Confocal microscopy indentation system for studying in situ chondrocyte mechanics

Medical Engineering & Physics 31 (2009) 1038–1042

Contents lists available at ScienceDirect

Medical Engineering & Physics journal homepage: www.elsevier.com/locate/medengphy

Communication

Confocal microscopy indentation system for studying in situ chondrocyte mechanics Sang-Kuy Han a,b,∗ , Pina Colarusso c , Walter Herzog b,a a

Department of Mechanical Engineering, Schulich School of Engineering, University of Calgary, Alberta, Canada Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, Alberta, Canada c Department of Physiology and Biophysics, Faculty of Medicine, University of Calgary, Alberta, Canada b

a r t i c l e

i n f o

Article history: Received 4 September 2008 Received in revised form 20 May 2009 Accepted 31 May 2009 Keywords: Articular cartilage Chondrocytes In situ cell mechanics Confocal microscope

a b s t r a c t Chondrocytes synthesize extracellular matrix molecules, thus they are essential for the development, adaptation and maintenance of articular cartilage. Furthermore, it is well accepted that the biosynthetic activity of chondrocytes is influenced by the mechanical environment. Therefore, their response to mechanical stimuli has been studied extensively. Much of the knowledge in this area of research has been derived from testing of isolated cells, cartilage explants, and fixed cartilage specimens: systems that differ in important aspects from chondrocytes embedded in articular cartilage and observed during loading conditions. In this study, current model systems have been improved by working with the intact cartilage in real time. An indentation system was designed on a confocal microscope that allows for simultaneous loading and observation of chondrocytes in their native environment. Cell mechanics were then measured under precisely controlled loading conditions. The indentation system is based on a light transmissible cylindrical glass indentor of 0.17 mm thickness and 1.64 mm diameter that is aligned along the focal axis of the microscope and allows for real time observation of live cells in their native environment. The system can be used to study cell deformation and biological responses, such as calcium sparks, while applying prescribed loads on the cartilage surface. It can also provide novel information on the relationship between cell loading and cartilage adaptive/degenerative processes in the intact tissue. © 2009 IPEM. Published by Elsevier Ltd. All rights reserved.

1. Introduction Articular cartilage is a specialized connective tissue that is composed of a solid matrix, a fluid phase, chondrocytes (cells), and small electrolytes (Na+ , Cl− , etc.) which are dissolved in the interstitial fluid. Healthy articular cartilage functions as a remarkable low-friction, wear-resistant, and load-bearing material that often lasts for a lifetime. However, sometimes joints become injured or diseased and the most common disease is osteoarthritis (OA), a condition that is associated with the breakdown of articular cartilage. It is well accepted that the integrity of articular cartilage depends, to a large extent, on the proper functioning of chondrocytes which synthesize and maintain the extracellular matrix and help the tissue adapt to changing mechanical demands. Chondrocyte activity is regulated by genetic factors, the composition of the extracellular matrix, and mechanical stimuli. Although the detailed regulatory mechanisms remain unknown, there is evi-

dence suggesting that the mechanical environment influences the biosynthetic activity of chondrocytes [1–5]. Most of our knowledge on the relationship between chondrocyte biosynthesis and mechanics is based on tests using isolated cells in agarose gel [6–7], cartilage explants [3,8–11], and fixed cartilage tissue [12]. However, the structural and mechanical environment of an articular cartilage cell, when isolated or extracted in an explant tissue, is different from that of a chondrocyte in its native environment [13], while cells studied in situ using fixed tissues represent, at best, steady-state rather than instantaneous properties [12]. Dynamic cell deformations, typically considered much more important than static steady-sate properties, are lost in fixed tissue approaches. Therefore, the purpose of this study was to design an indentation system on a confocal microscope that allows for simultaneous loading and observation of chondrocytes in their native environment, and to measure cell mechanics under controlled loading conditions. 2. Methods

∗ Corresponding author at: Human Performance Laboratory, Faculty of Kinesiology, University of Calgary, 2500 University Drive N.W., Alberta, Canada T2N 1N4. Tel.: +1 403 220 3841; fax: +1 403 220 2070. E-mail address: [email protected] (S.-K. Han).

2.1. Indentation system design We devised a system that allows for real-time observation of individual cells in fully intact articular cartilage under mechanical

1350-4533/$ – see front matter © 2009 IPEM. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.medengphy.2009.05.013

S.-K. Han et al. / Medical Engineering & Physics 31 (2009) 1038–1042

1039

actuator and collects data from the load cell and the displacement gauge. 2.2. Calibration test Calibration tests were performed to assess the accuracy of the confocal microscope indentation system. Two sets of calibration experiments were performed. The first involved microspheres of known size that were reconstructed using our new system; the second involved the imaging of actual chondrocytes in situ as performed in the experiments and comparing them to chondrocytes obtained histologically.

Fig. 1. Components of the confocal microscopy indentation system. For simultaneous mechanical stimulation and confocal observation, the indentation system is fixed to the motorized stage of an upright microscope. The indentation system is controlled by a piezo actuator and data are collected from a load cell and DVRT.

loading conditions. In order to achieve this, we designed a light transmissible indentation system that is mounted on the stage of a confocal microscope directly in front of the objective. With this set-up, cell deformations can be imaged directly underneath the loaded articular surface (Fig. 1). (a) Light transmissible indentor: The tip of the indentor is made of glass (Standard Flint R-6 Glass, transmission rate – 92%, refractive index – 1.52, thickness 0.17 mm diameter 1.64 mm; manufactured by Mindrum Precision Inc. CA, USA; Fig. 2a and b) and was mounted along the optical axis of the microscope’s objective (Figs. 1 and 2d). The indentor has a tapered shape that provides structural rigidity and allows for placing a small amount of water on the top surface for imaging with a water immersion objective (Fig. 2a and b). (b) Enclosure and mounting of specimen on the x–y stage: In order to avoid moving the objective, mechanical loading is applied to the bottom of the tissue sample pushing it up towards the underside of the fixed glass indentor. The enclosure, including the glass indentor, is fixed on the motorized stage of an upright confocal microscope (Zeiss LSM 510 META, Carl Zeiss, Inc., Germany; Fig. 2d) using a custom x–y stage. A series of vertically stacked images can be recorded without moving the indentor. (c) Load control and measurement components: The position of the cartilage specimen relative to the indentor is controlled using a piezo actuator (FPA-2000, DSM, TN, USA; Fig. 2a). Displacement of the specimen relative to the indentor, and compression of the cartilage, are measured using a linear displacement gauge, DVRT (HMG-DVRT-1.5, MicroStrain, VT, USA; Fig. 2a – accuracy of 300 nm) mounted on the indentation system. A load cell unit (ELW-D1, Entran, VA, USA; Fig. 2a) attached to the bottom of the specimen holder measures the contact force on the cartilage surface. In order to avoid lateral movement of the test specimen during loading, the specimen holder and the cylindrical enclosure were given a sliding fit (Fig. 2b). A custom-written program (LabView software, National instruments, TX, USA) controls the

2.2.1. Microsphere calibration Non fluorescent microspheres of 5.93 ␮m diameter (Polysciences Inc., PA, USA) were embedded in 2% agarose gel in a specimen holder. The agarose gel was stained with fluorescein conjugated dextran of 3 kDa molecular weight (excitation: 488 nm, emission: 500 nm. Molecular Probes, OR, USA) at a concentration of 0.8 mg/ml (0.26 mM). A 40×/0.8 N.A. and 0.17 mm coverglass corrected water immersion objective (Zeiss Inc., Germany) was used to capture microsphere images through the glass indentor. Optical sections were recorded at a spacing of 0.5 ␮m in the z direction. Twelve microspheres were selected for analysis. Image thresholds were calculated by optimal thresholding methods using intensity histograms from each single microsphere image [16]. Then, the average apparent height of the microspheres in the z direction was calculated. The correction factor for z distortion was obtained by dividing the known microsphere height by the average apparent height of microspheres [17,18]. 2.2.2. Comparison of chondrocyte morphology with histology analysis Four patellae from 16-month-old New Zealand white rabbits were harvested for quantifying chondrocyte morphology using the confocal microscopy indentation system. In order to preserve the intact cartilage, only the muscle and tendon surrounding the patella were dissected away. Fluorescein conjugated dextran of 3 kDa molecular weight (excitation: 488 nm, emission: 500 nm. Molecular Probes, OR, USA) was suspended in DMEM (Dulbecco’s Modified Eagle’s Medium, Gibco, OR, USA) at a concentration of 0.8 mg/ml (0.26 mM). The patella was incubated in the dextran solution for 4–8 h at 4 ◦ C prior to fluorescence confocal imaging. After staining, tissue samples were washed in dye-free phosphate buffered saline (PBS) for 20 min. The patella was attached to the specimen holder using dental cement (Fig. 3c). In order to provide a solid binding and rigidity between the patella and the dental cement, a self-tapping bone screw (diameter 1.17 mm and length 4.76 mm, Fine Science Tools, North Vancouver, BC, Canada; Fig. 3b) was inserted into the patella. The patella was immersed in a PBS solution throughout testing to prevent dehydration of the cartilage surface. A 40×/0.8 N.A and 0.17 mm coverglass corrected water immersion objective (Zeiss Inc., Germany) was used to capture cell images through the glass indentor, giving an image pixel size of 0.41 ␮m × 0.41 ␮m at a frame rate of 2.37 Hz (the corresponding scan speed was 1.61 ␮s per pixel). Each pixel contained an eightbit value representing the intensity of the image. Optical sections were recorded at a spacing of 0.5 ␮m in the z-direction. Twelve cells were selected from the superficial zone of the rabbit retropatellar cartilage located within 40 ␮m depth of the articular surface in the z direction. Three dimensional (3D) reconstruction of the chondrocytes was performed using a custom-written code (VTK, the Visualized toolkit: Kitware Inc.). Image thresholds were calculated by an optimal thresholding method using intensity histograms from each single cell image [16]. The middle slice, representing the middle of a cell in the axial direction, was selected for calculating the

1040

S.-K. Han et al. / Medical Engineering & Physics 31 (2009) 1038–1042

Fig. 2. (a) Expanded view of the indentation system, (b) cross-sectional view of the indentation system, circle: detail of the end of the indentor with the glass window, (c) photograph of the cylindrical enclosure with the entire indentation system, and (d) photograph of the cylindrical enclosure mounted on the x–y stage of the microscope.

threshold value for each cell, because the middle slice had the most even intensity distribution between object and background. A best fit ellipsoid method was used for quantitative analysis of morphological changes of individual cells [19]. Cell widths and depths were defined along the major and minor axes of the cross section taken perpendicular to the cell height, respectively. For validation of chondrocyte size and shape obtained with the new system, superficial zone chondrocytes were assessed using histological analysis. Patellae were fixed in a 10% neutral buffered formalin solution for 7 days. Following fixation, patellae were decalcified in a mixture of 10% formic acid and formaldehyde (Cal-Ex II solution, Fisher sciencific) for 14 days. Patellae were bisected along the mid-sagittal plane and dehydrated in graded alcohols, cleared in xylene and embedded in paraffin wax (Paraplast Plus paraffin wax, Fisher Sciencific). Sections of the cartilage and subchondral bone were cut in a sagittal plane at 10 ␮m thickness, adhered to slides (Superfrost Plus Slides, Fisher Science) and stained with toluidine blue. Sections were examined under a light microscope (oil-immersion 100×/1.25 N.A objective, Zeiss Axiostar plus, Zeiss Inc., Germany) and digital images of all sections were obtained (Axiovision Imaging system, Zeiss Inc., Germany). Imaging software was used to quantify chondrocyte morphology for 130 cells per patella for comparison with that obtained using the confocal indentation system. Cell width and height were defined as the major and minor axes of the cell, respectively. Cell width and height

were measured at the focal plane where cell width and height were maximum. 2.3. Cartilage specimen test Using the system described above, we performed an experiment with the retropatellar cartilage of a 15-month-old New Zealand white rabbit (Fig. 3a). The tissue was prepared as described above. 1 MPa and 2 MPa loads were applied to the patella surface using a constant ramp speed of 6 ␮m/s and then held constant until steady state was reached (1200 s [12]). Confocal images were taken before loading, 5 min and 20 min after loading. Confocal image acquisition and cell image analysis were done, as described above. After testing, cartilage thickness was measured using the needle indentation technique [14,15]. Compressive tissue strain was calculated by dividing tissue deformation measured from the DVRT by total tissue thickness measured using the needle probe. compressive tissue strain =

tissue deformation total tissue thickness

(1)

Average compressive local extracellular matrix (ECM) strain in the superficial zone was calculated by measuring the distance between identified paired cells (n = 9) in the unloaded (du ) and

Fig. 3. Cartilage sample preparation: (a) fully intact retropatellar tissue, (b) bone screw on the test sample, and (c) patella fixed in specimen holder.

S.-K. Han et al. / Medical Engineering & Physics 31 (2009) 1038–1042 Table 1 Comparison of chondrocyte morphology measured using confocal microscopy indentation system (n = 48) and histological analysis (n = 520). Confocal microscopy indentation system 5.0 ± 0.7 11.3 ± 1.8

Cell height (␮m) Cell width (␮m)

Histological analysis with light microscopy 4.9 ± 1.1 11.7 ± 2.0

loaded state (ds ) configuration, as described previously [8]. ECM strain =

du − ds du

(2)

Axial cell strain was defined by the engineering strain of cell height. Results are shown as mean ± standard deviation. 3. Results 3.1. Calibration tests Average diameter of microspheres in the x–y plane and height of microspheres in the z-direction were measured as 6.11 ± 0.09 ␮m and 7.26 ± 0.21, respectively using the confocal microscopy system. Based on these apparent values and the actual size of microspheres, the z-direction correction factor was calculated as 0.83. After applying the z-direction correction factor, chondrocyte height and width measured using the confocal microscopy system were statistically the same as those measured using vertical slices obtained histologically (Table 1).

Fig. 4. Normalized chondrocyte deformations as a function of time: (a) normalized cell height and (b) normalized cell volume. Chondrocyte deformations were measured before loading, 5 min after indentation, and 20 min after indentation (n = 12). Note the large change in cell shape in the 5 min of loading and the reduced rate of cell deformation for the next 15 min.

1041

3.2. Cartilage specimen test Total tissue strain, local ECM strain and cell strain increased with increasing pressure on the articular cartilage surface. Local axial ECM strains in the superficial layer were larger than axial tissue and cell strains. The unloaded average cell volume was 327.3 ± 41.2 ␮m3 (n = 24), while the corresponding average chondrocyte height, depth and width were 4.2 ± 0.4 ␮m, 14.8 ± 1.8 ␮m, and 10.7 ± 1.5 ␮m, respectively. With loading, chondrocyte height decreased by 11.4 ± 4.1% and 18.6 ± 4.1%, and chondrocyte volume decreased by 10.1 ± 3.3% and 12.1 ± 4.7% for the 1 MPa and 2 MPa loading conditions at 20 min after loading, respectively. Chondrocytes rapidly deformed in the first 5 min of loading and continued to deform at a decreased rate from 5 to 20 min of loading (Fig. 4). The average cell width increased from the unloaded to the 20 min loaded state by 2.5 ± 3.0% and 5.6 ± 7.6%, for the 1 MPa and 2 MPa conditions, respectively. The corresponding values for cell depth were 0.9 ± 6.8% and 3.7 ± 3.9%. The retropatellar articular cartilage thickness was 511 ± 10 ␮m, therefore compressive tissue strains for the 1 and 2 MPa conditions were 11.0% and 26.7% at 20 min of loading, respectively. The corresponding local axial ECM strains were 22 ± 10% and 35 ± 6%. Three-dimensional reconstruction of the chondrocytes showed the typical flat shape of the superficial cells in articular cartilage (Fig. 5).

Fig. 5. 3D reconstruction of cell shape before loading and 20 min after loading. The figure shows the shape of two exemplar cells before and after 1 MPa (a) and 2 MPa load application (b) compression of the cartilage surface. x and y directions are parallel to the articular surface, and z direction is perpendicular to the articular surface. Scale bar = 5 ␮m.

1042

S.-K. Han et al. / Medical Engineering & Physics 31 (2009) 1038–1042

4. Discussion The purpose of this study was to develop a novel loading system that allows for observation of chondrocytes in real time, under controlled loading conditions, and while fully embedded in their native tissue that is fully attached to its host bone. The system is based on a light transmissible indentor placed in front of the objective of a confocal microscope. Cell deformation in response to given loads can be measured for dynamic conditions. Furthermore, images can be collected in real time and without tissue fixation, thus allowing for repeat measurements of the same chondrocytes at different times of loading cycles. The pilot data in this study merely represent one of many possibilities for studying in situ chondrocyte mechanics. Nevertheless, they suggest that cell deformations are substantially smaller in this preparation than those observed in cartilage explants exposed to comparable nominal tissue strains [8,11]. There are several possibilities why chondrocyte deformations in our study were smaller than those observed in other studies. First and foremost we believe that the structural integrity of the cartilage in our in situ setup provides protection against excessive cell strains. In previous work on explant tissues, this integrity would be compromised when excising the tissue specimen, primarily near the cutting surface where chondrocyte deformations are typically evaluated. We plan on directly testing this hypothesis by performing in situ tests prior to and following the introduction of a cut surface to the specimen and measuring cell deformations in the fully intact and compromised tissue. We speculate that cell deformations increase with proximity to the cut surface indicating that indeed, loss of structural tissue integrity is directly reflected in cell deformation. Other possibilities for the observed differences in cell deformations between the in situ and previous explant studies include the tissue properties themselves and the differences in species and joints used in the various studies (i.e., rabbit retropatellar cartilage in our case vs. dog patellofemoral groove and pig femoral condyle in studies using cartilage explants [8,11]), as well as the precise location of the cells within the tissue (i.e., highly loaded vs. lightly or unloaded areas of the joint). Only measurements of surface zone cells in the rabbit retropatellar cartilage near a cutting surface will provide an answer as to the cause for the observed differences. Further developments planned for this system include deep penetration into cartilage tissue through multi-photon microscopy, increases in the rate of cell imaging and extension to simultaneous observation of cell deformation, calcium signalling and cell membrane rupture under various loading conditions. Acknowledgements The authors gratefully acknowledge Mr. Andrew Read in the Science Workshop at the University of Calgary and Dr. Salvatore Federico for discussions on the design of the indentation system.

Dr. Lisa Chilton and Dr. Andrea Clark for discussion on tissue preparation with fluorescent dye. We also acknowledge the financial support of The Canada Research Chair Programme for Molecular and Cellular Biomechanics, The Canadian Institutes of Health Research (CIHR), The Alberta Ingenuity Fund, The CIHR Group in Inflammatory Disease, and The AHFMR Interdisciplinary Team Grant on Bone and Joint Health (#200700596). Conflict of interest The authors declare that they have no competing interests. References [1] Gray M, Pizzanelli A, Grodzinsky A, Lee R. Mechanical and physiochemcial determinants of the chondrocyte biosynthetic response. J Orthop Res 1988;6:777–92. [2] Sah RL, Grodzinsky A, Plaas A, Sandy J. Effect of static and dynamic compression on matrix metabolism in cartilage explants. In: Kuettner K, Shleyerbach R, Peyron J, Hascall V, editors. Articular cartilage and osteoarthritis. New York: Raven Press; 1992. p. 373–92. [3] Kim YJ, Sah RY, Grodzinsky AJ, Plaas AHK, Sandy JD. Mechanical regulation of cartilage biosynthetic behaviour: Physical stimuli. Arch Biochem Biophys 1994;311:1–12. [4] Urban JPG. The chondrocyte: a cell under pressure. J Rheumatol 1994;33:901–8. [5] Urban JPG. Present perspectives on cartilage and chondrocyte mechanobiology. Biorheology 2000;37:185–90. [6] Freeman PM, Natarajan RN, Kimura JH, Andriacchi TP. Chondrocyte cells respond mechanically to compressive loads. J Orthop Res 1994;12:311–20. [7] Lee DA, Knight MM, Bolton JF, Idowu BD, Kayser MV, Bader DL. Chondrocyte deformation within compressed agarose constructs at the cellular and subcellular levels. J Biomech 2000;33:81–95. [8] Guilak F, Ratcliffe A, Mow VC. Chondrocyte deformation and local tissue strain in articular cartilage: a confocal microscopy study. J Orthop Res 1995;13:410–21. [9] Buschmann MD, Hunziker EB, Kim YJ, Grodzinsky AJ. Altered aggrecan synthesis correlates with cell and nucleus structure in statically compressed cartilage. J Cell Sci 1996;109:499–508. [10] Wu JP, Kirk TB, Milne N. A study of the shape change of sheep chondrocytes with application of compression to cartilage. In: Proceedings of conference on 7th Australian and New Zealand intelligent information systems conference. 2001. p. 95–9. [11] Choi JB, Youn I, Cao L, Leddy HA, Gilchrist CL, Setton LA, et al. Zonal changes in the three-dimensional morphology of the chondron under compression: the relationship among cellular, pericellular, and extracellular deformation in articular cartilage. J Biomech 2007;40:2596–603. [12] Clark AL, Barclay LD, Matyas JR, Herzog W. In situ chondrocyte deformation with physiological compression of the feline patellofemoral joint. J Biomech 2003;36:553–68. [13] Craig S. Effects of in-vivo joint loading on articular cartilage chondrocyte viability. Master Thesis, University of Calgary, 2003. [14] Hoch DH, Grodzinsky AJ, Koob TJ, Albert ML, Eyre DR. Early changes in the material properties of rabbit articular cartilage after meniscectomy. J Orthop Res 1983;1:4–12. [15] Herzog W, Diet S, Suter E, Mayzus P, Leonard TR, Muller C, et al. Material and functional properties of articular cartilage and patellofemoral contact mechanics in an experimental model of osteoarthritis. J Biomech 1998;31:1137–45. [16] Gonzalez RC, Woods RE. Digital image processing. 2nd ed Prentice-Hall; 2002. pp. 602–607. [17] Visser TD, Oud JL. Volume measurements in three-dimensional microscopy. Scanning 1994;16:198–200. [18] Guilak F. Volume and surface area measurement of viable chondrocytes in situ using geometric modelling of serial confocal sections. J Microsc 1994;173:245–56. [19] Alyassin AM, Lancaster JL, Down III JH. Evaluation of new algorithms for the interactive measurement of surface area and volume. Med Phys 1994;21:741–75.