical role in mechanotransduction (52). In osteoblasts, expression of osteopontin induced by mechanical strain was inhibited with addition of the actin cytoskeletal poison cytochalasin D (47), and c-fos and COX-2 expression were abolished when stress fiber formation was prevented (37). Increases in [Ca2+]i can regulate proteins that have diverse effects on actin assembly. Increased [Ca2+]i can initiate actin polymerization (22) and activate actin-severing proteins such as gelsolin (44). An interaction between Ca2+ and PtdIns(4,5)P 2 on actin cytoskeletal reorganization has also been described (22), and [Ca2+]i can also a play key role in modulation of the linkage between the cytoskeleton and integrins (3). These observations, coupled with the data presented here, suggest that shear-induced increases in [Ca2+]imay mediate the increase in gene expression by modulating the actin cytoskeleton.
Although many studies have shown that in vivo and in vitro mechanical loading induces known markers of an anabolic response in bone and osteoblasts, little is known about the cascade of cellular events leading to expression and production of these markers. However, several studies have indicated the importance of c-fos and COX-2 in the osteogenic response of bone to mechanical stimulation. One of the components of the AP-1 transcription factor complex, c-fos, appears to play an important role in normal bone physiology. In transgenic mice, overexpression of c-fos induces osteosarcomas and chondrosarcomas (17,18), whereas mice lacking c-fos develop osteopetrosis (53). Typically induced within minutes by a number of osteogenic stimulants (28, 37), in vivo loading experiments have demonstrated that expression of c-fos, but not c-jun or c-myc, is induced within 2 h of mechanical loading (49). These studies indicate the importance of c-fos in normal bone physiology and the response of bone to mechanical forces.
COX-2, the inducible isoform of cyclooxygenase, is a key enzyme in the production of prostaglandins. COX-2 can be induced by a number of growth factors and cytokines (24) and has been associated with the release of PGE2 and PGI2 in bone cells in response to mechanical stimuli in vitro (25). In vivo, the importance of prostaglandins and COX-2 were demonstrated when rats, subjected to tibial four-point bending or vertebral compression, were treated with indomethacin (8) and NS-398 (15). Indomethacin inhibits both COX-1 and COX-2, whereas NS-398 is specific for COX-2. Strain-induced increases in bone formation rate were completely inhibited by NS-398. These data indicate that both c-fos and COX-2 respond quickly to mechanical loading both in vivo and in vitro and that they are essential to the mechanically induced response of bone.
In summary, we show that fluid shear-induced actin cytoskeletal reorganization and c-fos and COX-2 production is dependent on IP3-mediated intracellular Ca2+ release but is independent of extracellular Ca2+ entry via membrane ion channels in MC3T3-E1 cells. Although we have yet to link changes in expression of c-fos and COX-2 to increases in expression and production of bone markers for osteogenesis, these data suggest that changes in [Ca2+]i in response to mechanical stimulation may play a pivotal role in the mechanical signaling cascade. However, it should be noted that changes in [Ca2+]i may not be sufficient to produce osteogenic responses in the osteoblast alone. Cellular signaling mechanisms, such as kinase activation localized to the focal adhesion attachment sites, may be required as well to produce the complete phenotypic response of osteoblasts to fluid shear.
We thank Dr. Yeou-Fang Hsieh for the installation of the fluid flow apparatus and Dr. Nasser E. Ajubi for helpful comments and discussion.
Address for reprint requests and other correspondence: R. L. Duncan, Dept. of Orthopaedic Surgery, Indiana Univ. School of Medicine, 541 Clinical Dr., Suite 600, Indianapolis, IN 46202-5111 (E-mail:).
This work was supported by NASA Grant NAG5–4917 (to R. L. Duncan), National Institute of Arthritis and Musculoskeletal and Skin Diseases Grant AR-43222 (to R. L. Duncan), and National Institutes of Health Musculoskeletal Training Grant T32, AR-07581 (to D. B. Burr).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.
- Copyright © 2000 the American Physiological Society