JOURNAL OF BONE AND MINERAL RESEARCH Volume 17, Number 4, 2002 © 2002 American Society for Bone and Mineral Research

The Mode of Mechanical Integrin Stressing Controls Intracellular Signaling in Osteoblasts ¨ RR,1 BARBARA NEBE,1 HAGEN POMMERENKE,1,2 CHRISTIAN SCHMIDT,1,2 FRIEDA DU 1 1 ¨ ¨ FRANK LUTHEN, PETRA MULLER, and JOACHIM RYCHLY1

ABSTRACT Following the idea that integrin receptors function as mechanotransducers, we applied defined physical forces to integrins in osteoblastic cells using a magnetic drag force device to show how cells sense different modes of physical forces. Application of mechanical stress to the ␤1-integrin subunit revealed that cyclic forces of 1 Hz were more effective to stimulate the cellular calcium response than continuous load. Cyclic forces also induced an enhanced cytoskeletal anchorage of tyrosine-phosphorylated proteins and increased activation of the focal adhesion kinase (FAK) and mitogen activated protein (MAP) kinase. These events were dependent on an intact cytoskeleton and the presence of intracellular calcium. Analyses of the intracellular spatial organization of the calcium responses revealed that calcium signals originate in a restricted region in the vicinity of the stressed receptors, which indicates that cells are able to sense locally applied stress on the cell surface via integrins. The calcium signals can spread throughout the cell including the nucleus, which shows that calcium also is a candidate to transmit mechanically induced information into different cellular compartments. (J Bone Miner Res 2002;17:603– 611) Key words:

mechanical forces, integrin, calcium, signal transduction, osteoblast

INTRODUCTION are a fundamental physiological factor in bone and may be the principal functional determinant of adult bone mass.(1,2) Experiments have revealed that physical forces act directly on the cellular level and bone cells are able to respond physiologically to mechanical stress.(3–5) It was shown further that cyclic mechanical loading appeared to be more effective to induce proliferation and gene expression than static loading.(6) In all these experiments, different types of cell deformation were induced by stretching or compressing the whole cell, which did not reveal any information on how the cell is able to perceive and transduce mechanical stimuli. Therefore, despite accu-

P

HYSICAL FORCES

The authors have no conflict of interest.

mulating data of cellular responses due to mechanical stress, which support the physiological relevance of physical forces, the molecular basis explaining how cells sense mechanical stimuli and realize signal transduction remains obscure. In the tissue, the extracellular matrix may represent the site where external forces are transmitted to the cell. Therefore, integrin receptors that mediate the interaction between cells and the extracellular matrix(7) are supposed to act as mechanotransducers.(8 –10) This view is supported by the findings that integrins can be induced to form a physical link to the cytoskeleton and therefore are able to transmit forces to intracellular structures.(11,12) Recently, we have developed a method that enables the application of drag forces on specific cell surface receptors.(13) We have found that mechanical stressing of the integrin subunits ␤1 or ␣2, with forces in physiological ranges, induced a significant

1

Department of Internal Medicine, University of Rostock, Rostock, Germany. These authors contributed equally to this work.

2

603

604

POMMERENKE ET AL.

intracellular calcium response in osteoblasts, which was not observed when a nonadhesion receptor was stressed and indicates that calcium plays a role in the regulation of mechanotransduction.(13) Intracellular calcium is a key signaling molecule in pathways induced by a variety of external factors. It acts as a regulator of different cellular processes including gene expression and cell proliferation.(14) Therefore, we were interested in whether intracellular calcium and the subsequent activation of further signaling events represent a molecular basis to sense different modes of integrin-mediated mechanical loading.

MATERIALS AND METHODS Cell preparation In general, cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies, Eggenstein, Germany) containing 10% fetal calf serum at 37°C and in a 5% CO2 atmosphere. For analyses of intracellular calcium, primary osteoblastic cells were used. Trabecular bone from the femoral head of orthopedic patients was obtained from the Orthopedic Clinic of the University of Rostock, Rostock, Germany (Dr. T. Schuhr). The bone pieces were minced and collected in DMEM. To 1 vol of this suspension, 1 vol of collagenase (2 mg/ml) and two parts of dispase (2.4 U/ml) were added and cultured for 2 h under continuous shaking. The supernatant containing osteoblastic cells was centrifugated and cells were plated into plastic culture vessels and then passaged twice before being seeded into wells of a 96-well Fluoro Nunc module or a chamber coverglass (Nunc A/S, Roskilde, Denmark). For biochemical analyses, the osteosarcoma cell line U-2 OS was used. One hundred microliters of cells in the medium containing 1 ⫻ 105 cells was seeded into wells of a 96-well culture module and grown to near confluence. Two hours before mechanical strain was applied, the cells were depleted of serum.

Preparation of beads Preparation of beads and incubation are described elsewhere.(13) In brief, paramagnetic microbeads (2.8 ␮m in size), then streptavidin or sheep anti-mouse antibody coated (Dynal, Hamburg, Germany) were used. These beads were coated with biotin-labeled anti-integrin ␤1 antibodies (Southern Biotechnologies Associates, Inc., Birmingham, AL, USA) or with unlabeled anti-integrin ␤1 antibodies (both are nonblocking; Immunotech, Hamburg, Germany),(15) respectively, or for controls with anti-CD71 (transferrin receptor) antibody (Immunotech). Fifty microliters of phosphate-buffered saline (PBS) containing beads (25 ␮g) were added to the cell monolayer in each well and incubated for 20 minutes at room temperature. On average, 2–7 microbeads were attached to the ␤1-integrin subunit on the dorsal surface of one cell.

Application of physical stress The magnetic device has been described in detail.(13) Briefly, the device consists of a coil system containing a

ferrite core of which the two poles are modeled differently to generate an inhomogeneous magnetic field. A culture well containing the prepared cells was located between the two poles of the device. Drag forces in the horizontal direction act on the magnetic beads that are attached to the receptors on the cell surface. The forces subjected to one bead were adjusted to 2 ⫻ 10⫺10N. For calcium measurements, the device was mounted on a stage of an inverted confocal laser scanning microscope (LSM-410; Carl Zeiss Jena, Germany). A continuous or cyclic stress was applied for 10 minutes. The frequency of the cyclic stress was 1 Hz (0.5 s on and 0.5 s off) or 0.1 Hz (5 s on and 5 s off). For biochemical analyses, the procedure was performed as described earlier.(16) Drag forces were applied for 30 minutes in continuous or cyclic mode (1 Hz). Application of a cyclic magnetic field did not increase the temperature of the medium, as measured with a laboratory thermometer. For comparison, cells were incubated with anti-integrin antibody– coated beads for 50 minutes for clustering.

Measurement of intracellular calcium Before the cell monolayer was incubated with magnetic microbeads, cells were loaded with the Ca2⫹ indicator fluo-3 according to a modified method of Vandenberghe and Ceuppens.(17) Briefly, 50 ␮l of 0.5 ␮M of fluo-3/ acetoxymethylester and Pluronic F-127 (Molecular Probes, Inc., Eugene, OR, USA) in PBS were added to the monolayer, incubated for 20 minutes at 37°C, and diluted with HEPES buffer (1:5). During mechanical receptor stressing, the global calcium responses were detected with the confocal microscope using a 10⫻ Plan-Neofluar objective. For excitation, a 488-nm argon-ion laser was used and the emission was detected at 515 nm. By analyzing a full frame (512 ⫻ 512), images of 20 individual cells were taken every 8 s using the “time series” software. The fluorescence intensities were related to the basic level detected with unstimulated fluo-3 loaded cells. For imaging of spatial intracellular calcium signals, a 40⫻ oil immersion objective Plan-Neofluar was used. In intervals of 2 s, images were taken and the recorded fluorescence intensities were expressed in false colors representing 256 gray values.

Preparation of Triton X-100 –soluble and –insoluble fractions The cell monolayers were washed in PBS and incubated with cell extraction buffer containing 1% Triton X-100, 20 mM of imidazole, 2 mM of MgCl2, 80 mM of KCl, and 2 mM of EGTA for 5 minutes at 4°C (pH 7.8). For the analysis of the soluble fraction, the supernatants were collected and precipitated with 1% trichloracetic acid for 15 minutes on ice. The pellets were washed in ice-cold acetone, dried, and boiled in sodium dodecyl sulfate (SDS) sample buffer. The Triton-nonsoluble fractions were collected in SDS sample buffer by Laemmli containing 50 mM of Tris/ HCl, pH 8.0, 6% (vol/vol) ␤-mercaptoethanol, and 5% (wt/vol) SDS. Lysates were subjected to gel electrophoresis as described below.

MECHANICAL INTEGRIN STRESSING

FIG. 1. Characteristic intracellular global calcium reactions of a representative osteoblastic cell during application of physical forces to the ␤1-integrin subunit with a frequency of 1 Hz for 10 minutes (⽧). The onset of the application of forces was at time zero and the changes in relative fluorescence intensities were recorded over the indicated time. Controls: anti–␤1-integrin coated beads but was not exposed to drag forces (f), cells without beads exposed to the magnetic field (similar line but not shown).

Immunoprecipitation To analyze activation of the focal adhesion kinase (FAK) and mitogen-activated protein (MAP) kinases such as ERK-1 and ERK-2, cells were lysed in precipitation buffer containing 50 mM of Tris/HCl, pH 7.4, 100 mM of NaCl, 50 mM of NaF, 40 mM of glycerophosphate, 5 mM of EDTA, 1 mM of sodium orthovanadate, 100 ␮M of phenylmethylsulfonyl fluoride (PMSF), 1 ␮M of leupeptin, 1 ␮M of pepstatin A, and 0.1% (vol/vol) Triton X-100 and clarified by centrifugation at 13,000g for 5 minutes at 4°C. The supernatant was incubated with 100 ␮l of anti–ERK-1 (p44) antibody (clone C-16), which also detects ERK-2, or 100 ␮l of anti-FAK antibody (both from Santa Cruz Biotechnology, Inc. Santa Cruz, CA, USA) for 1 h at room temperature followed by adding 50 ␮l of protein A agarose. After centrifugation, the pellet was washed in precipitation buffer and analyzed for phosphotyrosine, FAK, and MAP kinases by immunoblotting as described in the following section.

Gel electrophoresis and immunoblotting The samples were subjected to a 7.5% SDSpolyacrylamide gel electrophoresis (PAGE). The proteins were transferred to polyvinylidene difluoride (PVDF) membranes. To block nonspecific binding, membranes were incubated with buffer containing 5% milk powder. Immunoblotting for phosphotyrosine and different proteins was performed with alkaline phosphatase–labeled anti-phosphotyrosine antibody (dilution, 1:20,000), antiFAK (clone A-17), anti–ERK-1, and anti-actin (clone C-11; all diluted 1:1,000; all from Santa Cruz Biotechnology), respectively. To analyze activation of MAP kinases by the mobility shift assay, the precipitates were electrophoresed using a 14% SDS-polyacrylamide gel and subjected to immunoblotting with anti–ERK-1 (p44). All immunoblots were visualized with chemiluminescence (CDP star; Roche Molecular Biochemicals, Mannheim, Germany).

605

FIG. 2. Global intracellular calcium responses are dependent on the strength of the mechanical load applied to the ␤1-integrin. Analyses of the maximal calcium amplitudes of individual osteoblasts during a cyclic stress for 10 minutes (n ⫽ 13–20 cells). Note that the lowest force we were able to apply induced a calcium signal. Significantly higher signals were obtained with forces of 2 ⫻ 10⫺10N/bead. Controls: ⫺, untreated cells; ⫺/1 Hz, cells without beads but exposed to magnetic field.

To compare total protein levels associated with the cytoskeleton after different treatments of the cells, aliquots of Triton X-100 –nonsoluble fractions were subjected to a 14% electrophoresis gel. Then, the gel was fixed with 70% acetic acid for 2 h followed by staining with 1% Coomassie brilliant blue G 250 in 70% acetic acid for 2 h. To remove the background staining, the gel was washed in 50% acetic acid overnight.

Treatment with cytochalasin and calcium chelator To disrupt the actin filaments of the cytoskeleton, the cell monolayer was treated with 25 nM of cytochalasin D (Sigma, Deisenhofen, Germany) for 20 minutes at 37°C. Cells were washed and incubated with microbeads to perform the procedure for mechanical loading as described. For chelating intracellular calcium, the cells were preincubated with 5 ␮M of 1,2-bis-(o-aminophenoxy)-ethaneN,N,N⬘,N⬘-tetracetic acid, acetoxymethyl ester (BAPTAAM) for 15 minutes. Mechanical strain was then applied in the presence of 5 ␮M of BAPTA.

Statistics Data for global intracellular calcium responses of individual cells were presented as mean ⫾ SD and significance of different treatments were determined by analysis of variance (ANOVA) followed by unpaired Student’s t-test.

RESULTS Differential calcium responses by continuous and cyclic integrin loading Cells were stressed at the ␤1-integrin subunit and the global calcium signals of individual cells were recorded

606

POMMERENKE ET AL.

microscopically over a period of 10 minutes. The majority of cells responded with oscillating calcium signals, which had a maximum of eight spikes during the observation time regardless of whether a continuous or cyclic stress was applied. In Fig. 1, a characteristic calcium response of a cell is shown showing spikes with a duration of ⬃1 minute. In controls, when anti-integrin antibody– coated beads were incubated without subsequent mechanical stress or when pure cells without beads were subjected to the magnetic field, no significant calcium signals were detected. To test which pulling strength is required to induce a calcium signal, we reduced the forces applied to integrins (Fig. 2). The results revealed that the minimal force we were able to apply with our device (i.e., 2.5 ⫻ 10⫺11N/bead) was sufficient to induce a calcium response. Application of a cyclic stress with 1 Hz compared with a continuous stress revealed two differences (Fig. 3). First, mechanical stress evoked a higher maximal calcium reaction because of a 1-Hz stress (Fig. 3A), and, second, the onset of the calcium reaction was earlier (Fig. 3B). An influence on the frequency of the calcium spikes was not observed in all the experiments (Fig. 3C). The results indicate that the calcium response is controlled by the mode of application, that is, whether a continuous or a cyclic mechanical stress is applied. In an additional experiment, we compared the effect of 0.1 Hz with 1 Hz and a continuous stress. This experiment showed that compared with a 1-Hz stress, application of a stress with 0.1 Hz provoked a maximal calcium response of 58%, which was similar to a continuous stress (52% of the reaction with 1 Hz; data not shown).

Enhanced cytoskeletal anchorage of tyrosinephosphorylated proteins by cyclic mechanical stress

FIG. 3. Analyses of the intracellular global calcium responses during continuous and cyclic mechanical stress (1 Hz) applied to the ␤1integrin subunit. The time courses of calcium responses in individual cells (as shown in an example in Fig. 1) were analyzed with respect to maximal calcium amplitude, the onset of the first calcium signal, and the number of calcium spikes. Data of three independent experiments (1–3; n ⫽ 20 cells) were evaluated. (A) Maximal calcium amplitudes (in general the first spike). Significantly higher calcium amplitudes during cyclic integrin stress were obtained in experiments 1, 2, and 3 (p ⱕ 0.05, p ⱕ 0.001, and p ⱕ 0.001, respectively). (B) Time of the initial calcium response after the onset of the mechanical loading. Significantly earlier initial calcium signals were obtained during cyclic stress in all three experiments (p ⱕ 0.001) (C) Number of the calcium spikes during the experimental time. Significant differences only in experiment 3 (p ⱕ 0.01).

The results revealed that cyclic mechanical forces with a frequency of 1 Hz at the ␤1-integrin subunit induced a higher amount of tyrosine-phosphorylated proteins in the cytoskeletal fraction than static forces (Fig. 4A). Coomassie staining of total proteins after Triton X-100 extraction (Fig. 4B) indicated that the increased tyrosine phosphorylation was caused by an immobilization of these activated proteins, predominantly in the higher molecular weight range. A Western blot experiment (Fig. 4C) identified FAK, a representative protein in integrin-mediated signaling, in the insoluble fraction because of integrin stress and most effectively after cyclic stress. The immobilization of FAK to the cytoskeleton was confirmed by the shift of FAK from the soluble to the nonsoluble fraction (Fig. 4D). Disruption of the cytoskeleton by cytochalasin D supported the finding that the actin cytoskeleton serves as a structure for the immobilization of activated proteins because of integrin stress (Figs. 4A– 4C). This process requires intracellular calcium as shown by treatment with the calcium chelator BAPTA-AM (Figs. 4A– 4C). Together, the results suggest that the mode of physical integrin stress regulates the cytoskeletal immobilization of activated proteins.

MECHANICAL INTEGRIN STRESSING

FIG. 4. Cytoskeletal anchorage of proteins caused by differential mechanical loading of ␤1-integrin. (A) Tyrosine-phosphorylated proteins in the Triton X-100 –insoluble fraction: cyclic stress with 1 Hz (s or 1Hz), permanent stress (p), clustering with anti-integrin antibody coated beads (c), or not treated (⫺). For comparison, the CD71 receptor was mechanically stressed or cells were pretreated with BAPTA-AM (BAPTA) and cytochalasin D (CCD), respectively. Controls: cells without beads were exposed in the magnetic field to a frequency of 1 Hz and to a constant field (p). Note a higher level of phosphotyrosine because of cyclic stress. Stressing of the CD71 receptor and exposure of untreated cells to the magnetic field had no effect. BAPTA-AM and cytochalasin D blocked the enhanced tyrosine phosphorylation. Representative blot of three independent experiments. (B) Total protein content in the Triton X-100 –insoluble fraction after different treatments of the cells as described in panel A. The highest protein content in the cytoskeletal fractions was obtained because of cyclic mechanical integrin loading (1 Hz). Note that the immobilization was selective and proteins of the higher molecular weight range became anchored to the cytoskeleton (M, molecular weight marker). (C) FAK and actin content in the Triton X-100 –insoluble fraction after different treatments of the cells as described in panel A. The highest level of FAK was detected because of cyclic stress to ␤1-integrins (1 Hz). Stress to the CD71 receptor had no effect, and BAPTA-AM and cytochalasin D blocked the immobilization of FAK. The level of actin remained unaffected because of the different treatments. (D) Movement of FAK from the Triton X-100 –soluble fraction to the Triton X-100 –nonsoluble fraction. Cells were treated as described in panel A. In controls (⫺), after clustering (c) and permanent stressing (p) of the integrins, FAK was detected predominantly in the soluble fraction. After application of cyclic stress (1 Hz), FAK was detected in the insoluble fraction, which was paralleled with a loss of FAK in the soluble fraction.

607

FIG. 5. Activation of FAK and MAP kinases because of differential mechanical loading of the ␤1-integrin. Treatments are described in Fig. 4. (A) Tyrosine phosphorylation of FAK; (B) tyrosine phosphorylation of MAP kinases. Note that the most profound tyrosine phosphorylation of FAK and MAP kinases (ERK-2) was obtained due to a cyclic integrin stress. No effect due to mechanical stressing of the CD71 receptor and exposure of untreated cells to the magnetic field. BAPTA-AM and cytochalasin D blocked the enhanced activation. (C) Mobility shift assay for MAP kinases. Note that cyclic forces induced the slow migrating fraction of activated ERK-2 protein (ERK-P*). Representative blots of three experiments.

Enhanced activation of FAK and MAP kinases by cyclic integrin loading Activation of FAK and MAP kinases are significant events in integrin signaling. Our results revealed that cyclic forces were more effective to activate both FAK and MAP kinase ERK-2 than a continuous load detected by tyrosine phosphorylation of these proteins (Figs. 5A and 5B). Enhanced activation of MAP kinases caused by cyclic forces was confirmed also by the mobility shift in SDS-PAGE (Fig. 5C). Clustering of the receptors by incubation with beads had a lower effect and stressing the transferrin recep-

608

POMMERENKE ET AL.

FIG. 6. Imaging of the intracellular calcium organization during a cyclic mechanical stress of 1 Hz to the ␤1-integrin subunit. Typical time course of the cytosolic calcium propagation in a representative cell. Beads that are attached to the receptors are visible as yellow dots (autofluorescence). The onset of the application of mechanical stress is at time zero. After 16 s, a calcium signal initiates near a bead (arrow 1). After 18 s, two other beads induce a local calcium response (arrows 2 and 3). The calcium waves spread over the cell with a climax at 24 s and then decline. The images are presented in pseudocolor. The blue color represents the lowest and white represents the highest concentration of calcium. The time in seconds is indicated in the images (bar ⫽ 10 ␮m).

tor had no effect. Tyrosine phosphorylation of both FAK and MAP kinases was blocked completely in the presence of the Ca2⫹ chelator BAPTA-AM, which shows the role of intracellular calcium. Similarly, using cytochalasin D, the results indicate the requirement of an intact cytoskeleton for activation of FAK and MAP kinases.

Application of local mechanical stress determines the spatial origin and induces a transcellular spreading of calcium To further evaluate the role of intracellular calcium, we investigated the spatiotemporal calcium responses in individual cells by image analysis. Figure 6 illustrates a representative calcium reaction in an osteoblast. The beads that stress the receptors indicate the localization of the mechanical load. The calcium signals were initiated only near beads that stressed integrins. This indicates that for the local origin of the calcium signal, the presence of a bead was required. The calcium reaction of the cell shows that a primary calcium impulse may originate independently in different restricted regions of the cell. First, the calcium signal arose at the upper margin of the cell (Fig. 6, arrow 1), visible after 14 s of stress application. Although this calcium wave spreads two other

waves originated in different regions of the cell after 18 s (Fig. 6, arrows 2 and 3). As shown in this cell, in the majority of the cells, the individual waves spread over an extended area and combined to fill the entire cell interior. The climax of the calcium reaction may be accompanied by an accumulation of the highest concentration of calcium in the nucleus. The following decline of the calcium level in the nucleus paralleled that of the cytoplasm. To analyze the temporal course of the calcium response in different regions of the cell, we set four gates and determined the time-dependent alterations in calcium concentration (Fig. 7). These data show that the individual local calcium reactions do not have a simultaneous timing. In addition, as shown in two regions (blue and red areas), a lower calcium response may be initiated and then declined again before the onset of the main signal. Analyses of the spatiotemporal calcium reactions further revealed that in ⬃10% of the cells, the calcium signals remained restricted to the region where the calcium waves were initiated. An example is shown in Fig. 8, where a calcium signal originated due to an integrin stress in a very restricted region directly at the location of the bead after 12 s with a maximum after 20 s. The signal then declined again without spreading over the cell.

MECHANICAL INTEGRIN STRESSING

609

FIG. 7. Time courses of calcium responses in different regions of the cell during mechanical integrin stress. Four gates were set into the cell shown in Fig. 6. (A) Time courses of the calcium responses in these marked regions (gates). (B) The onset of the mechanical stress is at time zero. The curves show different intensities of the calcium responses in the four regions. It is further notable that the spikes of the signals are not synchronous (e.g., the cyan curve precedes the red curve). In the red and blue areas, transient calcium signals occurred before the onset of the main signals (bar ⫽ 10 ␮m).

DISCUSSION Our experiments revealed that a physical stress to integrins already with a low intensity evoked significant calcium signals. This confirms previous findings that show intracellular calcium is an early response in integrin signaling.(18) Most of these studies examined integrin-mediated calcium during cell spreading or adhesion to other cells that probably is a more complex mechanism that involves cell shape changes.(19,20) In our experiments we could not observe, at least by light microscopy that the cell membrane was distorted, although the forces applied at the receptorbound beads might induce a small highly localized distortion of the membrane. Previous studies have shown that the integrin- and stress-mediated rise of intracellular calcium is caused by both release from intracellular stores and an extracellular entry.(21,22) Calreticulin, which is able to interact with cytoplasmic domains of ␣-integrin subunits, is essential for integrin-mediated calcium entry and mediates adhesion.(23,24) When we compared the calcium reactions in individual cells during continuous and cyclic integrin stress with a

FIG. 8. Spatially restricted calcium response in a representative cell because of mechanical integrin stress. The yellow dot represents the bead attached to the ␤1-integrin subunit. The onset of the mechanical load is at time zero. After 12 s, an increase in the intracellular calcium response is visible with a maximum after 20 s (black area; arrow). This response declines and remains locally restricted. The images represent the substraction of the image before the onset of the mechanical load from the images during the mechanical load (bar ⫽ 30 ␮m).

frequency of 1 Hz, we found that cyclic stress was a stronger stimulus. Controls without beads but exposed to a cyclic magnetic field revealed no effect on the calcium reaction, which indicates that possible random electrical currents are not the cause for an increased calcium reaction because of cyclic forces. A lower frequency of 0.1 Hz had a similar effect as a continuous stress, which is consistent

610

with studies of mechanical loading of rat tibias showing no effect on bone formation at frequencies below 0.5 Hz but increasing induction of bone mass up to 2 Hz.(25) The actin cytoskeleton appears to be of major importance for integrin-mediated mechanotransduction.(10,26) A mechanical coupling between integrins and the cytoskeleton has been established(11,27) and, furthermore, the cytoskeleton serves as a structural scaffold to assemble signaling molecules to interact in biochemical reactions.(28,29) Here, we could show that a cyclic integrin stress is more effective to anchor tyrosinephosphorylated proteins at the cytoskeleton than a continuous stress. These proteins also included the FAK, which is essential for the formation of the cytoskeletal signaling structures and their functional activities because FAK recruits SH2 and SH3 domain– containing signaling proteins and is required for further signaling.(30,31) Earlier immunofluorescence analyses revealed that the accumulation of cytoskeletally associated proteins was strongest at the site of the stressing beads, but also in the ventral regions of the cell, a visible assembly of proteins was detected.(32) The cytoskeletal association of activated proteins depended on intracellular calcium, which stresses the relevance of calcium for mechanotransduction and forming a cytoskeletal signaling complex. In addition, this is supported by the finding that cytoskeletal linkage of ␤1-integrin was blocked by chelating of intracellular calcium.(12) Previous studies have shown that application of physical stress to cells induced an activation of FAK and MAP kinases.(9,33) We showed that mechanical stress applied to a defined integrin receptor activated FAK and MAP kinases and the magnitude was dependent on the mode of stress. MAP kinase activation could be dependent on cytoskeletally associated FAK because cytochalasin treatment inhibited both activation of FAK and MAP kinases, which supports earlier findings.(34) However, experiments also have indicated that activation of MAP kinases can be independent of FAK activation.(35) Analyses of the spatiotemporal distribution of calcium inside the cell because of a mechanical integrin stress revealed three main results: (1) the site of application of physical forces determined the spatial origin of the evoked calcium response; (2) calcium may spread over the cell including the nucleus, whereas spatially restricted spikes could precede the spreading of calcium over the entire cell; and (3) the calcium signal could remain confined to a narrow region in the cytoplasm. One conclusion is that cells are able to sense the localization of the applied stress on the cell surface. This is supported by findings that have shown a spatially restricted transmission of mechanical forces to the cytoskeleton using an optical trap(36) and highly localized responses of the actin and microtubule cytoskeleton to applied stress.(27) This local mechanotransduction appears also to have local physiological consequences because mechanical stress to the cell surface induced a relocation of the apparatus for protein synthesis to the site of signal reception.(37) Spreading of calcium caused by mechanical integrin loading in our experiments showed that calcium is an excellent candidate to transmit information into different compart-

POMMERENKE ET AL.

ments of the cell interior. Calcium can act directly at target sites in different cellular compartments.(38,39) The physiological relevance of calcium in the nucleus was shown in an elegant study showing that gene expression is controlled differently by nuclear and cytoplasmic calcium.(40) On the other hand, mechanical stress applied at the cell surface also can be transmitted by a direct mechanical coupling to the nucleus, which was mediated by a linkage between cytoskeleton and nucleus.(41) Although our results suggest a key role of intracellular calcium in the signaling pathway of mechanical integrin stress, it is reasonable that multiple routes exist that may not depend on calcium. For example, intracellular calcium induction by ligation of ␣v-integrin did not contribute to cell adhesion, whereas ␣5␤1 mediated cell adhesion without a calcium reaction.(20) Taken together, our results indicate that mechanical loading of integrins, notably, the mode of the stress controlled integrin signaling at different levels of the cascade by quantitative modulation of signaling events. We suggest that such quantitative modulations imply functional consequences that were supported by experiments in myoblasts, which established that quantitative changes in integrinmediated activation of paxillin, FAK, and MAP kinases decided whether the cells proliferated or withdrew from the cell cycle.(42)

ACKNOWLEDGMENT This work was supported by a grant from Bundesminister fu¨ r Bildung, Forschung und Technologie (BMBF; 01ZZ9601), and by a grant from the Deutsche Forschungsgemeinschaft (GK-Br 1255/4 –1).

REFERENCES 1. Frost HM 1992 The role of changes in mechanical usage set points in the pathogenesis of osteoporosis. J Bone Miner Res 7:253–261. 2. Duncan RL, Turner CH 1995 Mechanotransduction and the functional response of bone to mechanical strain. Calcif Tissue Int 57:344 –358. 3. Banes AJ, Tsuzaki M, Yamamoto J, Fischer T, Brigman B, Brown T, Miller L 1995 Mechanoreception at the cellular level: The detection, interpretation, and diversity of responses to mechanical signals. Biochem Cell Biol 73:349 –365. 4. Moalli MR, Caldwell NJ, Patil PV, Goldstein SA 2000 An in vivo model for investigations of mechanical signal transduction in trabecular bone. J Bone Miner Res 15:1346 –1353. 5. Wozniak M, Fausto A, Carron CP, Meyer DM, Hruska KA 2000 Mechanically strained cells of the osteoblast lineage organize their extracellular matrix through unique sites of alphavbeta3-integrin expression. J Bone Miner Res 15:1731– 1745. 6. Carano A, Siciliani G 1996 Effects of continuous and intermittent forces on human fibroblasts in vitro. Eur J Orthod 18:19 –26. 7. Hynes RO 1992 Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 69:11–25. 8. Ingber D 1991 Integrins as mechanochemical transducers. Curr Opin Cell Biol 3:841– 848. 9. MacKenna DA, Dolfi F, Vuori K, Ruoslahti E 1998 Extracellular signal-regulated kinase and c-Jun NH2-terminal kinase

MECHANICAL INTEGRIN STRESSING

10.

11.

12.

13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

activation by mechanical stretch is integrin-dependent and matrix- specific in rat cardiac fibroblasts. J Clin Invest 101: 301–310. Pavalko FM, Chen NX, Turner CH, Burr DB, Atkinson S, Hsieh YF, Qiu J, Duncan RL 1998 Fluid shear-induced mechanical signaling in MC3T3–E1 osteoblasts requires cytoskeleton-integrin interactions. Am J Physiol 275:C1591– C1601. Wang N, Butler JP, Ingber DE 1993 Mechanotransduction across the cell surface and through the cytoskeleton. Science 260:1124 –1127. Nebe B, Bohn W, Sanftleben H, Rychly J 1996 Induction of a physical linkage between integrins and the cytoskeleton depends on intracellular calcium in an epithelial cell line. Exp Cell Res 229:100 –110. Pommerenke H, Schreiber E, Durr F, Nebe B, Hahnel C, Moller W, Rychly J 1996 Stimulation of integrin receptors using a magnetic drag force device induces an intracellular free calcium response. Eur J Cell Biol 70:157–164. Berridge MJ 1993 Inositol trisphosphate and calcium signalling. Nature 361:315–325. Lehmann M, Rabenandrasana C, Tamura R, Lissitzky JC, Quaranta V, Pichon J, Marvaldi J 1994 A monoclonal antibody inhibits adhesion to fibronectin and vitronectin of a colon carcinoma cell line and recognizes the integrins alpha v beta 3, alpha v beta 5, and alpha v beta 6. Cancer Res 54:2102–2107. Schmidt C, Pommerenke H, Durr F, Nebe B, Rychly J 1998 Mechanical stressing of integrin receptors induces enhanced tyrosine phosphorylation of cytoskeletally anchored proteins. J Biol Chem 273:5081–5085. Vandenberghe PA, Ceuppens JL 1990 Flow cytometric measurement of cytoplasmic free calcium in human peripheral blood T lymphocytes with fluo-3, a new fluorescent calcium indicator. J Immunol Methods 127:197–205. Sjaastad MD, Nelson WJ 1997 Integrin-mediated calcium signaling and regulation of cell adhesion by intracellular calcium. Bioessays 19:47–55. Schwartz MA 1993 Spreading of human endothelial cells on fibronectin or vitronectin triggers elevation of intracellular free calcium. J Cell Biol 120:1003–1010. Schwartz MA, Denninghoff K 1994 Alpha v integrins mediate the rise in intracellular calcium in endothelial cells on fibronectin even though they play a minor role in adhesion. J Biol Chem 269:11133–11137. Glogauer M, Ferrier J, McCulloch CA 1995 Magnetic fields applied to collagen-coated ferric oxide beads induce stretchactivated Ca2⫹ flux in fibroblasts. Am J Physiol 269:C1093– C10104. Sjaastad MD, Lewis RS, Nelson WJ 1996 Mechanisms of integrin-mediated calcium signaling in MDCK cells: Regulation of adhesion by IP3- and store-independent calcium influx. Mol Biol Cell 7:1025–1041. Coppolino MG, Woodside MJ, Demaurex N, Grinstein S, St-Arnaud R, Dedhar S 1997 Calreticulin is essential for integrin-mediated calcium signalling and cell adhesion. Nature 386:843– 837. Kwon MS, Park CS, Choi K, Ahnn J, Kim JI, Eom SH, Kaufman SJ, Song WK 2000 Calreticulin couples calcium release and calcium influx in integrin-mediated calcium signaling. Mol Biol Cell 11:1433–1443. Turner CH, Forwood MR, Otter MW 1994 Mechanotransduction in bone: Do bone cells act as sensors of fluid flow? FASEB J 8:875– 878. Miyamoto S, Teramoto H, Coso OA, Gutkind JS, Burbelo PD, Akiyama SK, Yamada KM 1995 Integrin function: Molecular hierarchies of cytoskeletal and signaling molecules. J Cell Biol 131:791– 805.

611 27. Heidemann SR, Kaech S, Buxbaum RE, Matus A 1999 Direct observations of the mechanical behaviors of the cytoskeleton in living fibroblasts. J Cell Biol 145:109 –122. 28. Yamada KM, Miyamoto S 1995 Integrin transmembrane signaling and cytoskeletal control. Curr Opin Cell Biol 7:681– 689. 29. Burridge K, Chrzanowska-Wodnicka M 1996 Focal adhesions, contractility, and signaling. Annu Rev Cell Dev Biol 12:463– 518. 30. Richardson A, Parsons JT 1995 Signal transduction through integrins: A central role for focal adhesion kinase? Bioessays 17:229 –236. 31. Sieg DJ, Hauck CR, Schlaepfer DD 1999 Required role of focal adhesion kinase (FAK) for integrin-stimulated cell migration. J Cell Sci 112:2677–2691. 32. Rychly J, Pommerenke H, Durr F, Schreiber E, Nebe B 1998 Analysis of spatial distributions of cellular molecules during mechanical stressing of cell surface receptors using confocal microscopy. Cell Biol Int 22:7–12. 33. Toma CD, Ashkar S, Gray ML, Schaffer JL, Gerstenfeld LC 1997 Signal transduction of mechanical stimuli is dependent on microfilament integrity: Identification of osteopontin as a mechanically induced gene in osteoblasts. J Bone Miner Res 12:1626 –1636. 34. Schlaepfer DD, Hunter T 1997 Focal adhesion kinase overexpression enhances ras-dependent integrin signaling to ERK2/ mitogen-activated protein kinase through interactions with and activation of c-Src. J Biol Chem 272:13189 –13195. 35. Lin TH, Aplin AE, Shen Y, Chen Q, Schaller M, Romer L, Aukhil I, Juliano RL 1997 Integrin-mediated activation of MAP kinase is independent of FAK: Evidence for dual integrin signaling pathways in fibroblasts. J Cell Biol 136:1385– 1395. 36. Choquet D, Felsenfeld DP, Sheetz MP 1997 Extracellular matrix rigidity causes strengthening of integrin- cytoskeleton linkages. Cell 88:39 – 48. 37. Chicurel ME, Singer RH, Meyer CJ, Ingber DE 1998 Integrin binding and mechanical tension induce movement of mRNA and ribosomes to focal adhesions. Nature 392:730 –733. 38. Thorn P, Lawrie AM, Smith PM, Gallacher DV, Petersen OH 1993 Local and global cytosolic Ca2⫹ oscillations in exocrine cells evoked by agonists and inositol trisphosphate. Cell 74: 661– 668. 39. Nelson MT, Cheng H, Rubart M, Santana LF, Bonev AD, Knot HJ, Lederer WJ 1995 Relaxation of arterial smooth muscle by calcium sparks. Science 270:633– 637. 40. Hardingham GE, Chawla S, Johnson CM, Bading H 1997 Distinct functions of nuclear and cytoplasmic calcium in the control of gene expression. Nature 385:260 –265. 41. Maniotis AJ, Chen CS, Ingber DE 1997 Demonstration of mechanical connections between integrins, cytoskeletal filaments, and nucleoplasm that stabilize nuclear structure. Proc Natl Acad Sci USA 94:849 – 854. 42. Sastry SK, Lakonishok M, Wu S, Truong TQ, Huttenlocher A, Turner CE, Horwitz AF 1999 Quantitative changes in integrin and focal adhesion signaling regulate myoblast cell cycle withdrawal. J Cell Biol 144:1295–1309.

Address reprint requests to: Dr. Joachim Rychly Department of Internal Medicine Ernst-Heydemann-Str. 6 University of Rostock 18055 Rostock, Germany Received in original form November 16, 2000; in revised form September 3, 2001; accepted November 12, 2001.