- Email: [email protected]
A mechanical apparatus with microprocessor controlled stress profile for cyclic compression of cultured articular cartilage explants
J. Biomechamcs Printed
in Great
Vol. 22. No. 11112. pp. 1285-1291,
0021 ?290/89 $3 001 .I0
1989.
Pergamon
Britain
TECHNICAL
Press plc
NOTE
A MECHANICAL APPARATUS WITH MICROPROCESSOR CONTROLLED STRESS PROFILE FOR CYCLIC COMPRESSION CULTURED ARTICULAR CARTILAGE EXPLANTS
OF
JYRKI J. PARKKINEN, MIKKO J. LAMMI, SEFPO KARJAL.AINEN,* JUICKA LAAKKONEN,* ERKKI HYVARINEN,” AIMO TIIHONEN,* HEIKKI J. HELMINEN and MARKKU TAMMI Departments
of Anatomy
and *Instrumentation,
University
of Kuopio,
Kuopio,
Finland
Abstract-An apparatus was designed for mechanical compression of cultured articular cartilage explants 2-5 mm) driven by a stepping motor. A load cell under with a cylindrical plain-ended loading head(diameter the culture dish was applied for feedback regulation utilizing a microprocessor-based control unit. The operating programs allowed either continuous or cyclic loading, the latter with adjustable loading/resting ratio. The improvements in the present design compared with previously described apparatuses for similar purposes include: (I) the accurately controlled compression by a load cell and a rapid feedback circuit; (2) the wide range of selectable stresses (25 kPa-12.5 MPa) with both continuous and cyclic loading modes; (3) the ability to handle cycles as short as 1 s with 15ms peak loading phase. Using a 4 s cycle and 0.5 MPa load for 1.5 h resulted in a significantly enhanced incorporation of radiosulphate in cultured bovine articular cartilage explants, suggesting a stimulation of proteoglycan synthesis. Light and scanning electron microscopic examinations revealed a slight depression and superficial alterations in cartilage structure at the impact site following high pressures. We expect that this apparatus will help in revealing how articular cartilage tissue and chondrocytes respond to external mechanical stimuli.
INTRODUCTION A number ofstudies have shown that joint loading modulates articular cartilage metabolism. The effect of loading on articular cartilage has been studied mostly on experimental animals (Tammi et al., 1987). By in uiuo models the alterations of articular cartilage can be reliably quantified but the cellular mechanisms controlling those changes are difficult to examine and remain largely obscure (Stockwell, 1987). In uiuo experiments are also expensive and time-consuming. In vitro models offer a more direct way to study the cellular influences of articular cartilage loading under controlled circumstances (van Kampen and van de Stadt, 1987). The load has usually been applied by an apparatus which maintains a standardized environment for cell and tissue culture during loading. The apparatus constructed by Solomons et al. (1965) and Rodan et al. (1975) for a direct mechanical stress of bone in uitro appears to have served as the prototypes for some later constructions applied on cartilage. Cartilage has been compressed either by mechanical stress (Norton et al.. 1977; Jones et al., 1982; Palmoski and Brandt, 1984; Copray et al., 1985; Kllmfeldt, 1985) or hydrostatic pressure (Bourret and Rodan, 1976; Veldhuijzen et al., 1979; van Kampen et al., 1985; Kimura et al., 1985; Lippiello et al.. 1985; Klein-Nulend et al., 1986). In addition, two systems based on stretching of cartilage cells have been published (Lee et al., 1982; De Witt et al.. 1984). The loads created by these systems remain relatively low considering the forces
Received in finalform 12 June 1989. Abbreviations: C, control unit; CPU, central processing unit; D, driver unit; DTE. data terminal equipmdnt; DVMY digital volt meter: LC, load cell: LH. loading head: LMU. linear moving unit; LP. loading phase; PB<, phosphatebuffered saline; RP, resting phase; S, socket; SEM, scanning electron microscopy; Sh, shaft; TS, teflon seal. 1285
estimated to act. e.g., on human knee (Finlay and Repo, 1978) or hipjoints (Hodge et al., 1986). Only few of the apparatuses generate both static and cyclic loads and have a feedback control of the forces applied. In this paper we describe a novel apparatus for in vitro loading ofcultured articular cartilage explants, creating both continuous and cyclic compression over a wide range of forces. All operations of the apparatus are handled by microprocessors. MATERIALS AND METHODS The explants were taken by punch (diameter 8 mm) from the patellar surface of bovine knee joint. By aid of a scalpel, the uncalcified cartilage was separated from the subchondral bone and calcified cartilage. After fixation with a tissue adhesive (Histoacryl@,, B. Braun Melsungen AG, Melsungen, F. R. G.) on Petri dishes the explants were cultured in Eagle’s minimum essential medium with Earle’s salts supplemented with antibiotics (100 U ml- 1 penicillin and 100 fig ml- ’ streptomycin), 70 pgrnl- 1 sodium ascorbate (Sigma, St. Louis, U.S.A.) and 3 mM L-glutamic acid (Flow Lab., Irvine, U.K.) for 2 days. An unexposed control sample was immobilized on the same dish as the tested one. The plugs were labelled in 50 PCiml-’ (1.85 MBqml-‘) 35S0, (carrier free, Amersham Int., Little Chalfont, U.K.) for 1.5 h and simultaneously subjected to 0.5 MPa compression with a 4 mm diameter loading head. The duration of the maximum stress was 50 ms in a 4 s total cycle length. After loading, the samples were rinsed in cold phosphate-buffered saline (PBS) supplemented with 1 mM sodium sulphate for 30 min. The loaded and unloaded portions of the plug were separated with punch (diameter 5 mm) and heated in boiling water for 5 min to terminate synthetic activity. Cartilage samples were digested in papain (Sigma, St. Louis, U.S.A.) at 60°C overnight and unincorporated radiolabel was removed with PD10 columns (Sephadex G-25. Pharmacia Ab. Uppsala.
1286
Technical
Sweden). The incorporated “S-activity was determined in a liquid scintillation counter (LKB Ab, Bromma, Sweden). Statistical analysis was performed by paired Student’s t-test. For scanning electron microscopy (SEM) articular cartilage plugs (diameter 10 mm) were subjected to continuous or cyclic (cycle 1 s or 10 s load/20 s rest) load for 15 min with 2-mm diameter loading head. The stresses used were 0.5,4.0 and 12.5 MPa. The cartilage surface effects of the loading regime used in the radiosulphate experiment were checked with SEM. After loading, the samples were rinsed with PBS and fixed with 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, for 4 h. The plugs were washed with the cacodylate buffer and dehydrated in ascending ethanol series. After critical point drying in carbon dioxide for 2 h, the plugs were gold-coated and examined with a JEOL JSM-35 (Jeol Ltd, Tokyo, Japan) scanning electron microscope. For light microscopy thin slices (1 mm) were cut perpendicularly to the surface, fixed with formalin, embedded in paraffin, and stained with safranin 0 as described in detail before (Kiviranta et al., 1985). HARDWARE
The apparatus consisted of the load, driver and control units, illustrated in Figs l-3. The framework of the load unit, made of steel, held the stepping motor (Mae-motor Ltd, Offanengo, Italy). The stepping motor drove the linear moving unit (Neff Ltd. Waldenbach, F. R. G.), connected to the interchangeable loading head (diameter 2-5 mm) through a shaft. The shaft of the linear moving unit pierced the roof of the plastic incubation chamber through an air-tight teflon seal. A continuous flow of gas (5% carbon dioxide in air) was delivered into the chamber. Tight, interchangeable sockets
Note
for the Petri dish (diameters 35, 45 and 60 mm) were constructed on the load cell (Tedea Ltd, Herzliya, Israel). The load unit was operated in a temperature-controlled room at 37°C. The control unit processed the impulses from the load cell and transmitted them to the driver unit (Fig. 3). The feedback circuit between the different units is shown as a block diagram in Fig. 4. The movements of the stepping motor were controlled by the driver (Elektro-Huhtaset ky, Huittinen, Finland), either in a manual or programmable mode. RESULTS
Description of the apparatus The cartilage plugs were fixed onto Petri dishes with a tissue adhesive commonly used in clinical practice for small wounds to preclude lateral sliding of the tissue during loading. The effects of the adhesive on the viability of cartilage cells was tested by autoradiography. The results showed that the adhesive did not prevent the cartilage cells, even in the deepest zone, from incorporating radiolabel (data not shown). The weight created by the stepping motor can be selected at one gram intervals between 50 and 4000 g. A loading head with 2 mm diameter thus produces a maximum pressure of 12.5 MPa on the cartilage explant while the minimum load is 25 kPa using a 5 mm loading head. One step from the motor moves the loading head 1.6 pm. The cyclic loading was divided into fast (l-2 s) and slow (2-59 s) modes according to the cycle time. Examples of these load profiles are shown in Figs 5 and 6. During loading the compression was continuously monitored by the load cell. A
Block diagram
Load
!
unrt
Fig. 4. Block diagram of the electronic circuits. The arrows show the route of impulses. The impulse coming from the load unit is amplified and digitalized in the control unit. In the central processing unit (CPU) the input is compared with the values given by the operator. In the case of deviation by more than 5%, the stepping motor driver will automatically correct the load level. DVM, digital voltmeter.
DTP I
(3) Fig. 1. The steel-made framework the unit. The shaft of the linear
of the load unit with a stepping motor encased in the upper extension of moving unit (LMU) pierces the top of the polycarbonate incubation chamber.
Fig. 2. The load cell (LC) with the socket (S) for a Petri dish. The dish is tixed by the screws at the edges of the socket. An interchangeable, plain-ended loading head (LH) pierces the supporting plate during loading. Continuous gas flow is delivered into the chamber. TS, teflon seal; Sh, shaft.
Fig. 3. The control (C)and driver(D) units with data terminal equipment (DTE). On the right hand side. the cables pierce the wall of the temperature-controlled room to reach the load unit,
1287
(8) Fig. 7. The surface of bovine articular cartilage after 15 min loading with 12.5 MPa and 1 s cycle in vitro. The arrow shows the site where the loading head edge impacted the cartilage surface. The surface is smooth without damage but a slight depression under the loaded area in seen. Magnification x 130.
Fig. 8. The SEM structure of articular cartilage surface after 15 min loading with 12.5 MPa and 1 s cycle. The arrow shows the site where the loading head edge impacted the cartilage surface; the loaded side is to the right. There are slight surface alterations, especially on the loaded side. Magnification x 465.
1288
Technical
(b)
Note
1289
:: I
I
Fig. 5. (a) The cyclic loading as seen on the oscilloscope screen. The duration of the cycle is set to I s with 15 ms peak loading phase and 2000 g weight. Abscissa: 200 ms div-‘. (b) The peak loading phase (LP) from the same cycle as in (a). Abscissa: 20 msdiv-t.
I
I
I
I
H 0
Fig. 6. Loading with a 30 s cycle as registered by a pen recorder. The peak loading phase (LP) is 10 s and the resting phase (RP) 20s. The weight is 1000 g. Note the corrections of the load during the loading phase. Paper speed: 12 cm min - ’
1290
Technical Note
Table 1. The effect of 1.5 h cyclic loading (load 0.5 MPa, cycle length 4s, 50ms peak loading phase) on radioactive sulphate incorporation into glycosaminoglycans of bovine articular cartilage plugs of 8mm diameter (CPM/mg wet weight) Specimen (n=8) Central area Border area
Control (mean&-SD.) 1163+477 1816&-475
(t&t) <0.05 N.S.
Loaded (mean + SD.) 1687k429 1979&238
Central area denotes the part of the plug under the 4 mm diameter loading head, quantitatively isolated after loading by a 5 mm diameter punch, while the border area consists of the rest of the explant (width of the ring about 1.5 mm). Each culture contained two plugs, one of which was a control, glued on the same dish as the loaded plug. Two-tailed paired Student’s t-test was used for statistical evaluation. N.S., not significant.
corrective movement ensued when the measured value deviated from the preset value by 5% or more. In the continuous and the slower cyclic loading modes, correction was carried out at 0.5 s intervals, while with the faster cycle correction was performed in the next cycle. Experiments on cartilage plugs
The possibility for immediate cartilage surface degradation by the apparatus was evaluated by 15 min experiments using the small loading head (diameter 2 mm) to obtain the highest surface pressures. Light microscopically, the surface of articular cartilage plugs was smooth but a tiny dip was found in the margin area of some plugs if the load reached 4 MPa or more. The surface showed a slight depression under the loaded area, though no vertical clefts were identified even after 12.5 MPa load (Fig. 7). In the SEM examination alterations of articular cartilage surface were observed at the loading head margin area (Fig. 8). With the pressures and loading periods used for the metabolic studies of the cartilage (loading head diameter 4 mm, 0.5 MPa, 1.5 h loading, 4 s cycle) the exposition of collagen fibers in the loading head margin area of the explants was barely visible. Results from an experiment where the apparatus was used to study the effect of cyclic loading on proteoglycan synthesis is shown in Table 1. The incorporation of sulphate in the plugs, indicating the rate of synthesis of proteoglycans, showed a significant increase in the compressed cartilage under the loading head. DISCUSSION
The present apparatus was developed for relatively large articular cartilage plugs (diameter 5-10 mm) with a straight surface profile, such as those obtained from large domestic animals. Our system is not very suitable for loading of articular cartilage from small animal species owing to difficulties in obtaining large enough specimens. Bovine articular cartilage from 1-2-year-old animals was mostly used since samples from older animals showed occasional fibrillation, while ioints from younger animals yielded less cartilage surface for the experimems. For in vitro loading, Jones et al. (19821 ~~~ Palmoski and Brandt (1984). Kllmfeldt (1985k Lippiello et al. (1985) and Kim&a e;al. (1985) also used fragments or plugs of articular cartilage. In addition, several svstems have been designed for mechanical testing of epiphyses or epiphyseal &lls from chick embryo cartilage (Rodan et al.. 1975: Bourret and Rodan, 1976; Norton et al., 1977; Veldhuijzen’et al., 1979; De Witt et .a/., 1984; van Kampen et al., 1985). ~~,I
One of the novel features of the present in vitro loading apparatus is its capability to create both static and cyclic loads, the latter with a wide range of loading cycles. In this way it is possible to investigate the effects of both loading modes with the same apparatus. This is an important issue, since cyclic loading was suggested to have an anabolic influence on chondrocyte metabolism while continuous loading may have the opposite effect (Palmoski and Brandt, 1984). Only a few of the previously described apparatuses create different cyclic loading profiles whereas several papers deal with static loading only (Jones et al., 1982; Kllmfeldt, 1985; Lippiello et al., 1985). In our initial experiments the stimulation of proteoglycan synthesis by cyclic loading was confirmed (Table 1). Further investigations with different cycle lengths are clearly needed to gain a better insight into the regulation mechanisms of pressure on cartilage. The relative length of the off-phase in cyclic compression, i.e. the resting time between the loading sequences, which allows the rehydration of cartilage and a decrease in the local concentration of extracellular macromolecules, should be particularly important. In earlier studies the magnitude of applied forces has greatly varied. Palmoski and Brandt (1984) used 1.1 kPa for continuous compression of cartilage while the greatest static load hitherto reported has been 2.9 MPa (Jones et al., 1982; Kliimfeldt, 1985) and the largest hydrostatic loads were used by Kimura et al. (1985), from 350 kPa up to 2.7 MPa. For cyclic compression 11-13 kPa pressures have been applied by several investigators (Veldhuijzen et al., 1979; Palmoski and Brandt, 1984; van Kampen et al., 1985). The magnitude of the compressive forces acting on the cartilage surfaces of working joints are not well known, but according to the results of Morrison (1970), Maquet et al. (1975) and Harrington (1976), the physiological stress during the normal walking cycle may be of the order of 0.8-6.3 MPa (Finlay and Repo, 1978). In the human hip joint, peak forces exceeding 18 MPa have been measured (Hodge et al., 1986). The high end of the suggested physiological force range thus goes beyond the working limit of all previously published in vitro loading models. The present apparatus creates both static and cyclic compression at least up to 12.5 MPa, while forces as low as 25 kPa are also possible. The continuous monitoring of the weight is essential when using a direct, mechanical compression with a fast cycle. During compression under a sustained load the volume of cartilage is reduced due to the creep of the tissue. Therefore, to keep the compression at a preset level, a force monitoring system with a feedback circuit was included to obtain reproducible loads. Instead of measuring forces created by the stepping motor our apparatus controls forces directed onto cartilage plugs. The present apparatus allows a 5% variation in the load level, but no variation in the duration of the load or the time of experiment. The SEM and light microscopic investigations showed that the surface of articular cartilage plugs showed little damage even after a 12.5 MPa stress. A 0.5 MPa load with a 4s cycle (of which 50 ms is load phase) led to a stimulation of proteoglycan synthesis in the compressed area. The proteoglycan synthesis seemed to be increased also at the border area, possibly due to the stretching effect caused by the compression. It is generally accepted that mechanical stress plays an important role as a regulator of articular cartilage metabolism. A detailed investigation of the phenomenon requires an in vitro system which allows reproducible modulation of the loading forces and use of cycles over a wide range. We believe that this type of in vitro loading apparatus will be of great help in the characterization of the biomechanical control of chondrocyte metabolism. Acknowledgements-This work was supported by the grants from The University of Kuopio, The North Savo Fund of the
Technical
Finnish Cultural Foundation, The Research and Science Foundation of Farmos, The Paulo Foundation, The Academy of Finland and The Finnish Research Council for Physical Education and Sports, Ministry of Education. The authors arc most grateful to Mrs Eija Rahunen and Mrs Eija Voutilainen for skilful technical assistance. REFERENCES
Bourret, L. A. and Rodan, G. A. (1976) The role of calcium in the inhibition of CAMP accumulation in epiphyseal cartilage cells exposed to physiological pressure. J. cell. Physiol. 88. 353-362. Copray. J. C. V. M., Jansen. H. W. B. and Duterloo, H. S. (1985) An in vitro system for studying the effect of variable compressive forces on the mandibular condvlar cartilage of the rat. Archs oral Biol. 30, 305-311. _ De Witt. M. T.. Handlev. C. J.. Oakes. B. W. and Lowther. D. A. (1684) In vitro response of chondrocytes to mechanical loading. The effect of short term mechanical tension. Connect. Tiss. Res. 12, 97-109. Finlay, J. B. and Repo, R. U. (1978) Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Trans. biomed. Engng 25, 34-39. Harrington, I. J. (1976) A bioengineering analysis of force actions at knee in normal and pathological gait. Biomed. Enqng II, 167-172. Hodge, W. A., Fijan, R. S.. Carlson, K. L., Burgess, R. G. Harris. W. H. and Mann. R. W. (1986) Contact nressures in the human hip joint measured in oiuo. Proc. natn. Acad. Sci. U.S.A. 83, 2879-2883. Jones, I. A., Kllmfeldt, A. and Sandstriim, T. (1982) The effect of continuous mechanical pressure upon the turnover of articular cartilage proteoglycans in vitro. Ckn. Orthop. 165, 283-289. van Kampen, G. P. J. and van de Stadt, R. J. (1987) Cartilage and chondrocyte responses to mechanical loading in vitro. Joint loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., SHlmlnen, A.-M., Tammi, M.. Paukkonen. K. and Jurvelin, J.), pp. 112-125. Wright, Bristol. van Kampen, G. P. J.. Veldhuijzen, J. P., Kuijer, R., van de Stadt. R. J. and Schipper, C. A. (1985) Cartilage response to mechanical force in high-density chondrocyte cultures. Arth. Rheum. 28,419424. Kimura, J. H., Schipplein, 0. D., Kuettner, K. E. and Andriacchi, T. P. (1985) Effects of hydrostatic loading on extracellular matrix formation. Trans. orthop. Res. Sot. 10, 365.
c
22:*,.,2_,
Note
1291
Kiviranta, I., Jurvelin, J., Tammi, M., SZiminen, A.-M. and Helminen, H. J. (1985) Microspectrophotometric quantitation of glycosaminoglycans in articular cartilage sections with Safr&n 0. Hi&hemistry 82, 249-255. Klgmfeldt. A. (1985) Continuous mechanical nressure and joint tissue. &ani. J. Rheum. 14, 431437. Klein-Nulend, J., Veldhuijzen, J. P. and Burger, E. H. (1986) Increased calcification of growth plate cartilage as a result of compressive force in vitro. A&. Rheum. 29, 1002-1009. Lee, R. C., Rich, J. B., Kelley, K. M., Weiman, D. S. and Mathews, M. B. (1982) A comparison of in vitro cellular responses to mechanical and electrical stimulation. Am. Surg. 48, 567-574. Lippiello, L., Kaye, C., Neumata, T. and Mankin, H. J. (1985) In vitro metabolic response of articular cartilage segments to low levels of hydrostatic pressure. Connect. Tiss. Res. 13, 99-107. Maquet, P. G.. Van DeBerg, A. J. and Simonet, J. C. (1975) Femoro-tibia1 weight-bearing areas. J. Bone Joint Surg. SlA, 766-77 1. Morrison, J. B., (1970) The mechanics of the knee joint in relation to normal walking. J. Biomechanics 3, 5161. Norton, L. A., Rodan, G. A. and Bourret, L. A. (1977) Epiphyseal cartilage CAMP changes produced by electrical and mechanical perturbations. Gin. Orthop. 12.4, 59-68. Palmoski, M. J. and Brandt, K. D. (1984) Effects of static and compressive loading on articular cartilage plugs in uitro. Arch. Rheum. 27, 615-681. Rodan, G. A., Mensi, T. and Harvey, A. (1975) A quantitative method for application of compressive forces to bone in tissue culture. C&if Tiss. Res. 18, 125-131. Solomons, C. C., Shuster, D. and Kwan A. (1965) Biochemical effects of mechanical stress. I. Control of 3ZP release from rat femur in vitro. Aerospace Med. 36, 33-34. Stockwell. R. A. (1987) Structure and function ofchondrocvte under mechanical siress. Joint Loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., Sllmiinen, A.-M., Tammi, M., Paukkonen, K. and Jurvelin, J.), pp. 126-148. Wright, Bristol. Tammi, M., Paukkonen, K., Kiviranta, I., Jurvelin, J., Sllmlnen, A.-M. and Helminen, H. J. (1987) Joint loadinginduced alterations in articular cartilage. Joint Loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., SZmBnen, A.-M.. Tammi, M., Paukkonen, K. and Jurvelin, J.), pp. 64-88. Wright, Bristol. Veldhuijzen, J. P., Bourret. L. A. and Rodan, G. A. (1979) In tritrostudies of the effect of intermittent compressive forces on cartilage cell proliferation. J. cell. Physiol. 98, 299-306.
in Great
Vol. 22. No. 11112. pp. 1285-1291,
0021 ?290/89 $3 001 .I0
1989.
Pergamon
Britain
TECHNICAL
Press plc
NOTE
A MECHANICAL APPARATUS WITH MICROPROCESSOR CONTROLLED STRESS PROFILE FOR CYCLIC COMPRESSION CULTURED ARTICULAR CARTILAGE EXPLANTS
OF
JYRKI J. PARKKINEN, MIKKO J. LAMMI, SEFPO KARJAL.AINEN,* JUICKA LAAKKONEN,* ERKKI HYVARINEN,” AIMO TIIHONEN,* HEIKKI J. HELMINEN and MARKKU TAMMI Departments
of Anatomy
and *Instrumentation,
University
of Kuopio,
Kuopio,
Finland
Abstract-An apparatus was designed for mechanical compression of cultured articular cartilage explants 2-5 mm) driven by a stepping motor. A load cell under with a cylindrical plain-ended loading head(diameter the culture dish was applied for feedback regulation utilizing a microprocessor-based control unit. The operating programs allowed either continuous or cyclic loading, the latter with adjustable loading/resting ratio. The improvements in the present design compared with previously described apparatuses for similar purposes include: (I) the accurately controlled compression by a load cell and a rapid feedback circuit; (2) the wide range of selectable stresses (25 kPa-12.5 MPa) with both continuous and cyclic loading modes; (3) the ability to handle cycles as short as 1 s with 15ms peak loading phase. Using a 4 s cycle and 0.5 MPa load for 1.5 h resulted in a significantly enhanced incorporation of radiosulphate in cultured bovine articular cartilage explants, suggesting a stimulation of proteoglycan synthesis. Light and scanning electron microscopic examinations revealed a slight depression and superficial alterations in cartilage structure at the impact site following high pressures. We expect that this apparatus will help in revealing how articular cartilage tissue and chondrocytes respond to external mechanical stimuli.
INTRODUCTION A number ofstudies have shown that joint loading modulates articular cartilage metabolism. The effect of loading on articular cartilage has been studied mostly on experimental animals (Tammi et al., 1987). By in uiuo models the alterations of articular cartilage can be reliably quantified but the cellular mechanisms controlling those changes are difficult to examine and remain largely obscure (Stockwell, 1987). In uiuo experiments are also expensive and time-consuming. In vitro models offer a more direct way to study the cellular influences of articular cartilage loading under controlled circumstances (van Kampen and van de Stadt, 1987). The load has usually been applied by an apparatus which maintains a standardized environment for cell and tissue culture during loading. The apparatus constructed by Solomons et al. (1965) and Rodan et al. (1975) for a direct mechanical stress of bone in uitro appears to have served as the prototypes for some later constructions applied on cartilage. Cartilage has been compressed either by mechanical stress (Norton et al.. 1977; Jones et al., 1982; Palmoski and Brandt, 1984; Copray et al., 1985; Kllmfeldt, 1985) or hydrostatic pressure (Bourret and Rodan, 1976; Veldhuijzen et al., 1979; van Kampen et al., 1985; Kimura et al., 1985; Lippiello et al.. 1985; Klein-Nulend et al., 1986). In addition, two systems based on stretching of cartilage cells have been published (Lee et al., 1982; De Witt et al.. 1984). The loads created by these systems remain relatively low considering the forces
Received in finalform 12 June 1989. Abbreviations: C, control unit; CPU, central processing unit; D, driver unit; DTE. data terminal equipmdnt; DVMY digital volt meter: LC, load cell: LH. loading head: LMU. linear moving unit; LP. loading phase; PB<, phosphatebuffered saline; RP, resting phase; S, socket; SEM, scanning electron microscopy; Sh, shaft; TS, teflon seal. 1285
estimated to act. e.g., on human knee (Finlay and Repo, 1978) or hipjoints (Hodge et al., 1986). Only few of the apparatuses generate both static and cyclic loads and have a feedback control of the forces applied. In this paper we describe a novel apparatus for in vitro loading ofcultured articular cartilage explants, creating both continuous and cyclic compression over a wide range of forces. All operations of the apparatus are handled by microprocessors. MATERIALS AND METHODS The explants were taken by punch (diameter 8 mm) from the patellar surface of bovine knee joint. By aid of a scalpel, the uncalcified cartilage was separated from the subchondral bone and calcified cartilage. After fixation with a tissue adhesive (Histoacryl@,, B. Braun Melsungen AG, Melsungen, F. R. G.) on Petri dishes the explants were cultured in Eagle’s minimum essential medium with Earle’s salts supplemented with antibiotics (100 U ml- 1 penicillin and 100 fig ml- ’ streptomycin), 70 pgrnl- 1 sodium ascorbate (Sigma, St. Louis, U.S.A.) and 3 mM L-glutamic acid (Flow Lab., Irvine, U.K.) for 2 days. An unexposed control sample was immobilized on the same dish as the tested one. The plugs were labelled in 50 PCiml-’ (1.85 MBqml-‘) 35S0, (carrier free, Amersham Int., Little Chalfont, U.K.) for 1.5 h and simultaneously subjected to 0.5 MPa compression with a 4 mm diameter loading head. The duration of the maximum stress was 50 ms in a 4 s total cycle length. After loading, the samples were rinsed in cold phosphate-buffered saline (PBS) supplemented with 1 mM sodium sulphate for 30 min. The loaded and unloaded portions of the plug were separated with punch (diameter 5 mm) and heated in boiling water for 5 min to terminate synthetic activity. Cartilage samples were digested in papain (Sigma, St. Louis, U.S.A.) at 60°C overnight and unincorporated radiolabel was removed with PD10 columns (Sephadex G-25. Pharmacia Ab. Uppsala.
1286
Technical
Sweden). The incorporated “S-activity was determined in a liquid scintillation counter (LKB Ab, Bromma, Sweden). Statistical analysis was performed by paired Student’s t-test. For scanning electron microscopy (SEM) articular cartilage plugs (diameter 10 mm) were subjected to continuous or cyclic (cycle 1 s or 10 s load/20 s rest) load for 15 min with 2-mm diameter loading head. The stresses used were 0.5,4.0 and 12.5 MPa. The cartilage surface effects of the loading regime used in the radiosulphate experiment were checked with SEM. After loading, the samples were rinsed with PBS and fixed with 2% glutaraldehyde in 0.1 M cacodylate buffer, pH 7.4, for 4 h. The plugs were washed with the cacodylate buffer and dehydrated in ascending ethanol series. After critical point drying in carbon dioxide for 2 h, the plugs were gold-coated and examined with a JEOL JSM-35 (Jeol Ltd, Tokyo, Japan) scanning electron microscope. For light microscopy thin slices (1 mm) were cut perpendicularly to the surface, fixed with formalin, embedded in paraffin, and stained with safranin 0 as described in detail before (Kiviranta et al., 1985). HARDWARE
The apparatus consisted of the load, driver and control units, illustrated in Figs l-3. The framework of the load unit, made of steel, held the stepping motor (Mae-motor Ltd, Offanengo, Italy). The stepping motor drove the linear moving unit (Neff Ltd. Waldenbach, F. R. G.), connected to the interchangeable loading head (diameter 2-5 mm) through a shaft. The shaft of the linear moving unit pierced the roof of the plastic incubation chamber through an air-tight teflon seal. A continuous flow of gas (5% carbon dioxide in air) was delivered into the chamber. Tight, interchangeable sockets
Note
for the Petri dish (diameters 35, 45 and 60 mm) were constructed on the load cell (Tedea Ltd, Herzliya, Israel). The load unit was operated in a temperature-controlled room at 37°C. The control unit processed the impulses from the load cell and transmitted them to the driver unit (Fig. 3). The feedback circuit between the different units is shown as a block diagram in Fig. 4. The movements of the stepping motor were controlled by the driver (Elektro-Huhtaset ky, Huittinen, Finland), either in a manual or programmable mode. RESULTS
Description of the apparatus The cartilage plugs were fixed onto Petri dishes with a tissue adhesive commonly used in clinical practice for small wounds to preclude lateral sliding of the tissue during loading. The effects of the adhesive on the viability of cartilage cells was tested by autoradiography. The results showed that the adhesive did not prevent the cartilage cells, even in the deepest zone, from incorporating radiolabel (data not shown). The weight created by the stepping motor can be selected at one gram intervals between 50 and 4000 g. A loading head with 2 mm diameter thus produces a maximum pressure of 12.5 MPa on the cartilage explant while the minimum load is 25 kPa using a 5 mm loading head. One step from the motor moves the loading head 1.6 pm. The cyclic loading was divided into fast (l-2 s) and slow (2-59 s) modes according to the cycle time. Examples of these load profiles are shown in Figs 5 and 6. During loading the compression was continuously monitored by the load cell. A
Block diagram
Load
!
unrt
Fig. 4. Block diagram of the electronic circuits. The arrows show the route of impulses. The impulse coming from the load unit is amplified and digitalized in the control unit. In the central processing unit (CPU) the input is compared with the values given by the operator. In the case of deviation by more than 5%, the stepping motor driver will automatically correct the load level. DVM, digital voltmeter.
DTP I
(3) Fig. 1. The steel-made framework the unit. The shaft of the linear
of the load unit with a stepping motor encased in the upper extension of moving unit (LMU) pierces the top of the polycarbonate incubation chamber.
Fig. 2. The load cell (LC) with the socket (S) for a Petri dish. The dish is tixed by the screws at the edges of the socket. An interchangeable, plain-ended loading head (LH) pierces the supporting plate during loading. Continuous gas flow is delivered into the chamber. TS, teflon seal; Sh, shaft.
Fig. 3. The control (C)and driver(D) units with data terminal equipment (DTE). On the right hand side. the cables pierce the wall of the temperature-controlled room to reach the load unit,
1287
(8) Fig. 7. The surface of bovine articular cartilage after 15 min loading with 12.5 MPa and 1 s cycle in vitro. The arrow shows the site where the loading head edge impacted the cartilage surface. The surface is smooth without damage but a slight depression under the loaded area in seen. Magnification x 130.
Fig. 8. The SEM structure of articular cartilage surface after 15 min loading with 12.5 MPa and 1 s cycle. The arrow shows the site where the loading head edge impacted the cartilage surface; the loaded side is to the right. There are slight surface alterations, especially on the loaded side. Magnification x 465.
1288
Technical
(b)
Note
1289
:: I
I
Fig. 5. (a) The cyclic loading as seen on the oscilloscope screen. The duration of the cycle is set to I s with 15 ms peak loading phase and 2000 g weight. Abscissa: 200 ms div-‘. (b) The peak loading phase (LP) from the same cycle as in (a). Abscissa: 20 msdiv-t.
I
I
I
I
H 0
Fig. 6. Loading with a 30 s cycle as registered by a pen recorder. The peak loading phase (LP) is 10 s and the resting phase (RP) 20s. The weight is 1000 g. Note the corrections of the load during the loading phase. Paper speed: 12 cm min - ’
1290
Technical Note
Table 1. The effect of 1.5 h cyclic loading (load 0.5 MPa, cycle length 4s, 50ms peak loading phase) on radioactive sulphate incorporation into glycosaminoglycans of bovine articular cartilage plugs of 8mm diameter (CPM/mg wet weight) Specimen (n=8) Central area Border area
Control (mean&-SD.) 1163+477 1816&-475
(t&t) <0.05 N.S.
Loaded (mean + SD.) 1687k429 1979&238
Central area denotes the part of the plug under the 4 mm diameter loading head, quantitatively isolated after loading by a 5 mm diameter punch, while the border area consists of the rest of the explant (width of the ring about 1.5 mm). Each culture contained two plugs, one of which was a control, glued on the same dish as the loaded plug. Two-tailed paired Student’s t-test was used for statistical evaluation. N.S., not significant.
corrective movement ensued when the measured value deviated from the preset value by 5% or more. In the continuous and the slower cyclic loading modes, correction was carried out at 0.5 s intervals, while with the faster cycle correction was performed in the next cycle. Experiments on cartilage plugs
The possibility for immediate cartilage surface degradation by the apparatus was evaluated by 15 min experiments using the small loading head (diameter 2 mm) to obtain the highest surface pressures. Light microscopically, the surface of articular cartilage plugs was smooth but a tiny dip was found in the margin area of some plugs if the load reached 4 MPa or more. The surface showed a slight depression under the loaded area, though no vertical clefts were identified even after 12.5 MPa load (Fig. 7). In the SEM examination alterations of articular cartilage surface were observed at the loading head margin area (Fig. 8). With the pressures and loading periods used for the metabolic studies of the cartilage (loading head diameter 4 mm, 0.5 MPa, 1.5 h loading, 4 s cycle) the exposition of collagen fibers in the loading head margin area of the explants was barely visible. Results from an experiment where the apparatus was used to study the effect of cyclic loading on proteoglycan synthesis is shown in Table 1. The incorporation of sulphate in the plugs, indicating the rate of synthesis of proteoglycans, showed a significant increase in the compressed cartilage under the loading head. DISCUSSION
The present apparatus was developed for relatively large articular cartilage plugs (diameter 5-10 mm) with a straight surface profile, such as those obtained from large domestic animals. Our system is not very suitable for loading of articular cartilage from small animal species owing to difficulties in obtaining large enough specimens. Bovine articular cartilage from 1-2-year-old animals was mostly used since samples from older animals showed occasional fibrillation, while ioints from younger animals yielded less cartilage surface for the experimems. For in vitro loading, Jones et al. (19821 ~~~ Palmoski and Brandt (1984). Kllmfeldt (1985k Lippiello et al. (1985) and Kim&a e;al. (1985) also used fragments or plugs of articular cartilage. In addition, several svstems have been designed for mechanical testing of epiphyses or epiphyseal &lls from chick embryo cartilage (Rodan et al.. 1975: Bourret and Rodan, 1976; Norton et al., 1977; Veldhuijzen’et al., 1979; De Witt et .a/., 1984; van Kampen et al., 1985). ~~,I
One of the novel features of the present in vitro loading apparatus is its capability to create both static and cyclic loads, the latter with a wide range of loading cycles. In this way it is possible to investigate the effects of both loading modes with the same apparatus. This is an important issue, since cyclic loading was suggested to have an anabolic influence on chondrocyte metabolism while continuous loading may have the opposite effect (Palmoski and Brandt, 1984). Only a few of the previously described apparatuses create different cyclic loading profiles whereas several papers deal with static loading only (Jones et al., 1982; Kllmfeldt, 1985; Lippiello et al., 1985). In our initial experiments the stimulation of proteoglycan synthesis by cyclic loading was confirmed (Table 1). Further investigations with different cycle lengths are clearly needed to gain a better insight into the regulation mechanisms of pressure on cartilage. The relative length of the off-phase in cyclic compression, i.e. the resting time between the loading sequences, which allows the rehydration of cartilage and a decrease in the local concentration of extracellular macromolecules, should be particularly important. In earlier studies the magnitude of applied forces has greatly varied. Palmoski and Brandt (1984) used 1.1 kPa for continuous compression of cartilage while the greatest static load hitherto reported has been 2.9 MPa (Jones et al., 1982; Kliimfeldt, 1985) and the largest hydrostatic loads were used by Kimura et al. (1985), from 350 kPa up to 2.7 MPa. For cyclic compression 11-13 kPa pressures have been applied by several investigators (Veldhuijzen et al., 1979; Palmoski and Brandt, 1984; van Kampen et al., 1985). The magnitude of the compressive forces acting on the cartilage surfaces of working joints are not well known, but according to the results of Morrison (1970), Maquet et al. (1975) and Harrington (1976), the physiological stress during the normal walking cycle may be of the order of 0.8-6.3 MPa (Finlay and Repo, 1978). In the human hip joint, peak forces exceeding 18 MPa have been measured (Hodge et al., 1986). The high end of the suggested physiological force range thus goes beyond the working limit of all previously published in vitro loading models. The present apparatus creates both static and cyclic compression at least up to 12.5 MPa, while forces as low as 25 kPa are also possible. The continuous monitoring of the weight is essential when using a direct, mechanical compression with a fast cycle. During compression under a sustained load the volume of cartilage is reduced due to the creep of the tissue. Therefore, to keep the compression at a preset level, a force monitoring system with a feedback circuit was included to obtain reproducible loads. Instead of measuring forces created by the stepping motor our apparatus controls forces directed onto cartilage plugs. The present apparatus allows a 5% variation in the load level, but no variation in the duration of the load or the time of experiment. The SEM and light microscopic investigations showed that the surface of articular cartilage plugs showed little damage even after a 12.5 MPa stress. A 0.5 MPa load with a 4s cycle (of which 50 ms is load phase) led to a stimulation of proteoglycan synthesis in the compressed area. The proteoglycan synthesis seemed to be increased also at the border area, possibly due to the stretching effect caused by the compression. It is generally accepted that mechanical stress plays an important role as a regulator of articular cartilage metabolism. A detailed investigation of the phenomenon requires an in vitro system which allows reproducible modulation of the loading forces and use of cycles over a wide range. We believe that this type of in vitro loading apparatus will be of great help in the characterization of the biomechanical control of chondrocyte metabolism. Acknowledgements-This work was supported by the grants from The University of Kuopio, The North Savo Fund of the
Technical
Finnish Cultural Foundation, The Research and Science Foundation of Farmos, The Paulo Foundation, The Academy of Finland and The Finnish Research Council for Physical Education and Sports, Ministry of Education. The authors arc most grateful to Mrs Eija Rahunen and Mrs Eija Voutilainen for skilful technical assistance. REFERENCES
Bourret, L. A. and Rodan, G. A. (1976) The role of calcium in the inhibition of CAMP accumulation in epiphyseal cartilage cells exposed to physiological pressure. J. cell. Physiol. 88. 353-362. Copray. J. C. V. M., Jansen. H. W. B. and Duterloo, H. S. (1985) An in vitro system for studying the effect of variable compressive forces on the mandibular condvlar cartilage of the rat. Archs oral Biol. 30, 305-311. _ De Witt. M. T.. Handlev. C. J.. Oakes. B. W. and Lowther. D. A. (1684) In vitro response of chondrocytes to mechanical loading. The effect of short term mechanical tension. Connect. Tiss. Res. 12, 97-109. Finlay, J. B. and Repo, R. U. (1978) Instrumentation and procedure for the controlled impact of articular cartilage. IEEE Trans. biomed. Engng 25, 34-39. Harrington, I. J. (1976) A bioengineering analysis of force actions at knee in normal and pathological gait. Biomed. Enqng II, 167-172. Hodge, W. A., Fijan, R. S.. Carlson, K. L., Burgess, R. G. Harris. W. H. and Mann. R. W. (1986) Contact nressures in the human hip joint measured in oiuo. Proc. natn. Acad. Sci. U.S.A. 83, 2879-2883. Jones, I. A., Kllmfeldt, A. and Sandstriim, T. (1982) The effect of continuous mechanical pressure upon the turnover of articular cartilage proteoglycans in vitro. Ckn. Orthop. 165, 283-289. van Kampen, G. P. J. and van de Stadt, R. J. (1987) Cartilage and chondrocyte responses to mechanical loading in vitro. Joint loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., SHlmlnen, A.-M., Tammi, M.. Paukkonen. K. and Jurvelin, J.), pp. 112-125. Wright, Bristol. van Kampen, G. P. J.. Veldhuijzen, J. P., Kuijer, R., van de Stadt. R. J. and Schipper, C. A. (1985) Cartilage response to mechanical force in high-density chondrocyte cultures. Arth. Rheum. 28,419424. Kimura, J. H., Schipplein, 0. D., Kuettner, K. E. and Andriacchi, T. P. (1985) Effects of hydrostatic loading on extracellular matrix formation. Trans. orthop. Res. Sot. 10, 365.
c
22:*,.,2_,
Note
1291
Kiviranta, I., Jurvelin, J., Tammi, M., SZiminen, A.-M. and Helminen, H. J. (1985) Microspectrophotometric quantitation of glycosaminoglycans in articular cartilage sections with Safr&n 0. Hi&hemistry 82, 249-255. Klgmfeldt. A. (1985) Continuous mechanical nressure and joint tissue. &ani. J. Rheum. 14, 431437. Klein-Nulend, J., Veldhuijzen, J. P. and Burger, E. H. (1986) Increased calcification of growth plate cartilage as a result of compressive force in vitro. A&. Rheum. 29, 1002-1009. Lee, R. C., Rich, J. B., Kelley, K. M., Weiman, D. S. and Mathews, M. B. (1982) A comparison of in vitro cellular responses to mechanical and electrical stimulation. Am. Surg. 48, 567-574. Lippiello, L., Kaye, C., Neumata, T. and Mankin, H. J. (1985) In vitro metabolic response of articular cartilage segments to low levels of hydrostatic pressure. Connect. Tiss. Res. 13, 99-107. Maquet, P. G.. Van DeBerg, A. J. and Simonet, J. C. (1975) Femoro-tibia1 weight-bearing areas. J. Bone Joint Surg. SlA, 766-77 1. Morrison, J. B., (1970) The mechanics of the knee joint in relation to normal walking. J. Biomechanics 3, 5161. Norton, L. A., Rodan, G. A. and Bourret, L. A. (1977) Epiphyseal cartilage CAMP changes produced by electrical and mechanical perturbations. Gin. Orthop. 12.4, 59-68. Palmoski, M. J. and Brandt, K. D. (1984) Effects of static and compressive loading on articular cartilage plugs in uitro. Arch. Rheum. 27, 615-681. Rodan, G. A., Mensi, T. and Harvey, A. (1975) A quantitative method for application of compressive forces to bone in tissue culture. C&if Tiss. Res. 18, 125-131. Solomons, C. C., Shuster, D. and Kwan A. (1965) Biochemical effects of mechanical stress. I. Control of 3ZP release from rat femur in vitro. Aerospace Med. 36, 33-34. Stockwell. R. A. (1987) Structure and function ofchondrocvte under mechanical siress. Joint Loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., Sllmiinen, A.-M., Tammi, M., Paukkonen, K. and Jurvelin, J.), pp. 126-148. Wright, Bristol. Tammi, M., Paukkonen, K., Kiviranta, I., Jurvelin, J., Sllmlnen, A.-M. and Helminen, H. J. (1987) Joint loadinginduced alterations in articular cartilage. Joint Loading. Biology and Health of Articular Structures (Edited by Helminen, H. J., Kiviranta, I., SZmBnen, A.-M.. Tammi, M., Paukkonen, K. and Jurvelin, J.), pp. 64-88. Wright, Bristol. Veldhuijzen, J. P., Bourret. L. A. and Rodan, G. A. (1979) In tritrostudies of the effect of intermittent compressive forces on cartilage cell proliferation. J. cell. Physiol. 98, 299-306.