|Cancer||Cell death||Cell cycle||Cytoskeleton||Exo/endocytosis||Differentiation||Division||Organelles||Signalling||Stem cells||Trafficking|
Mechanical loading modulates chondrocyte primary cilia incidence and length
Susan R McGlashan*1, Martin M Knight†, Tina T Chowdhury†, Purva Joshi*, Cynthia G Jensen*, Sarah Kennedy* and Charles A Poole*‡
*Department of Anatomy with Radiology, University of Auckland, Auckland, New Zealand, †School of Engineering and Materials Science, Queen Mary, University of London, London, U.K., and ‡Section of Orthopaedic Surgery, Department of Medical and Surgical Sciences, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand
The pathways by which chondrocytes of articular cartilage sense their mechanical environment are unclear. Compelling structural evidence suggests that chondrocyte primary cilia are mechanosensory organelles. This study used a 3D agarose culture model to examine the effect of compressive strain on chondrocyte cilia. Chondrocyte/agarose constructs were subjected to cyclic compression (0–15%; 1 Hz) for 0.5–48 h. Additional constructs were compressed for 48 h and allowed to recover for 72 h in uncompressed free-swelling conditions. Incidence and length of cilia labelled with anti-acetylated α-tubulin were examined using confocal microscopy. In free-swelling chondrocytes, these parameters increased progressively, but showed a significant decrease following 24 or 48 h compression. A 72 h recovery partially reversed this effect. The reduced cilia incidence and length were not due to increased cell division. We therefore propose that control of primary cilia length is an adaptive signalling mechanism in response to varying levels and duration of mechanical loads during joint activity.
Key words: three-dimensional agarose culture, articular cartilage, cilia resorption, compressive strain, primary cilium
Abbreviations: DIC, differential interference contrast, DMEM, Dulbecco’s minimal essential medium, ECM, extracellular matrix, FCS, fetal calf serum, IFT, intraflagellar transport, PCNA, proliferating cell nuclear antigen
1To whom correspondence should be addressed (email firstname.lastname@example.org).
The chondrocytes within articular cartilage sense and respond to mechanical loading by altering the synthesis and catabolism of the extracellular matrix (Grodzinsky et al., 2000). Mechanical loading is also known to modulate chondrocyte cytoskeletal organization, although the function of this remodelling is unclear (Knight et al., 2006). During joint loading, several forms of mechanical stimuli are active, including cell deformation, hydrostatic pressure, osmotic changes, shear forces and fluid flow (Urban, 1994; Buckwalter et al., 2006). However, the mechanism(s) through which chondrocytes convert mechanical signals into biological responses in both healthy and diseased tissues is only partially understood.
Research to date suggests that a variety of mechanotransductory processes could occur via the primary cilium. Primary cilia are highly conserved, single cytoplasmic organelles with a 9+0 microtubular architecture and are found in virtually all vertebrate cells (Wheatley et al., 1996; Satir and Christensen, 2008). Chondrocytes express a single primary cilium, which projects from the cell surface into the ECM (extracellular matrix) and interacts with molecules such as collagen and glycoproteins via receptors that include integrins and NG2 (Meier-Vismara et al., 1979; Wilsman et al., 1980; Poole et al., 1985, 2001; Jensen et al., 2004; McGlashan et al., 2006). The base of the primary cilium also has a close structural relationship with the Golgi apparatus, suggesting that the cilium could be involved in the polarised secretion of ECM macromolecules during cartilage turnover (Poole et al., 1997; Song et al., 2007). In several other cell types, including renal epithelial cells and endothelium, primary cilia function as mechanosensors via the activation of intracellular Ca2+ (Praetorius and Spring, 2003: Nauli and Zhou, 2004). Studies have shown that mechanical loading also activates intracellular Ca2+ signalling in chondrocytes (Roberts et al., 2001; Pingguan-Murphy et al., 2005). However, while there is compelling structural evidence supporting a mechanosensory role for the chondrocyte primary cilium, especially in relation to ECM production, there are currently no functional data to support this hypothesis. Therefore, in a first step towards examining the roles of the cilium in cartilage homeostasis, this study tests the hypothesis that chondrocyte primary cilia incidence and length are sensitive to mechanical load.
2. Materials and methods
2.1. Preparation of chondrocyte/agarose constructs
Full-depth articular cartilage was removed from bovine metacarpophalangeal joints of 18-month-old steers. Chondrocytes were isolated from the tissue by sequential enzyme digestion with pronase (700 U/ml) and collagenase (100 U/ml) and seeded in 3% agarose (type VII) at 4×106 cells/ml as previously described (Lee et al., 2000). The chondrocyte–agarose suspension was gelled at 4°C in sterile moulds creating cylindrical constructs (5 mm diameter, 5 mm height). Constructs were cultured at 37°C/5% CO2 for up to 144 h in supplemented DMEM (Dulbecco’s minimal essential medium) with and without 20% FCS (fetal calf serum). Media were changed every 2–3 days.
2.2. Mechanical loading
To examine the effects of mechanical loading, constructs were cultured in DMEM+20% FCS for 24 h and then subjected to cyclic compression using a well-characterized loading system within a modified cell culture incubator (Zwick Testing Machines) (Lee and Bader, 1997). Cyclic compression was applied at 0–15% strain at a frequency of 1 Hz for 0.5, 6, 24 and 48 h. Unloaded control constructs were cultured under free-swelling conditions for the same time periods. Each construct was maintained in 1 ml of DMEM+20% FCS throughout. A separate group of constructs, called ‘48+72R’, were cultured for 24 h prior to compression, compressed for 48 h and then cultured in free-swelling conditions for a further 72 h. Control constructs were incubated under free-swelling conditions for 144 h.
At each time point, free-swelling and compressed constructs were fixed in 4% paraformaldehyde for 1 h at 37°C and then washed in PBS. Chondrocyte/agarose constructs underwent standard histological processing into paraffin. Sections 12 μm thick were rehydrated and underwent microwave antigen retrieval in 0.01 M citrate buffer for 2×5 min and were allowed to cool. Sections were permeabilised using 0.5% (v/v) Triton-X-100 in PBS for 5 min, followed by 5% (v/v) goat serum in PBS (Sigma–Aldrich) for 30 min at room temperature. Sections were incubated overnight at 4°C with primary antibodies raised against either acetylated α-tubulin for primary cilia staining (6-11-B; 1:500: Sigma–Aldrich) or PCNA (proliferating cell nuclear antigen; 1:500: Dako Cytomation). Sections were washed in PBS+0.1% BSA and subsequently incubated with goat anti-mouse Alexa 488 (1:400; Invitrogen) for 2 h at room temperature, then rinsed in PBS+0.1% BSA. Cell nuclei were labelled with Hoechst 33258 (100 nM; Sigma–Aldrich) for 15 min at room temperature, rinsed in PBS+0.1% BSA and mounted with Prolong Gold (Invitrogen). Negative controls comprised sections not incubated with primary antibody.
2.4. Analysis of cilia incidence, cilia length and cell proliferation
To quantify the incidence of primary cilia within each construct, 30 randomly selected fields of view, each containing 4–10 cells were examined using a Nikon epi-fluorescent microscope and a ×100 oil immersion objective. At least three constructs were examined for each condition from two separate experiments. Corresponding DIC (differential interference contrast) microscopy images were also acquired. The percentage of cells expressing a primary cilium was calculated for each experimental condition. For cilia length measurement, sections were imaged using a Leica TCS-SP2 confocal laser scanning microscope with a ×100 oil immersion objective lens and ×4 optical zoom. Due to the differences in the z resolution of the microscope compared with the x and y planes, only cilia that were approximately 90° to the incident light were selected. This ensured that the maximum z depth measured was 1.5 μm. Serial optical z sections (Δz = 0.35 μm) were used to create 2D projections of cilia for a minimum of 200 cells per condition enabling cilia length to be measured using Image J software (NIH Image). Cell proliferation was assessed by counting the number of PCNA-positive and total nuclei from a minimum of three constructs for each condition from two separate experiments. The number of PCNA-positive nuclei was expressed as a percentage of total nuclei. All differences were assessed using an unpaired Student’s t test with P-values less than 5% considered statistically significant.
3.1. Chondrocytes possess primary cilia in 3D culture
Chondrocytes in agarose showed a rounded morphology similar to chondrocytes in situ. Primary cilia were either lying against the chondrocyte cell membrane or extending into the extracellular microenvironment (Figure 1A). In free-swelling cultures, cilia incidence progressively increased, with a 2-fold increase between 24 and 144 h of culture (Figure 1B). At 144 h, there was no significant difference between cilia incidence in chondrocytes cultured in the presence or absence of serum (P>0.05; data not shown). Cilia length increased with time in culture between 24 and 48 h, reaching steady state after 48 h, with no further change in length up to 144 h (Figure 1C).
3.2. The incidence of chondrocyte primary cilia is regulated by compressive strain
Following the application of cyclic compression for 0.5 and 6 h, cilia incidence showed no statistically significant difference compared with free-swelling controls (P>0.05; Figure 2). However, following 24 h of compression, the percentage of ciliated cells was significantly reduced compared with free-swelling chondrocytes (P<0.05). Cilia incidence further decreased following 48 h of compression, with a statistically significant 74% reduction compared with free-swelling controls (P<0.001; Figure 2). In contrast, removal of compression and culture for an additional 72 h induced a 2-fold increase in cilia incidence (P<0.01); however, cilia incidence remained significantly less compared with free-swelling controls cultured for 144 h (P<0.001).
3.3. Compression-induced reduction in cilia length is reversible
No significant differences were found between cilia length of free-swelling controls and constructs subjected to dynamic compression for 0.5, 6 and 24 h (P>0.05; Figure 3). However, the application of cyclic compression for 48 h reduced cilia length by approximately 30% when compared with free-swelling controls. The mean cilia length of free-swelling constructs was 2.2 μm compared with 1.3 μm for compressed constructs (P<0.05; Figure 3). Chondrocyte cilia length in the ‘48+72R’ group showed values similar to free-swelling controls (P<0.05; Figure 3). This reversal in cilia length for the ‘48+72R’ test condition was also significantly greater when compared with constructs subjected to cyclic compression for 48 h (P<0.05; Figure 3).
3.4. Compression-induced reduction in cilia incidence is not due to an increase in cell proliferation
To assess whether a reduction in cilia incidence was due to an increase in the number of dividing cells, chondrocyte/agarose constructs were labelled with an antibody raised against PCNA. Cells in free-swelling conditions showed an increase in the percentage of proliferating cells with time in culture (Figure 4). The application of cyclic compression yielded a peak in cell proliferation at 24 h (P<0.05) followed by a significant reduction in the number of proliferating cells following 48 h and ‘48+72R’ (Figure 4; P<0.05).
This study used a well-established 3D culture model that maintains chondrocyte morphology and phenotype similar to that in situ (Aydelotte and Kuettner, 1988; Lee et al., 2000). This model system facilitates the application of cyclic cellular deformation through gross compression of the agarose construct (Knight et al., 1998; Pingguan-Murphy et al., 2005). Loading regimes identical to those employed in this study have previously been shown to modulate chondrocyte matrix synthesis and proliferation in agarose without any loss of cell viability (Lee et al., 1997).
Following isolation from the ECM and seeding into agarose gel, chondrocytes were allowed to recover in culture for 24 h. At this time point, cilia incidence was lower than values obtained from studies in situ (McGlashan et al., 2008) and could be due to a loss of cilia during collagenase treatment and centrifugation. After 144 h of free-swelling culture, cilia incidence had reached a mean percentage of 47% (Figure 1B). This incidence is similar to that found in bovine patella cartilage, which ranges between 40% and 67% depending on the cartilage zone of origin (McGlashan et al., 2008). These data suggest that primary cilia were lost during the cell isolation procedure but recovered within 144 h of culture.
Serum-free culture conditions have been previously shown to induce maximal cilia expression in several cell types (Schneider et al., 2005; Pugacheva et al., 2007; Kiprilov et al., 2008). We therefore compared the effect of serum on cilia incidence in free-swelling constructs cultured for up to 144 h, but found no significant differences, perhaps a result of the relatively low chondrocyte proliferation rates in agarose cultures. Therefore, all further experiments were conducted using media supplemented with serum.
Along with increased primary cilia frequency, cilia length also increased with time in culture under free-swelling conditions, reaching a steady state of approximately 2.2 μm after 48 h in culture (Figure 1C), and suggesting that chondrocyte cilia had reached their ‘set control length’ (Marshall et al., 2005; Wemmer and Marshall, 2007). These values corresponded to previously reported mean cilia lengths of 1.5 and 2.3 μm in canine and bovine articular cartilage, respectively (Wilsman, 1978; McGlashan et al., 2008). Since both the frequency and length of chondrocyte cilia, as measured by positive staining with antibodies against acetylated α-tubulin, are similar in agarose and in situ, we are confident that the 3D agarose system was a suitable model to examine primary cilia function in articular chondrocytes.
The application of cyclic compression affected primary cilia in a time-dependent manner. There was no significant change in the percentage of ciliated cells and cilia length following compression for both 0.5 and 6 h (Figures 2 and 3). However, there were statistically significant reductions in cilia incidence after 24 and 48 h of loading, and any remaining cilia that could be detected were significantly shorter than controls. Although loading is known to modulate cytoskeletal organization, including the microtubule network which is linked to the primary cilium, we have previously shown that these loading modalities for chondrocytes in agarose do not induce microtubule disruption or breakdown (Idowu et al., 2000). Further studies are needed to determine the role of the cilium in matrix turnover and to examine whether duration-dependent compression leads to ciliary de-sensitization as a result of ciliary shortening and/or resorption.
The precise processes governing cilia length and cilia resorption have been the focus of intense study in recent years. Cilia and flagella are assembled and maintained by a process termed IFT (intraflagellar transport), in which protein complexes called IFT particles are moved bi-directionally along the axoneme by the coordinated action of the motors dynein and kinesin (Rosenbaum and Witman, 2002). The signalling mechanisms by which these processes are controlled remain unclear; however, numerous kinases and other complexes such as Nek, HDAC and the Aurora kinases have been implicated in either or both ciliary assembly and disassembly (Santos and Reiter, 2008). Our previous studies of bovine articular cartilage in situ have shown that cilia length varies with the cartilage zone of origin (McGlashan et al., 2008) and thus may be related to the local biomechanical environment, which changes with depth into the tissue (Guilak et al., 1995). Mechanical forces have been reported to affect primary cilia assembly/disassembly in other cell types. In epithelial or endothelial cells in vitro, exposure to either laminar fluid flow (Iomini et al., 2004), or orbital shaking (Resnick and Hopfer, 2008), results in a reduction or total loss of cilia. Interestingly, these findings correlate with in vivo studies, where cilia are only found in areas of low shear stresses, and are absent from cells exposed to continuous high laminar shear stress within blood vessels (Van der Heiden et al., 2008). Similarly, studies have shown alterations in renal cilium length associated with disruption of physiological fluid shear stress (Wang et al., 2008). Iomini et al. further showed that IFT ceases following high shear stress, and the IFT 71 homologue CMG 1 can no longer be detected in endothelial cell cilia, resulting in cilia resorption (Iomini et al., 2004). Although IFT components were not examined in the current study, we believe that a similar process may have occurred as a result of compression, therefore suggesting that mechanical stimulation also influences chondrocyte ciliary assembly/disassembly rates.
Chondrocyte cilia shortening and resorption may be an adaptive mechanism to minimize cellular exposure to overt or prolonged sensory signals. A similar effect has been extensively reported in studies of Chlamydomonas flagella, which are shed when subjected to unfavourable physiological conditions such as low pH or restrictive temperatures. Removal of the stimulus results in a reversal of the shortening and a return to original steady-state flagella length (Quarmby, 2004). We believe that similar mechanisms occurred in this study during the compression and recovery phases and that mechanosensitive cilia length control is an adaptive mechanism that adjusts chondrocyte sensitivity to the local biomechanical environment. Since similar effects have been observed in several cell types, we suggest that mechanosensitive cilia length control is likely to be a highly conserved function of mechanically responsive cells.
Primary cilia are known to resorb during cell division in several cell types (Jensen et al., 1979; Tucker et al., 1979; Jensen et al., 1987; Pugacheva et al., 2007). Since previous studies have shown that cyclic compression up-regulates chondrocyte proliferation (Buschmann et al., 1995; Lee and Bader, 1997), we examined whether the compression-induced reduction in cilia incidence was due to an increase in the proportion of dividing cells. Although the percentage of proliferating cells progressively increased in free-swelling conditions (Figure 4), there was also an associated increase in cilia incidence (Figure 1B). In compressed cells, PCNA labelling revealed increased cell proliferation within 24 h followed by a reduction after 48 h of compression. The present data support a previous study in which a longer duration of dynamic compression was inhibitory for cell proliferation (Chowdhury et al., 2003). Our data suggest that decreased cilia incidence is independent of cell proliferation, since when cilia incidence is maximally reduced (following 48 h of compression), the proportion of proliferating cells is also reduced compared with free-swelling controls (Figure 4). In addition, given that cell viability remains above 95% during the experimental period, it is unlikely that the reduction in cilia incidence during the loading period is due to cell apoptosis. Therefore, these findings suggest that there is no direct correlation between cell proliferation and cilia incidence under the conditions used in this study.
In conclusion, we have previously suggested that changes in the incidence and/or length of chondrocyte primary cilia may influence the mechanosensitivity of the cell (McGlashan et al., 2008). In the present study, we have shown, using a physiologically relevant model, that the process controlling the incidence and length of chondrocyte primary cilia is mechanosensitive. A feedback mechanism may exist such that the mechanosensitivity of cartilage is down-regulated in cases of prolonged or chronic mechanical loading. The effect of this regulation is unclear. Further studies will reveal which part of the process governing cilia length is mechanosensitive and will investigate if IFT is also influenced by compressive strain.
Susan McGlashan designed the study, conducted tissue culture, immunohistochemistry, microscopy, image and data analysis and wrote the manuscript. Tina Chowdhury conducted tissue culture, mechanical loading experiments and critically reviewed the manuscript. Purva Joshi and Sarah Kennedy conducted immunohistochemistry, microscopy and image analysis. Cynthia Jensen, Martin Knight and Charles Poole contributed intellectually to the experimental design, data interpretation and preparation of the manuscript.
We thank the Biomedical Imaging Research Unit for use of its microscopy facilities.
This work was conducted, in part, during a study visit by S.R.M. to Queen Mary, University of London, funded by
Aydelotte, MB and Kuettner, KE (1988) Differences between sub-populations of cultured bovine articular chondrocytes. I. Morphology and cartilage matrix production. Connect Tissue Res 18, 205-22
Chowdhury, TT, Bader, DL, Shelton, JC and Lee, DA (2003) Temporal regulation of chondrocyte metabolism in agarose constructs subjected to dynamic compression. Arch Biochem Biophys 417, 105-11
Idowu, BD, Knight, MM, Bader, DL and Lee, DA (2000) Confocal analysis of cytoskeletal organisation within isolated chondrocyte sub-populations cultured in agarose. Histochem J 32, 165-74
Iomini, C, Tejada, K, Mo, W, Vaananen, H and Piperno, G (2004) Primary cilia of human endothelial cells disassemble under laminar shear stress. J Cell Biol 164, 811-7
Jensen, CG, Davison, EA, Bowser, SS and Rieder, CL (1987) Primary cilia cycle in ptk1 cells: Effects of colcemid and taxol on cilia formation and resorption. Cell Motil Cytoskeleton 7, 187-97
Jensen, CG, Poole, CA, McGlashan, SR, Marko, M, Issa, ZI, Vujcich, KV and Bowser, SS (2004) Ultrastructural, tomographic and confocal imaging of the chondrocyte primary cilium in situ. Cell Biol Int 28, 101-10
Kiprilov, EN, Awan, A, Desprat, R, Velho, M, Clement, CA and Byskov, AG (2008) Human embryonic stem cells in culture possess primary cilia with hedgehog signaling machinery. J Cell Biol 180, 897-904
Knight, MM, Lee, DA and Bader, DL (1998) The influence of elaborated pericellular matrix on the deformation of isolated articular chondrocytes cultured in agarose. Biochim Biophys Acta 1405, 67-77
Knight, MM, Toyoda, T, Lee, DA and Bader, DL (2006) Mechanical compression and hydrostatic pressure induce reversible changes in actin cytoskeletal organisation in chondrocytes in agarose. J Biomech 39, 1547-51
Lee, DA, Knight, MM, Bolton, JF, Idowu, BD, Kayser, MV and Bader, DL (2000) Chondrocyte deformation within compressed agarose constructs at the cellular and sub-cellular levels. J Biomech 33, 81-95
Marshall, WF, Qin, H, Rodrigo Brenni, M and Rosenbaum, JL (2005) Flagellar length control system: Testing a simple model based on intraflagellar transport and turnover. Mol Biol Cell 16, 270-8
Pingguan-Murphy, B, Lee, DA, Bader, DL and Knight, MM (2005) Activation of chondrocyte calcium signalling by dynamic compression is independent of number of cycles. Arch Biochem Biophys 444, 45-51
Poole, CA, Flint, MH and Beaumont, BW (1985) Analysis of the morphology and function of primary cilia in connective tissues: A cellular cybernetic probe? Cell Motil 5, 175-93
Poole, CA, Jensen, CG, Snyder, JA, Gray, CG, Hermanutz, VL and Wheatley, DN (1997) Confocal analysis of primary cilia structure and colocalization with the golgi apparatus in chondrocytes and aortic smooth muscle cells. Cell Biol Int 21, 483-94
Poole, CA, Zhang, ZJ and Ross, JM (2001) The differential distribution of acetylated and detyrosinated alpha-tubulin in the microtubular cytoskeleton and primary cilia of hyaline cartilage chondrocytes. J Anat 199, 393-405
Pugacheva, EN, Jablonski, SA, Hartman, TR, Henske, EP and Golemis, EA (2007) Hef1-dependent Aurora A activation induces disassembly of the primary cilium. Cell 129, 1351-63
Roberts, SR, Knight, MM, Lee, DA and Bader, DL (2001) Mechanical compression influences intracellular Ca2+ signaling in chondrocytes seeded in agarose constructs. J Appl Physiol 90, 1385-91
Schneider, L, Clement, CA, Teilmann, SC, Pazour, GJ, Hoffmann, EK, Satir, P and Christensen, ST (2005) PDGFRαα signaling is regulated through the primary cilium in fibroblasts. Curr Biol 15, 1861-6
Van der Heiden, K, Hierck, BP, Krams, R, de Crom, R, Cheng, C and Baiker, M (2008) Endothelial primary cilia in areas of disturbed flow are at the base of atherosclerosis. Atherosclerosis 196, 542-50
Wang, L, Weidenfeld, R, Verghese, E, Ricardo, SD and Deane, JA (2008) Alterations in renal cilium length during transient complete ureteral obstruction in the mouse. J Anat 213, 79-85
Received 28 July 2009/9 December 2009; accepted 25 January 2010
Published as Cell Biology International Immediate Publication 25 January 2010, doi:10.1042/CBI20090094
© The Author(s) Journal compilation © 2010 Portland Press Limited
ISSN Print: 1065-6995
ISSN Electronic: 1095-8355
Published by Portland Press Limited on behalf of the International Federation for Cell Biology (IFCB)