Calcium regulates cyclic compression-induced early changes in chondrocytes during in vitro cartilage tissue formation

Cell Calcium 48 (2010) 232–242

Contents lists available at ScienceDirect

Cell Calcium journal homepage: www.elsevier.com/locate/ceca

Calcium regulates cyclic compression-induced early changes in chondrocytes during in vitro cartilage tissue formation Igal Raizman, J.N. Amritha De Croos, Robert Pilliar, Rita A. Kandel ∗ CIHR-BioEngineering of Skeletal Tissue Team, Mount Sinai Hospital, University of Toronto, Toronto, Canada M5G 1X5

a r t i c l e

i n f o

Article history: Received 25 June 2010 Received in revised form 7 September 2010 Accepted 8 September 2010 Available online 6 October 2010 Keywords: Chondrocytes Tissue-engineering Nifedipine Gadolinium Calcium

a b s t r a c t A single application of cyclic compression (1 kPa, 1 Hz, 30 min) to bioengineered cartilage results in improved tissue formation through sequential catabolic and anabolic changes mediated via cell shape changes that are regulated by ␣5␤1 integrin and membrane-type metalloprotease (MT1-MMP). To determine if calcium was involved in this process, the role of calcium in regulating cell shape changes, MT1-MMP expression and integrin activity in response to mechanical stimulation was examined. Stimulation-induced changes in cell shape and MT1-MMP expression were abolished by chelation of extracellular calcium, and this effect was reversed by re-introduction of calcium. Spreading was inhibited by blocking stretch-activated channels (with gadolinium), while retraction was prevented by blocking the L-Type voltage-gated channel (with nifedipine); both compounds inhibited MT1-MMP upregulation. Calcium A23187 ionophore restored cellular response further supporting a role for these channels. Calcium regulated the integrin-mediated signalling pathway, which was facilitated through Src kinase. Both calcium- and integrin-mediated pathways converged on ERK-MAPK in response to stimulation. While both integrins and calcium signalling mediate chondrocyte mechanotransduction, calcium appears to play the major regulatory role. Understanding the underlying molecular mechanisms involved in chondrocyte mechanotransduction may lead to the development of improved bioengineered cartilage. © 2010 Elsevier Ltd. All rights reserved.

1. Introduction Articular cartilage covers the articulating ends of bones in synovial joints. Due, in part, to lack of vascular supply cartilage possesses a limited ability for self-repair, and with injury or disease fails to properly heal. Current attempts to address this issue have been only partially successful [1]. Consequently, repair of cartilage defects with in vitro-formed cartilage tissue is a promising alternative approach that would circumvent many limitations inherent in current treatment approaches. Unfortunately the biomechanical properties of bioengineered cartilage tissue are inferior to the native tissue, a factor which precludes its utilization clinically [2]. We have shown previously that the application, as short as 30 min, of compressive forces enhances collagen and proteoglycan accumulation by articular chondrocytes grown on the surface of a porous ceramic substrate [2,3]. This increase in matrix appears to be dependent on a biphasic remodelling process, which involves an initial catabolic phase (characterized by ERK activation and upregulation of MT1-MMP) followed by a subsequent anabolic response

∗ Corresponding author at: Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, 600 University Avenue, Suite 6-500, Toronto, Ontario, Canada M5G 1X5. Tel.: +1 416 586 8516; fax: +1 416 586 8719. E-mail address: [email protected] (R.A. Kandel). 0143-4160/$ – see front matter © 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.ceca.2010.09.006

which increases the amount of extracellular matrix (ECM) synthesis and accumulation [4,5]. Furthermore, it was observed that this stimulation-induced increase in matrix synthesis is dependent upon transient chondrocyte spreading and subsequent retraction, mediated through the ␣5␤1 integrin and was dependent on the upregulation of membrane-type 1 metalloproteinase (MT1-MMP), which was observed to occur immediately following mechanical stimulation [6]. However, the underlying mechanism(s) by which chondrocytes sense mechanical forces and convert them to the previously delineated pathways are poorly understood. For example, it is not known whether ERK activation occurs prior to cell spreading. As other studies have shown that calcium fluxes occur in mechanically stimulated chondrocytes [7–10], and can occur within seconds of stimulation [11,12], it has been proposed that the first signalling event in response to mechanical stimuli may involve calcium. Of the multitude of calcium channels described, chondrocytes have been shown to express the L-Type Voltage Gated Calcium Channel (VGCC) and the cationic stretch activated cationic channels (SAC) that are permissive to calcium ion fluxes [13–15]. Studies in canine chondrocytes have shown that selective inhibitors of VGCCs are able to reduce the expression of MMP-3 and -13 [16], while studies in human pulmonary artery endothelial cells have shown that calcium signalling regulates the binding activity of AP-1 (a known regulator of MT1-MMP expression) [17]. Moreover, chondrocytes appear to reorganize their actin cytoskeleton in

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

a calcium-dependent manner [18], and intracellular calcium levels have been shown to influence aggrecan, collagen types I, II, and X expression levels [8,19,20]. Interestingly, ion channels colocalize with ␤1 integrins [21,22], and some studies have shown that cross-talk between the ion channels and integrins exist which can regulate cell function [23]. As such, calcium signalling could be involved in the regulation of both the cell morphology changes, and the anabolic/catabolic response that occurs in our system. The role of calcium in mediating enhanced cartilage tissue formation in response to cyclic compression has not been elucidated. The aim of this study, therefore, was to determine whether the initial chondrocyte response that leads to the cyclic compressioninduced increase in matrix accumulation, namely cell spreading and MT1-MMP expression, are regulated by calcium. In addition, because of the putative role of ␣5␤1 integrin in this process, we investigated whether the calcium signalling pathway is involved in regulating the integrin associated kinase Src. Understanding the mechanism by which mechanical compression results in increased matrix accumulation may further help us to improve the properties of in vitro-formed cartilage tissue. 2. Methods 2.1. Cell culture Chondrocytes were isolated by sequential enzymatic digestion from bovine (6–9 months old) metacarpal–phalangeal joints as described previously [24]. Briefly, cartilage harvested from 2 to 3 joints was pooled together to obtain sufficient cells, and digested with 0.5% protease (Sigma Chemical Co., MI) for 1 h at 37 ◦ C, followed by digestion with 0.1% collagenase (Roche Molecular Biochemicals, IN) overnight under standard culture conditions (37 ◦ C, 5% CO2 ). Chondrocytes were seeded in Ham’s F-12 supplemented with 5% fetal bovine serum (Sigma Chemical Co.) at a density of 160,000 cells/mm2 on the top surface of porous substrates (calcium polyphosphate (CPP) 2 mm height × 4 mm diameter). Cultures were grown at 37 ◦ C under standard cell culture conditions [2]. 2.2. Mechanical stimulation Following 3 days of static culture, cells were subjected to uniaxial, confined cyclic compression (1 kPa, 1 Hz, 30 min, MACH-1, Biomomentum, Montreal, Canada) as described previously [3,4]. The strain applied to the tissue was estimated to be 1.4%. This was calculated by determining the difference in displacement of the mechanical tester in the presence or absence of tissue given a constant force of 1 kPa and divided by the tissue height (0.3 ± 0.06 mm). To avoid dislodging the small amount of tissue formed by this time, the cells were subjected to compressive force through a compliant 2% agarose gel disk (3.5 mm diameter × 8 mm height) that had been placed on top of the cells as described previously [3,4]. The control constructs were maintained under identical culture conditions but did not receive any mechanical stimulation. 2.3. Calcium studies To investigate the role of calcium in mechanotransduction, constructs were pre-incubated in serum-free Hams F-12 alone (as a control) or supplemented with EGTA (10 ␮M), or EGTA (10 ␮M) combined with CaCl2 (5 ␮M), for 4 h prior to mechanical stimulation. These samples were then subjected to mechanical stimulation under the conditions described above. To determine which calcium permissive channels are involved, chondrocytes were preincubated for 4 h in the absence or presence of the ion channel inhibitors, nifedipine and gadolinium (Sigma) which were

233

dissolved in DMSO and then diluted with HAMs F-12 to form a 20 ␮M solution (0.01% DMSO). The role of calcium fluxes was further confirmed by utilizing the A23187 ionophore (10 ␮M) in conjunction with the inhibitors. To investigate the involvement of Src, constructs were preincubated in the presence of PP2 (4-amino5-(4-chlorophenyl)-7-(t-butyl)pyrazolo[3,4-d]pyrimidine) or its negative control PP3 (4-amino-7-phenylpyrazol[3,4-d]pyrimidine) (20 ␮M, Calbiochem), for 4 h prior to mechanical stimulation. 2.4. Chondrocyte morphology and area Following mechanical stimulation, tissue-CPP constructs were washed three times in cold phosphate buffered saline (PBS) and fixed in 2.5% glutaraldehyde for at least an hour. Subsequently, constructs were rinsed three times in PBS, dehydrated in graded ethanol (up to 100%), critical point dried (Bal-Tec, Liechtenstein; CPD 030), and sputter coated with gold (Denton Vacuum, NJ; Desk II). The cells at the surface of the CPP substrate were visualized by secondary electron imaging under scanning electron microscopy (SEM, FEI, OR; XL30). Resulting images were imported into Image J (http://rsb.info.nih.gov/ij/) and chondrocyte area was determined by tracing the cell outline and calculating the area using the Image J program, utilizing the scale bar in the images. Only cells that were in direct contact with the substrate were measured. At least three constructs and a total of at least 100 cells were examined for each condition. 2.5. Western blot analysis The tissue was removed from the CPP and homogenized with a motorized pestle on ice (100 ␮l) in RIPA buffer A (50 mM Tris–HCl pH 7.4, 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 1 mM EDTA) supplemented with the protease inhibitors NaF (1 mM) and phenylmethylsulfonyl fluoride (1 mM) which were added immediately prior to use. Protein extracts were clarified by centrifugation for 10 min at 12,000 × g and protein content quantified utilizing a BCA protein assay kit (Pierce, Rockford, IL). Proteins from each sample (20 ␮g) were separated by sodiumdodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) (12% acrylamide, 1.5 h at 150 V), electroblotted (1.5 h, 0.3 mA) onto nitrocellulose membranes (BioTrace NT, Pall Life Sciences, Pensacola, FL) then incubated with antibodies reactive with total ERK1/2 or phospho-ERK1/2 (1:2000; Cell Signaling, Beverly, MA) in Tris-buffered saline, 0.1% Tween-20 (10 mM Tris, pH 7.5, 150 mM NaCl, 0.05% v/v Tween-20) containing 5.0% non-fat dry milk overnight at 4 ◦ C. Immunoreactivity was visualized using secondary antibody conjugated with horseradish peroxidase (1:2000, Cell Signaling) and enhanced chemiluminescence (ECL Plus; Amersham Biosciences). 2.6. Semi-quantification of gene expression by RT-PCR To determine the expression of MT1-MMP following mechanical stimulation, total RNA was extracted from stimulated and unstimulated tissues using Trizol (Invitrogen, Carlsbad, CA) according to the manufacturer’s directions. Total RNA was reverse transcribed (RT) into cDNA using reverse transcriptase (Superscript II, Invitrogen) according to the manufacturer’s instructions. Relative gene expression was examined by semi-quantitative RT-PCR using Taq Polymerase (Invitrogen) and primers in reactions designed to amplify the sequence of interest in the linear range as outlined previously [5]. 18S rRNA was the housekeeping gene against which gene expression was normalized. PCR products were separated on a 1.6%-agarose gel stained with ethidium bromide. Band intensity was semi-

234

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

Fig. 1. Characterization of chondrocyte response to cyclic compression. Changes in (A and B) Erk phosphorylation (A: densitometry; B: Western blot); (C) cell morphology as visualized by scanning electron microscopy and (D) chondrocyte cell area over time during cyclic compression. *Statistically significant different (p = 0.05) as compared to unstimulated control constructs at the respective time points. Results are expressed as mean ± SEM (n = 6).

quantified by densitometry using Lab Works software (V4.0, Media Cybernectics). 2.7. Statistical analysis All conditions were performed in at least duplicate and all experiments repeated at least three times. The data were pooled and expressed as mean ± standard error of the mean (SEM). The results were evaluated using a one-way analysis of variance (ANOVA) with a Bonferroni post hoc correction. Statistical significance was assigned at p < 0.05. 3. Results 3.1. Characterization of ERK phosphorylation and cell spreading following cyclic compression ERK MAP Kinase activation and cell spreading are amongst the early changes induced in chondrocytes by mechanical stimulation. To examine which of them occurred first we examined the time course of ERK phosphorylation (Fig. 1A and B) and cell size (Fig. 1C and D) at various times up to 30 min following the initiation of mechanical stimulation. Significant upregulation of ERK phosphorylation was observed by 5 min (5-fold increase), with the levels peaking by 15 min (10-fold increase) of stimulation. Albeit still significantly elevated compared to control constructs, a decline in phosphorylation was noted with 25 min of stimulation. Chondrocyte area peaked (two-fold increase) within 5 min of cyclic compression and remained spread over the 30 min examined. This suggested that changes in cell shape and activation of the MAP Kinase signalling were rapid. As maximal cell spreading occurred prior to maximal ERK activation, it suggested that cell spreading was the earlier change so only cell shape changes were quantified in the subsequent experiments.

3.2. Extracellular calcium regulates cell morphology Given that the chondrocytes responded within 5 min to cyclic compression, and since calcium signalling is known to be involved in mechanotransduction and regulation of cell morphology, the role of extracellular calcium in regulating cell spreading was investigated. As shown in Fig. 2 cells that were stimulated in the presence of EGTA-treated media failed to spread following mechanical stimulation (EGTA treated: 58.8 ± 1.81 ␮m2 ; untreated: 86.5 ± 5.44 ␮m2 ). This was not due to a delayed response, as no spreading was seen as late as 6 h after stimulation. As EGTA can chelate ions other than calcium, we confirmed that calcium was responsible for this effect by adding calcium back into the media after EGTA treatment but prior to mechanical stimulation. The amount of calcium that needed to be added to the media was determined through dose–response experiments which ensured that the cation binding capacity of EGTA was saturated. Addition of calcium into the EGTA-treated media allowed cells to respond to cyclic compression similar to cells stimulated in untreated media as determined morphologically by SEM and by measuring cell area (Fig. 2B). 3.3. Cell spreading and retraction are regulated by two distinct ion channels in chondrocytes As extracellular calcium influenced chondrocyte response to cyclic compression, the role of two calcium transitory channels, the voltage-gated calcium channel and the cationic stretch-activated channels, in regulating these cell shape changes were investigated (Fig. 3). Involvement of these channels would also provide additional support for the involvement of ion (calcium) fluxes regulating cell response to cylic compression. Mechanical stimulation in the presence of nifedipine, a selective inhibitor of the L-Type VGCC, had no effect on cell spreading immediately following mechani-

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

235

Fig. 2. Chondrocyte cell shape changes following cyclic compression are regulated by extracellular calcium. (A) Representative SEM micrographs of unstimulated chondrocytes (A1), chondrocytes immediately after stimulation (A2), unstimulated chondrocytes in EGTA-treated media at 0 h (A3) and 6 h (A5), stimulated chondrocytes in EGTA-treated media at 0 h (A4) and 6 h (A6), unstimulated chondrocytes in EGTA-treated media supplemented with calcium at 0 h (A7), and stimulated chondrocytes in EGTA-treated media supplemented with calcium at 0 h (A8). (B) Quantification of chondrocyte area in control (C) and stimulated (S) constructs at 0 and 6 h post-mechanical stimulation under different conditions. The results of three separate experiments were pooled and expressed as mean ± SEM (n = 6). *Statistically significant difference (p < 0.05) as compared to respective unstimulated samples.

cal stimulation (0 h); however while untreated cells retracted to their original cell size by 6 h, those exposed to nifedipine remained significantly more spread relative to the unstimulated controls. In contrast, constructs stimulated in the presence of gadolinium (Gd3+ ), a selective inhibitor of the stretch activated cationic channel, did not exhibit cyclic compression-induced cell spreading, suggesting that the L-Type VGCC is involved in cell retraction while the stretch activated channel is involved in regulating cell spreading. To support the role of calcium and to exclude an indirect effect of the inhibitors in the regulation of cell morphology, constructs were mechanically stimulated in the presence of calcium ionophore A23187 and either nifedipine or gadolinium.

Treatment with the A23187 ionophore partially reversed the inhibition of gadolinium, and mechanically stimulated chondrocytes spread immediately following mechanical stimulation, compared to unstimulated chondrocytes. Similarly mechanical stimulation in the presence of both nifedipine and the A23187 ionophore reversed the effects of nifedipine-inhibition as chondrocytes had partially retracted by 6 h post-mechanical stimulation as compared to vehicle treated constructs (Fig. 3E). 3.4. Extracellular calcium regulated MT1-MMP expression We had determined previously that MT1-MMP gene and protein expression following cyclic compression plays a role in modulating

236

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

Fig. 3. Chondrocyte cell spreading and contraction following mechanical stimulation are controlled by two different ion channels. Chondrocytes underwent cyclic compression in the presence or absence of nifedipine (10 ␮M), or gadolinium (10 ␮M), with or without A23187 ionophore. The control samples received only vehicle (0.01% DMSO in HAM’S-F12). Chondrocytes were visualized by SEM at 0 h (A) or 6 h (B) and area quantified by morphometry (C–E). Representative SEM images of (A1) unstimulated, (A3) stimulated, (A2) unstimulated and gadolinium treated, (A4) stimulated and gadolinium treated, (A5) unstimulated and nifedipine treated, and (A6) stimulated and nifedipine treated. (B1) unstimulated, (B3) stimulated, (B2) unstimulated and gadolinium treated, (B4) stimulated and gadolinium treated, (B5) unstimulated and nifedipine treated, and (B6) stimulated and nifedipine treated. Densitometry results (C–E) from three separate experiments were pooled and expressed as mean ± SEM. *Statistically significant difference (p < 0.05) as compared to respective unstimulated samples.

cell shape changes observed following cyclic loading [6]. Thus we examined whether blocking the calcium influx prevented changes in MT1-MMP mRNA expression following stimulation (Fig. 4). As expected, a significant upregulation in MT1-MMP gene expression, approximately four-fold, was seen in constructs stimulated in the absence of any chemical pre-treatment. In contrast, no

increase in MT1-MMP mRNA was seen in constructs that were mechanically stimulated in the presence of EGTA-treated media. Upregulation of MT1-MMP expression was reversed by addition of calcium to the EGTA-treated media, to levels similar to that seen in mechanically stimulated untreated control constructs. Unstimulated control constructs placed in EGTA-treated media showed

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

237

Fig. 4. Upregulation of MT1-MMP gene expression in response to mechanical stimulation in chondrocytes is dependent on extracellular calcium. Cells were stimulated in the presence of EGTA (10 ␮M), EGTA (10 ␮M) and CaCl2 (5 ␮M), or vehicle alone (PBS in HAM’S F-12). (A) Representative agarose gel showing PCR products; (B) semiquantification by densitometry. Results were pooled and expressed as mean ± SEM (n = 6). *Statistically significant difference (p < 0.05) as compared to unstimulated control constructs at the respective time points.

increased MT1-MMP gene expression (approximately three-fold) in the absence of mechanical stimulation suggesting a role for calcium in regulation of basal levels of MT1-MMP in chondrocytes (Fig. 4B). Blocking L-Type VGCC and SAC by treatment with nifedipine or gadolinium respectively prevented the induction of MT1-MMP mRNA expression following compression as compared to untreated controls (Fig. 5) (p < 0.05). Interestingly, similar to that observed for EGTA-treated constructs, both nifedipine and gadolinium induced upregulation of MT1-MMP in the absence of mechanical stimulation. Treatment of constructs with the A23187 ionophore reversed the inhibitory effects of both nifedipine and gadolinium resulting in upregulation of MT1-MMP gene expression following cyclic compression supporting the role of calcium in these processes (Fig. 6). 3.5. Involvement of integrin signalling in cell spreading and MT1-MMP expression The ␣5␤1 integrin has also been shown to regulate chondrocyte spreading (and MT1-MMP expression) following cyclic compression. Integrins translate the signal by activation of adapter proteins such as Src. To further evaluate the role of calcium in this process we examined whether calcium ionophore would reverse the inhibitory effects induced by blocking integrin activation through the use of the Src inhibitor, PP2. Negative controls were PP3 (a compound that does not affect Src) and vehicle only supplemented media. PP2 prevented cell spreading and upregulation of MT1-MMP following cyclic expression. Cell spreading did occur in the controls (vehicle and PP3) so the cells were able to respond to mechanical stimulation (Fig. 7). The ionophore A23187 was able to overcome the inhibitory effect of PP2 (p < 0.05) as cell size increased with cyclic compression although not to the same level observed in control constructs (vehicle only). Increasing the ionophore concentration did not have any further effect on cell shape suggesting that it was

Fig. 5. Upregulation of MT1-MMP in response to mechanical stimulation in chondrocytes is dependent upon calcium. (A and B) Chondrocytes were stimulated in the presence or absence of nifedipine (10 ␮M), gadolinium (10 ␮M), or vehicle alone (control). (A) Representative agarose gel showing PCR products. (B) Expression was semi-quantified by densitometry, corrected for 18sRNA and then expressed as a fold change relative to untreated but mechanically stimulated cells. (C) MT1-MMP gene expression in chondrocytes under these treatment conditions but in the absence of mechanical stimulation. Results from three separate experiments were pooled and expressed as mean ± SEM (n = 6). *Statistically significant difference (p < 0.05) as compared to untreated control constructs at the respective time points.

unlikely that the partial effect was due to an insufficient amount of the drug. 3.6. Regulation of ERK MAP Kinase activation To examine whether calcium also regulates ERK MAPK activation we examined the effect of calcium channel blockers on ERK phosphorylation (Fig. 8). Cyclic compression resulted in increased ERK phosphorylation. Inhibition of either calcium channel (SAC or L-Type VGCC) did not prevent ERK phosphorylation following stimulation, and actually increased the phosphorylation in Gd3+treated constructs, although variability in the responsiveness of the cells to these drugs was observed. Mechanical stimulation in the presence of PP2 did not prevent ERK phosphorylation, however when the cells were stimulated in the presence of both the Src

238

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

Fig. 6. Treatment with calcium ionophore reversed the effects of nifedipine/gadolinium inhibition on MT1-MMP gene expression following cyclic compression. (A) Representative agarose gel illustrating the effects of A23187 ionophore treatment on MT1-MMP gene expression in nifedipine or gadolinium treated constructs. (B) Gene expression was semi-quantified by densitometry. The data from three separate experiments were pooled and expressed as mean ± SEM (n = 6). *Statistically significant difference (p < 0.05) as compared to unstimulated control constructs at the respective time points.

Fig. 7. Treatment with the Src inhibitor PP2 inhibited cell spreading and upregulation of MT1-MMP following cyclic compression. (A) Treatment with PP2 abolished cell spreading following mechanical stimulation, which was observed in vehicle and PP3 (negative control) treated cells. (B) Representative gel showing MT1-MMP gene expression under different conditions. (C) Semi-quantification of gene expression showed that treatment of cells with PP2 prior to mechanical stimulation also abolished the increase in MT1-MMP mRNA which was seen in vehicle and PP3 treated constructs. *Statistically significant difference (p < 0.05) as compared to unstimulated control constructs at the respective time points.

inhibitor, PP2, and nifedipine, ERK phosphorylation was inhibited completely (p < 0.05). 4. Discussion This study demonstrated that the transient cell spreading and retraction, as well as the associated upregulation of MT1-MMP that follows a single application of cyclic compression are regulated via calcium as treatment of the media with EGTA abolished these changes and re-introduction of calcium was able to reverse this inhibitory effect. Cell spreading and cell retraction were regulated through two distinct calcium permissive ionic channels, the cationic stretch-activated channel (SAC) and the L-Type voltage gated channel (VGCC), respectively. Despite the differential regulation of cell morphology, inhibition of either channel prevents

upregulation of MT1-MMP gene expression following stimulation. Treatment with the calcium ionophore A23187 partially reversed the inhibitory effects of nifedipine and gadolinium on cell morphology, and completely reversed the effects of stimulation on MT1-MMP expression. Inhibition of the integrin associated kinase Src abolished cell spreading and MT1-MMP upregulation and treatment with the A23187 ionophore re-established this cellular response. The two pathways, calcium and integrin associated kinase Src converged on the activation of ERK MAP Kinase, as inhibition of both pathways are required to prevent ERK phosphorylation in response to mechanical stimulation. The role of calcium in regulating these early events was demonstrated through the use of EGTA, which abolished the stimulation-induced changes, and the reintroduction of calcium into the media which returned the cellular response to mechanical

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

239

Fig. 8. Calcium and integrin signalling regulate cell response through regulation of ERK phosphorylation following mechanical stimulation. (A) Treatment of cells with both the Src inhibitor PP2 and the calcium ionophore partially reversed the effects of PP2 in terms of cell area. (B) Treatment with gadolinium did not abolish ERK phosphorylation, neither did treatment with nifedipine. Co-treatment with PP2 and nifedipine completely abolished ERK phosphorylation following mechanical stimulation. (C) Representative Western blots illustrating the effects of PP2, nifedipine and gadolinium treatment on ERK phosphorylation in stimulated constructs. *Statistically significant difference (p < 0.05) as compared to unstimulated control constructs at the respective time points. Results of three separate experiments were pooled and expressed as mean ± SEM (n = 6).

stimulation. While EGTA can chelate a variety of cations (namely Ca2+ , Na2+ , K+ ), it was the re-introduction of calcium into the media that returned cell response to normal levels, suggesting that the mechanism regulating this response is attributed to calcium and not other cations. The role of calcium influx was further confirmed through the use of the A23187 ionophore alone or in conjunction with nifedipine or gadolinium. A23187 ionophore has been previously shown to elevate intracellular calcium levels in chondrocytes in our laboratory [25]. It is interesting that ionophore treatment significantly reversed the inhibitory effects of the nifedipine and gadolinium on cell morphology but the response did not return the same level as untreated cells. The reasons for this incomplete reversal are not known, but one possible explanation may be that the localization of calcium, and not merely its presence, following mechanical stimulation is an important aspect of this response. Localized calcium currents have been previously documented to influence local signalling cascades and thus activate signalling pathways in specific locations in the cell [26,27]. Cell spreading was shown to be regulated through the SAC, which is known to be activated through direct physical deformation of the cellular membrane [27,28]. Owing to previous studies implicating SAC as the initial step in mechanotransduction [29,30], it was not surprising that we found that this channel was involved in our system. Unfortunately, it is not possible to characterize the expression of the SAC in these cells, as the channel has not been fully characterized [31], but its presence is supported by cell responsiveness to gadolinium [10,13,32]. Chondrocyte retraction following mechanical stimulation appeared to be regulated by ion fluxes via

L-Type VGCC. While this channel has been shown to be involved in mechanotransduction in other studies [33–35], its relationship to chondrocyte morphology has not been explored by others. In preliminary studies we were able to determine that articular cartilage chondrocytes express VGCC mRNA as determined by RT-PCR (data not shown), similar to that observed by Zuscik et al., in growth plate chondrocytes [15]. The sequential role of the SAC and the L-Type VGCC is not unexpected as it has been shown in smooth muscle cells that one of the functional outcomes of SAC activation is cell depolarization and subsequent activation of VGCCs [27]. Chelation of extracellular calcium abolished the stimulationinduced upregulation of MT1-MMP suggesting that either calcium directly regulates MT1-MMP expression, or that MT1-MMP expression is dependent upon cell morphology changes. The former is supported by several studies showing that calcium can regulate the binding activity of AP-1 (a transcription factor that regulates MT1-MMP transcription) in human pulmonary artery endothelial cells [17,36], and that L-Type VGCC generated currents are implicated in direct regulation of AP-1 [37]. It is interesting to note that in inhibitor-treated, but mechanically unstimulated samples, an increase in MT1-MMP gene expression was observed. Since this change was observed without any significant morphological change in cell shape it suggests that calcium signalling may regulate constitutive expression of MT1-MMP and that a calcium signalling pathway is involved regulating MT1-MMP expression in chondrocytes. Further supporting this notion is the restoration of MT1-MMP gene expression by ionophore treatment in the presence of the ion channel blocker nifedipine. This is in keeping with previous work

240

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

Fig. 9. Proposed mechanism by which chondrocytes respond to mechanical stimulation.

by Lohi et al. who showed that ionophore treatment can induce MMP expression, such as MMP-9 and MMP-2. Although they did not observe a change in MT1-MMP in their system [38], it is possible that this is a result of the different cell type (fibrosarcoma) [38,39]. Due to the previously documented role of ␣5␤1 integrin in mediating cell shape and MT1-MMP expression [6], we examined the role of the kinases Src and FAK, which are known downstream modulators of integrin activation in regulating these changes. The response to cyclic compression was facilitated through Src, as its inhibition abolished both cell spreading and MT1-MMP upregulation. No changes in FAK phosphorylation were seen with mechanical stimulation (data not shown). There are several possible explanations for the preferential involvement of Src. It is possible that compression favours activation of Src as different types of mechanical forces have been shown to differentially regulate phosphorylation of FAK and Src [40]. Alternatively, it is possible that the cell–matrix interactions formed during spreading are somewhere between focal adhesions and fibrillar adhesions. Fibrillar adhesions are known to play a role in matrix reorganization [41], and it is believed that their formation (through maturation) is dependent upon Src activity [42]. The dual regulation of cell morphology and MT1-MMP is not unexpected as others have shown that both integrins and calcium channels are components of a cellular mechanosensing complex. VGCCs are known to co-localize with ␤1 integrins as shown in mouse limb-bud and articular chondrocytes [21,22], and studies using vascular smooth muscle cells showed that ␣5␤1 integrin can regulate the function of the L-Type VGCC [23]. It is now known that Src can bind to both the II-III linker and C-terminal regions of the ␣1C subunit of the L-Type VGCC and thus regulate its activity

[43]. Similarly, since the ␣5␤1 integrin can be localized with MT1MMP in intercellular contacts [16], and Src phosphorylation has been previously shown to regulate MT1-MMP activity [44], it is not surprising that inhibition of Src can influence the mechanosensitive upregulation of MT1-MMP following stimulation. Our data suggests that calcium plays a major regulatory role in mediating this response, possibly to the extent of lying upstream to ␣5␤1 and regulating its activation. This idea is supported by the observation that blocking calcium signalling completely abolished cell spreading in response to mechanical stimulation, and that ionophore treatment was able to partially reverse the effects of Src inhibition. Moreover, the ability of calcium to influence basal levels MT1-MMP expression in unstimulated samples further implicates calcium in exerting a major regulatory role in mediating chondrocyte responses. However, as the activation of ERK is not only dependant on calcium, this raises the possibility that parallel pathways exist where both signalling cascades converge to regulate ERK, which then exerts a regulatory role on AP-1 and MT1-MMP expression. To fully address this, it would be necessary to visualize calcium fluxes in response to mechanical stimulation in real-time and determine whether inhibition of Src activity influences these currents. Unfortunately, due to the technical limitations resulting from mechanically stimulating biphasic 3D constructs, it was not possible for us to do these studies. Lastly, it is worth noting that the observation of significant changes in cell morphology occurring prior to the peak in phosphorylation of ERK suggests that physical changes in cell shape precede/initiate the downstream signalling cascades such as ERK MAP Kinase. Past studies in our lab indicated that MT1-MMP upregulation in response to cyclic compression is dependent on ERK activation [5], and owing to MT1-MMP’s role in facilitating cel-

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

lular retraction, it is possible that the activation of ERK may not be involved in cell spreading but rather the subsequent events. This notion is supported by studies which observed that physical deformations in the cellular membrane comprise the first step in mechanotransduction [29,30]. Based on the results obtained, we would like to suggest a pathway through which mechanical forces result in improved formation of cartilage tissue in vitro (Fig. 9): cyclic compression causes cellular deformation which activates SAC, resulting in calcium influx. This influx results in cell spreading and activation of the L-Type VGCC and ␣5␤1 integrin with its associated Src kinase. This results in the phosphorylation of ERK and subsequent upregulation of MT1-MMP. Simultaneously, Src signalling results in increased phosphorylation of protein such as Endophilin A2 (a protein involved in MT1-MMP endocytosis), which inhibits endocytosis and prolongs the proteolytic activity of MT1-MMP [44]. The increased degradation of the ECM would then weaken any cell-matrix adhesions causing the cells to retract to their original morphology, and the subsequent matrix degradation products either directly or by release of growth factors will signal the cells to increase synthesis of novel matrix molecules (for example, release of growth factors has been shown previously to increase proteoglycan synthesis by chondrocytes [45]). In summary, calcium currents through SAC and L-Type VGCC channels appear to regulate the transient cell spreading and retraction, respectively, that occur following the application of a short period of cyclic compression. Calcium exerts a regulatory role on the integrin-mediated signalling pathways that is facilitated through the kinase Src. Further understanding of the underlying molecular mechanisms involved in chondrocyte mechanotransduction will ultimately lead to the development of improved bioengineered cartilage and may lead to a better treatment for osteoarthritis and other cartilage pathologies. Acknowledgements I.R. was supported by a fellowship from CAN-Arthritis Society and an Ontario Student Scholarship. The research was supported by a grant from the Canadian Institutes of Health Research (CIHR). We thank Mr. Harry Bojarski and Ryding-Regency Meat Packers for providing the tissues for these studies, Dr S. Prada for manufacturing the CPP substrates and Dr. J. Wang for technical assistance. References [1] E.M. Darling, K.A. Athanasiou, Biomechanical strategies for articular cartilage regeneration, Ann. Biomed. Eng. 31 (2003) 1114–1124. [2] S.D. Waldman, C.G. Spiteri, M.D. Grynpas, R.M. Pilliar, R.A. Kandel, Long-term intermittent compressive stimulation improves the composition and mechanical properties of tissue-engineered cartilage, Tissue Eng. 10 (2004) 1323–1331. [3] S.D. Waldman, D.C. Couto, M.D. Grynpas, R.M. Pilliar, R.A. Kandel, A single application of cyclic loading can accelerate matrix deposition and enhance the properties of tissue-engineered cartilage, Osteoarthritis Cartilage 14 (2006) 323–330. [4] J.N. De Croos, S.S. Dhaliwal, M.D. Grynpas, R.M. Pilliar, R.A. Kandel, Cyclic compressive mechanical stimulation induces sequential catabolic and anabolic gene changes in chondrocytes resulting in increased extracellular matrix accumulation, Matrix Biol. 25 (2006) 323–331. [5] J.N. De Croos, B. Jang, S.S. Dhaliwal, M.D. Grynpas, R.M. Pilliar, R.A. Kandel, Membrane type-1 matrix metalloproteinase is induced following cyclic compression of in vitro grown bovine chondrocytes, Osteoarthritis Cartilage 15 (2007) 1301–1310. [6] C. Spiteri, I. Raizman, R.M. Pilliar, R.A. Kandel, Matrix accumulation by articular chondrocytes during mechanical stimulation is influenced by integrinmediated cell spreading, J. Biomed. Mater. Res. A 94 (2010) 122–129. [7] M. Zhang, J.J. Wang, Y.J. Chen, Effects of mechanical pressure on intracellular calcium release channel and cytoskeletal structure in rabbit mandibular condylar chondrocytes, Life Sci. 78 (2006) 2480–2487. [8] A.I. Alford, C.E. Yellowley, C.R. Jacobs, H.J. Donahue, Increases in cytosolic calcium, but not fluid flow, affect aggrecan mRNA levels in articular chondrocytes, J. Cell. Biochem. 90 (2003) 938–944. [9] P. D’Andrea, A. Calabrese, I. Capozzi, M. Grandolfo, R. Tonon, F. Vittur, Intercellular Ca2+ waves in mechanically stimulated articular chondrocytes, Biorheology 37 (2000) 75–83.

241

[10] F. Guilak, R.A. Zell, G.R. Erickson, D.A. Grande, C.T. Rubin, K.J. McLeod, et al., Mechanically induced calcium waves in articular chondrocytes are inhibited by gadolinium and amiloride, J. Orthop. Res. 17 (1999) 421–429. [11] T. Kono, T. Nishikori, H. Kataoka, Y. Uchio, M. Ochi, K. Enomoto, Spontaneous oscillation and mechanically induced calcium waves in chondrocytes, Cell Biochem. Funct. 24 (2006) 103–111. [12] B. Pingguan-Murphy, D.A. Lee, D.L. Bader, M.M. Knight, Activation of chondrocytes calcium signalling by dynamic compression is independent of number of cycles, Arch. Biochem. Biophys. 444 (2005) 45–51. [13] G.L. Perkins, A. Derfoul, A. Ast, D.J. Hall, An inhibitor of the stretch-activated cation receptor exerts a potent effect on chondrocyte phenotype, Differentiation 73 (2005) 199–211. [14] Y. Shao, M. Alicknavitch, M.C. Farach-Carson, Expression of voltage sensitive calcium channel (VSCC) L-type Cav1.2 (alpha1C) and T-type Cav3.2 (alpha1H) subunits during mouse bone development, Dev. Dyn. 234 (2005) 54–62. [15] M.J. Zuscik, T.E. Gunter, J.E. Puzas, R.N. Rosier, Characterization of voltagesensitive calcium channels in growth plate chondrocytes, Biochem. Biophys. Res. Commun. 234 (1997) 432–438. [16] B.G. Galvez, S. Matias-Roman, M. Yanez-Mo, F. Sanchez-Madrid, A.G. Arroyo, ECM regulates MT1-MMP localization with beta1 or alphavbeta3 integrins at distinct cell compartments modulating its internalization and activity on human endothelial cells, J. Cell Biol. 159 (2002) 509–521. [17] I. Fantozzi, S. Zhang, O. Platoshyn, C.V. Remillard, R.T. Cowling, J.X. Yuan, Hypoxia increases AP-1 binding activity by enhancing capacitative Ca2+ entry in human pulmonary artery endothelial cells, Am. J. Physiol. Lung Cell Mol. Physiol. 285 (2003) L1233–L1245. [18] G.R. Erickson, D.L. Northrup, F. Guilak, Hypo-osmotic stress induces calciumdependent actin reorganization in articular chondrocytes, Osteoarthritis Cartilage 11 (2003) 187–197. [19] D.K. Bonen, T.M. Schmid, Elevated extracellular calcium concentrations induce type X collagen synthesis in chondrocyte cultures, J. Cell Biol. 115 (1991) 1171–1178. [20] K. Deshmukh, B.D. Sawyer, Synthesis of collagen by chondrocytes in suspension culture: modulation by calcium, 3 :5 -cyclic AMP, and prostaglandins, Proc. Natl. Acad. Sci. U.S.A. 74 (1977) 3864–3868. [21] A. Mobasheri, S.D. Carter, P. Martin-Vasallo, M. Shakibaei, Integrins and stretch activated ion channels; putative components of functional cell surface mechanoreceptors in articular chondrocytes, Cell Biol. Int. 26 (2002) 1– 18. [22] M. Shakibaei, A. Mobasheri, Beta1-integrins co-localize with Na, K-ATPase, epithelial sodium channels (ENaC) and voltage activated calcium channels (VACC) in mechanoreceptor complexes of mouse limb-bud chondrocytes, Histol. Histopathol. 18 (2003) 343–351. [23] X. Wu, G.E. Davis, G.A. Meininger, E. Wilson, M.J. Davis, Regulation of the Ltype calcium channel by alpha 5beta 1 integrin requires signaling between focal adhesion proteins, J. Biol. Chem. 276 (2001) 30285–30292. [24] J. Boyle, B. Luan, T.F. Cruz, R.A. Kandel, Characterization of proteoglycan accumulation during formation of cartilagenous tissue in vitro, Osteoarthritis Cartilage 3 (1995) 117–125. [25] D. Howarth, K.P. Pritzker, T.F. Cruz, R.A. Kandel, Calcium ionophore A23187 stimulates production of a 144 kDa gelatinase, J. Rheumatol. 20 (1993) 97– 101. [26] P.A. Janmey, The cytoskeleton and cell signaling: component localization and mechanical coupling, Physiol. Rev. 78 (1998) 763–781. [27] H. Zou, L.M. Lifshitz, R.A. Tuft, K.E. Fogarty, J.J. Singer, Visualization of Ca2+ entry through single stretch-activated cation channels, Proc. Natl. Acad. Sci. U.S.A. 99 (2002) 6404–6409. [28] X. Wu, M.J. Davis, Characterization of stretch-activated cation current in coronary smooth muscle cells, Am. J. Physiol. Heart Circ. Physiol. 280 (2001) H1751–H1761. [29] F. Guilak, Compression-induced changes in the shape and volume of the chondrocyte nucleus, J. Biomech. 28 (1995) 1529–1541. [30] M.M. Knight, Z. Bomzon, E. Kimmel, A.M. Sharma, D.A. Lee, D.L. Bader, Chondrocyte deformation induces mitochondrial distortion and heterogeneous intracellular strain fields, Biomech. Model. Mechanobiol. 5 (2006) 180–191. [31] R. Maroto, A. Raso, T.G. Wood, A. Kurosky, B. Martinac, O.P. Hamill, TRPC1 forms the stretch-activated cation channel in vertebrate cells, Nat. Cell Biol. 7 (2005) 179–185. [32] J.K. Mouw, S.M. Imler, M.E. Levenston, Ion-channel regulation of chondrocyte matrix synthesis in 3D culture under static and dynamic compression, Biomech. Model. Mechanobiol. 6 (2007) 33–41. [33] E. Gomez, C. Pritchard, T.P. Herbert, cAMP-dependent protein kinase and Ca2+ influx through L-type voltage-gated calcium channels mediate Rafindependent activation of extracellular regulated kinase in response to glucagon-like peptide-1 in pancreatic beta-cells, J. Biol. Chem. 277 (2002) 48146–48151. [34] C. Guibert, T. Ducret, J.P. Savineau, Voltage-independent calcium influx in smooth muscle, Prog. Biophys. Mol. Biol. 98 (2008) 10–23. [35] L. Lacinova, Voltage-dependent calcium channels, Gen. Physiol. Biophys. 24 (Suppl. 1) (2005) 1–78. [36] C.A. Seguin, M. Bojarski, R.M. Pilliar, P.J. Roughley, R.A. Kandel, Differential regulation of matrix degrading enzymes in a TNFalpha-induced model of nucleus pulposus tissue degeneration, Matrix Biol. 25 (2006) 409–418. [37] D.R. Premkumar, R.R. Mishra, J.L. Overholt, M.S. Simonson, N.S. Cherniack, N.R. Prabhakar, L-type Ca(2+) channel activation regulates induction of c-fos transcription by hypoxia, J. Appl. Physiol. 88 (2000) 1898–1906.

242

I. Raizman et al. / Cell Calcium 48 (2010) 232–242

[38] J. Lohi, J. Keski-Oja, Calcium ionophores decrease pericellular gelatinolytic activity via inhibition of 92-kDa gelatinase expression and decrease of 72-kDa gelatinase activation, J. Biol. Chem. 270 (1995) 17602–17609. [39] J.F. Whitfield, Calcium signals and cancer, Crit. Rev. Oncog. 3 (1992) 55–90. [40] S. Lehoux, B. Esposito, R. Merval, A. Tedgui, Differential regulation of vascular focal adhesion kinase by steady stretch and pulsatility, Circulation 111 (2005) 643–649. [41] E. Zamir, B. Geiger, Molecular complexity and dynamics of cell–matrix adhesions, J. Cell Sci. 114 (2001) 3583–3590. [42] E.H. Danen, P. Sonneveld, C. Brakebusch, R. Fassler, A. Sonnenberg, The fibronectin-binding integrins alpha5beta1 and alphavbeta3 differentially mod-

ulate RhoA-GTP loading, organization of cell matrix adhesions, and fibronectin fibrillogenesis, J. Cell Biol. 159 (2002) 1071–1086. [43] E. Dubuis, N. Rockliffe, M. Hussain, M. Boyett, D. Wray, D. Gawler, Evidence for multiple Src binding sites on the alpha1c L-type Ca2+ channel and their roles in activity regulation, Cardiovasc. Res. 69 (2006) 391–401. [44] X. Wu, B. Gan, Y. Yoo, J.L. Guan, FAK-mediated src phosphorylation of endophilin A2 inhibits endocytosis of MT1-MMP and promotes ECM degradation, Dev. Cell. 9 (2005) 185–196. [45] C. Gaissmaier, J.L. Koh, K. Weise, Growth and differentiation factors for cartilage healing and repair, Injury 39 (Suppl. 1) (2008) S88–S96.