- Email: [email protected]
Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells
ELSEVIER
Journal of Orthopaedic Research 21 (2003) 597-603
Journal of Orthopaedic Research www.elsevier.com/locate/orthres
Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells Mehran Kasra 'I
Vijay Goel ', James Martin ', Shea-Tien Wang d , Woosung Choi e, Joseph Buckwalter
Department of Mechanical, Aerospace and Biomedical Engineering, The Universitjj of Tennessee, Knoxville, TN 37969. USA Depurtment of Biomedical Ettgineering, University of Toledo, Toledo, OH 43606, USA D e y t m e n t of Orthopaedic Surgery, University of' Iowa Hospitals and Clinics. 1on.u City, I A 52242, USA Department of Truumatology & Orthopaedics, Veterans General Hospital-Taipei, Taipei, TaiMiun Department of Orthopaedic Surgery, Duejeon St. Mary's Hospitul, Catholic University of Koreu, Daejeon. South Korea
Received 17 May 2002; accepted 10 January 2003
Abstract
The pathogenesis of vibration-induced disorders of intervertebral disc at the cellular level is largely unknown. The objective of this study was to establish a method to investigate the ranges of constructive and destructive hydrostatic loading frequencies and amplitudes in preventing or inducing extracellular disc matrix degradation. Using a hydraulic chamber, normal rabbit intervertebral disc cells were tested under dynamic hydrostatic loading. Monolayer cultures of disc outer annulus cells and 3-dimensional (3-D) alginate cultures of disc nucleus pulposus cells were tested. Effects of different loading amplitudes (3-D culture, 0-3 MPa; monolayer, 0-1.7 MPa) and frequencies (1-20 Hz) on disc collagen and protein metabolism were investigated by measuring 3H-proline-labeled proteins associated with the cells in the extracellular matrix and release of 'H-proline-labeled molecules into culture medium. High frequency and high amplitude hydrostatic stress stimulated collagen synthesis in cultures of outer annulus cells whereas the lower amplitude and frequency hydrostatic stress had little effect. For the same loading duration and repetition, neither treatment significantly affected the relative amount of protein released from the cell layers, indicating that protein degradation and stability were unaffected. In the 3-D nucleus culture, higher amplitude and frequency increased synthesis rate and lowered degradation. In this case, loading amplitude had a stronger influence on cell response than that of loading frequency. Considering the ranges of loading amplitude and frequency used in this study, short-term application of high loading amplitudes and frequencies was beneficial in stimulation of protein synthesis and reduction of protein degradation. 0 2003 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. Kqwwrds: Dynamic; Hydrostatic pressure; Intervertebral disc; Cells; Rabbit
Introduction Low back pain is the primary cause of disability in active age group of the society, playing a major role in the medical, social and economic structure of industrial countries. The etiology of low back pain is still unclear and therapeutic approaches are often ineffective [5,28]. Human intervertebral discs undergo age-related degenerative changes that contribute to some of the most common causes of impairment and disability for middle aged and older persons, such as back pain [2]. Potential * Corresponding author. Tel.: + 1-865-974-7673; fax: + 1-865-9748564. E-mail address: [email protected] (M. Kasra).
causes of the age-related degeneration of intervertebral discs include declining nutrition, loss of viable cells, cell senescence, post-translational modification of matrix proteins, accumulation of degraded matrix molecules, and fatigue failure of the matrix. Degenerative changes include decreased diffusion, decreased cell viability, decreased proteoglycan synthesis, and alteration in collagen distribution [2,8]. In vivo intradiscal pressure measurement in healthy individuals and in patients with ongoing back problem showed that the intradiscal pressure in degenerated discs was significantly reduced compared with that of normal discs [31]. Changes in matrix compressive stress would inhibit disc cell metabolism throughout the disc, and could lead to progressive deterioration of the matrix. Surviving cells are
0736-0266/03/$ - see front matter 0 2003 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. doi: 10.10l6/S0736-0266(03)00027-5
598
M. Kusra et ul. I Jourtiul of' Ortl?opuc.dic Rrsrurch 21 (2003) 597403
not synthetically inactive but are, rather, producing inappropriate matrix products during aging and degeneration [6]. Disc degeneration is a multifactorial phenomenon. Both biomechanical and biologic factors have been implicated in cases of accelerated degeneration [ 1,23,27]. Mechanical factors seem to play an important role. The intervertebral disc is routinely subjected to compressive loads that alter with posture and muscle activity and can produce pressures >2 MPa in human lumbar disks in vivo [30-321. Intervertebral disc pressure varies from about 0.19 at rest up to 3 MPa, above which vertebral fracture may occur [30]. Intervertebral disc cell death is dependent on the magnitude and duration of spinal loading [21,22]. Cell death, in turn, has been associated with disc degeneration in humans [2,6]. Chronically applied compressive forces and immobilization caused changes in the biomechanical behavior and biochemical composition of rat tail intervertebral discs [15]. Therefore, maintenance of appropriate stress or hydrostatic pressure within the disc may be an important basis for strategies to alleviate disc degeneration and initiate disc repair. Different investigators have shown that hydrostatic pressure directly affects the synthesis of collagen and proteoglycan by the intervertebral disc cells [3,10,14,17]. These studies suggested that a physiologic level of hydrostatic pressure might act as an anabolic factor for stimulation of proteoglycan synthesis. This may be essential for maintaining the matrix of the disc. However abnormal pressures, higher or lower pressure than physiological levels, may have a catabolic effect such as reduction of proteoglycan synthesis rate. These studies investigated only the effect of quasi-static hydrostatic loads. During daily occupational activities intervertebral disc is exposed to dynamic oscillatory hydrostatic loads, characterized by wide frequency spectrum and variable amplitude. A physiologic level of amplitude and frequency of hydrostatic pressure may be essential for maintaining the matrix of the disc, while an abnormal amplitude and frequency of hydrostatic pressure may accelerate disc degeneration. Long-term occupational exposure to whole-body vibration increases the risk of disc degeneration and the consequent back pain [4,27, 29,331. The pathogenesis of vibration-induced disorders is still not completely clear and there is no effective treatment. Although the potential effects of vibrational stress on extracellular matrix (ECM) assembly and degradation are particularly relevant to the clinical findings of the vibration-induced disorders, the effects of vibrational loads on disc cells are largely unknown. Considering the wide range of amplitude and frequency spectrum of in vivo dynamic loading on the human spine, the objective of this study was to establish a method to investigate the duration and ranges of
constructive and destructive loading frequencies and amplitudes in preventing or inducing extracellular disc matrix (ECM) degradation. We also investigated which of loading frequency and loading amplitude has more influence in affecting disc cell response to vibration.
Materials and methods Both monolayer and 3-D culture systems were tested. Disc outer annulus cells were tested in a monolayer culture and disc nucleus cells in a 3-D alginate culture. The use of different culture formats for these two cell types is based on findings which show that in vivo conditions for annulus cells are well approximated in monolayer culture while in vivo conditions for nucleus cells are better reproduced in alginate culture [13]. Rabbit disc cells were used in this study [?0,26]. Effects of two different loading levels, low and high, were first investigated by testing the monolayer disc annulus culture. In this case the effect of different loading repetition (number of days) was also investigated. This was then followed by a separate experiment with a different protocol testing 3-D nucleus cell cultures at different loading levels of low, medium. and high frequencies and amplitudes. Cells cultured in tissue culture dishes (Costar, Corning. N Y ) were placed in a hydraulic chamber filled with culture medium. The load was imposed through the fluid medium by a piston. The piston chamber assembly was sealed tight, bleeding any air out of the system. The chamber-cell assembly was fixed in a servo-hydraulic MTS mechanical testing system (MTS Corp., Minnesota) and a haversine compressive cyclic load was applied by the machine actuator (Fig. I ) . The applied hydrostatic pressure was calculated by dividing the applied load by the piston cross-sectional area. Monoltiycw disc unnu1u.r c,rll.s
Cells were isolated by sequential digestion of adult rabbit (New Zealaiid White) annulus fibrosis with bovine testicular hyaluronidase (I600 plml. 60 min) (Sigma, St. Louis, MO) followed by collagenase
Fig. I . Piston-chamber assembly installed in an Instron servohydraulic mechanical testing system. A haversine compressive cyclic load was applied by the machine actuator on the piston. The piston transferred the load to the cells placed in the chamber filled with medium.
599
M. Kusru e l ul. I Journul u f Orlhopaedic Resrurch 21 (2003) 597-603
type IA + pronase E (0.5 mglml, 16 h) (Sigma). First passage cells were seeded in 35 mm culture dishes (600,000 cells/dish) in medium (Dulbecco's Modified Eagle MediumllO%t fetal calf serum) (Life Technologies, Rockville. MD) and incubated overnight before the first hydrostatic stress treatment. The first passage cells were divided in three groups (seven dishes in each group), group I was the control group, group I1 was exposed to a cyclic hydrostatic pressure of 0.3 MPa amplitude and 1 Hz frequency simulating a low level physiological loading condition, and group 111 was exposed to a preload pressure of 0.8 MPa and a cyclic haversine pressure of 1.7 MPa at 20 Hz frequency (maximum of 2.5 MPa) simulating a high level loading condition within an in vivo physiological level [30-321. The cell cultures of different groups were loaded in the hydraulic chamber (Fig. 1) daily for 30 min. After three days three dishes of each groups were removed for analyses and the remainder four dishes were kept loaded up to nine days. In this case, the effect of loading repetition (number of loading days) was observed.
tracted in 4 M guanidine buffer (4 M guanidine HCI, I.O'%] Triton X-100, I0 mM EDTA. 0.8 mg/ml benzamidine, 1.0 mM phenylmethylsulfonyl fluoride. 50 mM sodium acetate; pH 6.8) and the extracts were fractionated on Sephadex G-50 (Amersham Pharmacia, Piscataway, NJ) columns. The column procedure removes any free (unincorporated) 'H-proline from both the medium and cell extracts so that all CPM values reflect only the fraction of 'H-proline that was incorporated into protein [I 1,121. Column eluates were analyzed by liquid scintillation counting to determine counts per minute (CPM) and DNA was measured in the cell extract by fluoronietric assay [19]. CPM data were normalized to DNA (CPM/pg) to control for differences in cell number from culture to culture. The sum of the CPM in both medium and extract (CPM,,,,I) was calculated to determine the amount of collagens synthesized during the labeling period. The ratio ( R ) of CPM released to the medium (CPM,,,,,,,) to the total CPM, (CPM,,,d,,,,,,lCPM,,,,i), was then calculated to determine the amount of collagen degraded after loading.
Tlir.ee-~lir~irn,sio~~~~l disc nuc1c~u.scc4l.s
Stutisticul anulysrs
Rabbit disc nucleus cells were maintained in monolayer culture for up to one week after their isolation. The monolayer cultures were trypsinized and the cells counted. Alginate cultures were prepared essentially as described in the literature [9]. Trypsinized disc cells were pelleted by centrifugation and the cell pellets re-suspended at a concentration of 1.5 x loh cells per ml in a sterile solution consisting of I50 mM NaCl and I .2'%1(w:v) alginate (Kelco, San Diego, CA). Individual alginate beads containing -25,000 cells each were formed by dripping the suspension through a 20 gauge needle into a shallow dish containing 102 mM CaCl?. Freshly polymerized beads were washed in Hanks balanced salt solution and transferred to a 96 well tissue culture plate ( 1 bead per well of 5-mm diameter). 0.25 ml medium was added and the cultures were incubated at 37 "C in a humidified chamber with 5% C 0 2 . Alginate cultures were divided into 10 groups (four dishes in each group) and exposed to different cyclic hydrostatic loading conditions simulating normal and abnormal loading condition within in vivo physiological levels. The applied dynamic hydrostatic pressure had a static component (preload) plus a haversine dynamic component (amplitude) both increasing from low to high levels. Three levels of high (H), medium (M), and low (L) amplitudes (A) (preload + amplitude; HA = 2 + 3. MA = I + I .5, LA =0.3 + 0.75 MPa) and frequencies (F) ( H F = 2 0 , M F = 10, L F = 1 Hz) were chosen. Group I was exposed to the high amplitude and high frequency (HA. HF); group 2 (HA. MF); group 3 (HA, LF); group 4 (MA, HF); group 5 (MA. MF); group 6 (MA, LF); group 7 (LA, HF); group 8 (LA, MF); group 9 (LA, LF) and group 10 was the control group. The cell cultures of different groups were placed in the hydraulic chamber (Fig. I ) and loaded daily for 30 min. After three days dishes of each groups were removed l o r analyses.
For both monolayer and 3-D alginate experiments, analysis of variance (ANOVA) with a significance level of p < 0.05 was first performed to detect the existence of any significant differences among different groups. Tukey's multiple comparison test was then used to determine significant differences between any two groups. For the 3-D alginate experiment, Pearson's correlation test was performed to determine any significant correlations between the dependent variable, ratio (R). and the independent variable, loading amplitude ( A ) . When there was a significant correlation present (p < 0.05). linear regression technique was used to find a relation between the dependent and independent variables.
of col/u,yen i~rc/uhoii.snr
Loading effects on collagen synthesis and degradation were determined essentially as described [25]. In this metabolic radiolabeling procedure. cells are first incubated in medium containing 'H-proline (L-[2.3,4,S3H] proline, 102 Cilmmol, Amersham, Piscataway, NJ) which is incorporated into collagen and other proteins. 'H-prolinelabeled collagens are deposited outside the cells where they form an insoluble ECM. In our experiments cultures were incubated following treatment in labeling medium (culture medium containing 20 pCilml 'H-proline and 25 pg/ml ascorbate) (Sigma) on days 2 and 8 for 16 h. This allowed sufficient time for the 'H-proline to be incorporated into proteins and for the labeled proteins to be secreted and added to the ECM. Thus, after 16 h in labeling medium (on days 3 and 9) the cells had established an essentially insoluble ECM containing 'H-prolinelabeled collagens. The labeling medium was then replaced with fresh medium containing unlabeled proline. All proteins synthesized after the medium change were not labeled with 3H-proline. Three hours later the cells were subjected to different loading regimes. After loading the cultures were incubated for an additional 36 h to allow sufficient time for the matrix destabilizing effects of loading to result in the release of soluble. 'H-proline-labeled collagen peptides into the medium. Such peptides may include the products of collagen fibril degradation and incompletely processed procollagens. Medium and cells were then ex-
Results Monoluyer annulus cells The results are indicated as mean 31 standard deviation (SD). The data in Fig. 2 show the total amount of 'H-proline incorporated at three and nine days (CPM/ pg DNA). The histogram shows that incorporation in group 111 was -1.4-fold greater than that in groups I1 and I on both days (3 and 9). The differences between group 111 and groups I1 or I were significant on both three and nine days (p < 0.05) but the difference between groups I1 and I was not significant. Fig. 3 shows the
0
400j
Day 3
Day9
300
I
n
111
Treatment Fig. 2. Total 'H-proline incorporated by monolayer annulus cells under no loading (group I: control). low level loading (group 11: 0.3 MPa, 1 Hz), and high level loading (group 111: 1.7 MPa. 20 Hz). Incorporation was measured after three and nine days of loading.
M. Kasra c’t al. I Journnl qf Orthopaedic Reseurch 21 (2003) 597403
600
9 were respectively lower and higher than those of day 3 (p < 0.05).
Day3
_. I
Three-dimensional nucleus cells The significant interactions among the 10 (loading amplitude, frequency) condition pairings are illustrated in Fig. 4. Examination of the released to total collagen ratio ( R = CPM,edlu,/CPM,o,aI) indicated no significant differences between the control group and the three groups with 0.75 MPa amplitudes. There was a significant drop (p < 0.05) in the ratio for the 1.5 MPa amplitude loading conditions, but no significant differences at this load level across the three frequency values. There was a further drop (p < 0.05) in the ratio for the 3.0 MPa amplitude loading conditions, and the cultures demonstrated monotonic decreases with increased frequency. In this case, the “loading amplitude, frequency” conditions (3.0, 1) and (3.0,20) were significantly different from each other (p < 0.05). In general, the results indicated a significant increase in the CPM,,,,, and CPM,,,, (better synthesis) and a decrease in the ratio ( R ) (less degradation) by increasing loading amplitude or frequency. The effect of frequency was significant only at the highest loading amplitude (3.0 MPa), while the effect of loading amplitude was significant at all loading frequencies (Fig. 4). Comparing the data of the three loading amplitudes (0.75, 1.5, and 3.0 MPa) in three groups, a significant negative correlation was found between ratio ( R ) and loading amplitude ( A ) . Fig. 5
Y
I1
I
111
Treatment Fig. 3. Percentage of ’H-proline-labeled protein released to the medium (ratio of released to total incorporated) by monolayer annulus cells under no loading (group I: control), low level loading (group 11: 0.3 MPa, 1 Hz), and high level loading (group 111: 1.7 MPa, 20 Hz). Incorporation was measured after three and nine days of loading.
percentage of counts in the medium to the total incorporated (‘h released = 100 x CPM,ed,,,/CPMt,t,t) on days 3 and 9. Although the histogram indicates that this percentage increased slightly from group I to groups I1 or 111, these differences were not significant. Therefore, the variations of total incorporated (Fig. 2) and released (Fig. 3) ’H-proline with loading condition follow the same trend for both days 3 and 9. However, for the loading groups I1 and 111 as well as the control group I, the amounts of total ‘H-proline incorporated (Fig. 2) and the percentage of released ’H-proline (Fig. 3) of day
0.6
0,X.*
Groups with any of these symbols in common are significantly different(p < 0.05) I
5m
O 0.5
I I
I
X
X,O 7
I I
x,0
I
m -
I
0
0 m
I
0.4
T
-
r
I I I
Y
0
c
2 0.3 U a,
In
8
0.2
??
L
0
0 .c
0.1
2 0.0
Loading amplitude, Frequency (MPa, Hz) Fig. 4. ’H-proline-labeled protein released to the medium (ratio of released to total ‘H-proline incorporated) versus loading condition (loading amplitude, frequency). The figure shows the change of ratio with loading frequency for different loading amplitudes (3-D nucleus cells). Dash lines separate the groups of loading conditions of different amplitudes. In this case, the groups with any of the symbols “ x ” and ‘.o” in common are significantly different. The decrease of ratio with increasing amplitude is significant at all the frequencies (p < 0.05). The symbol “*” indicates significant differences within the groups with the same loading amplitude. The decrease of ratio with increasing frequency is significant 0, < 0.02) only at the highest amplitude (3 MPa).
M. Kusru rt ul. I Journal of Orthopurciic Rrsrurclz 21 (2003) 597403 c
ar
0.55
U I
2
050-
0
R = - 0 1 A + 0 52 (r = 0 93,p < 0 05)
0.45 I
r-"
0.40
-
.h.
4
0.35 -
- 0.30 G)
-g 0 ._ +
0.25 0.20 0.15
Fig. 5. Variation of ratio of released collagen ( R ) versus loading amplitude ( A ) (MPa) within the frequency range of 1-20 Hz and loading amplitude range of 0.75-3.0 MPd. The ratio decreases significantly by increasing the loading amplitude (p < 0.05).
shows the linear regression plot describing the ratio ( R ) as a linear function of loading amplitude ( A ) within the frequency range of 1-20 Hz and loading amplitude range of 0.75-3.0 MPa: R
=
-0.1A
+ 0.52
(Y = 0.93,
p < 0.05)
Discussion In this study we mainly focused on introducing a method to investigate the effect dynamic hydrostatic pressure on intervertebral disc cells over a wide range of amplitude and frequency. Using a hydraulic chamber and transfer of load through a fluid medium, the stiffness of the system could be much higher than that of the traditional way of load transfer through the air in a pressure vessel. This allowed us to impose a wider range of amplitudes and frequencies. To show the applicability of our method, we used the device to assess the effects of human in vivo loading amplitudes [30-321 on monolayer cultures of rabbit outer annulus cells and 3-D cultures of rabbit nucleus pulposis cells in two different experiments with different protocols. Rabbit intervertebral discs undergo progressive degeneration of the entire inner annulus and nucleus, similar to those found in degenerative human intervertebral discs [20]. However, it is not known if the responses of rabbit disc cells to loading are similar to the responses of human cells. We found that high frequency, high amplitude hydrostatic stress treatment stimulated 3H-proline incorporation in monolayer cultured rabbit outer annulus cells whereas low amplitude, low frequency hydrostatic stress had little effect (Fig. 2). In contrast, neither treatment significantly affected the relative amount of radiolabel released from the cell layers, indicating that protein degradation and stability was unaffected. Analyzing the cell cultures of day 9 versus day 3, we found a reduction in the total
60 I
amount of 3H-proline incorporated (protein synthesis) (Fig. 2), and an increase in the percentage of 'H-proline released (protein degradation) (Fig. 3). This may indicate simply the degenerative effect of number of culturing days in our monolayer culture system rather than that of loading repetition. Because repetition (number of days) affected not only the loading groups I1 and 111 but also the no loading control group I (Figs. 2 and 3). In the 3-D nucleus culture, high amplitude and frequency increased proteir, synthesis and lowered protein degradation. In this case, loading amplitude had a stronger influence on cell response than that of loading frequency (Figs. 4 and 5). Our results suggested a linear relationship between the ratio of released 3H-proline (protein degradation) and the loading amplitude (Fig. 5). However, this linear relationship may only be valid for the loading condition and culturing system used in this study. This relationship does not account for any singularity due to the existence of a critical frequency within or above the range of the frequencies applied in this study (1-20 Hz). An example of a critical frequency may be the in-vivo resonant frequency of the human spine (5-8 Hz) [27]. The cumulative effect of loading (number of loading days) on cell response was not investigated for the 3-D nucleus culture. Several studies have previously shown that application of non-physiological loads can lead to cellular changes and even loss of cells in the intervertebral disc and surrounding tissues [14,21,24]. Lotz and Chin reported that the degree of apoptosis correlated with the magnitude and duration of the applied stress [21]. However, the results of another study [16] found that the proteoglycan content of disc under sustained compressive load increased, indicating that compression here led to stimulation of matrix 4ynthesis by disc cells. In partial agreement with that study we found that short-term (3day) application of loads at physiologic amplitude stimulated protein synthesis in both monolayer and 3-D cultures. However, increasing the number of loading treatments to nine days inhibited protein synthesis and stimulated protein degradation in annulus cells. Therefore, loading history or number of culturing days appears to play an important role in disc cell responses. These findings are consistent with data showing that long-term load-induced changes to the disc cells' environment can lead to cell death [ 17,181. It has also been shown that the same load will likely have different effects on cells from different regions of the disc. Cells from the nucleus, annulus, and endplate are phenotypically distinct, show differences in metabolic response to load [17], and are surrounded by matrices with different mechanical properties. Highly significant differences were observed in the morphology, cytoskeletal arrangement, and biomechanical properties of the nucleus pulposus cells as compared with anulus fibrosus or transition zone cells [7]. The loading response
602
M . KUVU et ul. I Journul of Orthopueriir Researcli 21 (2003) 5 9 7 4 0 3
and shape of a cell from a specific region (for example, nucleus region) may be different if prepared in different cultures of monolayer or 3-D. Human nucleus cells are spindle-shaped in monolayer culture and they become rounded when seeded into 3-D alginate culture [3]. Furthermore, the in vivo loading of cells in the annulus region is not purely hydrostatic and has a higher shear stress component than that in the nucleus region. In our study we compared the responses of cells from different regions (nucleus and outer annulus), and different culturing methods (monolayer and 3-D alginate) under hydrostatic pressure. The low and high dynamic loading conditions were qualitatively similar in both cases, but we used higher loading amplitude in the analysis of the 3-D nucleus cell culture. In both cases, the high loading condition significantly stimulated protein synthesis. However, decreased protein degradation under the high loading condition was only observed in the 3-D nucleus cell culture. The higher loading amplitude used for the 3-D cultures than for the monolayer cultures may also account for the different responses of the two culture systems. Increasing the loading amplitude is expected to stimulate stronger cell responses. However, above a certain loading amplitude (3 MPa in this study), loading frequency likely becomes also significant (Fig. 4). Our findings indicate that short-term application of high loading amplitudes and frequencies is likely to be beneficial to the disc by stimulating protein synthesis and reducing protein degradation. In contrast, the cumulative effects of prolonged loading may be destructive as observed in in vivo conditions [4,27,29,33].This study is expected to shed new light on the mechanism of vibration-induced intervertebral disc disorders at a cellular level. It also proposes that vibration loading can likely be used as a preventive or a treatment tool if applied in a controlled manner. It also introduces a new method for applying hydrostatic pressure with a wide range of frequency and amplitude on cell cultures to determine destructive and constructive loading amplitudes and frequencies.
Acknowledgements This work was supported by the departments of Biomedical Engineering and Orthopedic Surgery at the University of Iowa.
References [I] Adams MA, Freeman BJ, Morrison HP, Nelson IW, Dolan P. Mechanical initiation of intervertebral disc degeneration. Spine 200025: 1625-36. [2] Buckwalter JA. Aging and degeneration of the human intervertebral disc. Spine 1995;20:1307 14.
Chiba K, Anderson GB, Masuda K, Momohara S, Williams JM, Thonar EJ. A new culture system to study the metabolism of the intervertebral disc in vitro. Spine 1998;23: 1821-7. Dupuis H. Medical and occupational preconditions for vibrationinduced spinal disorders: occupational disease no. 21 10 in Germany. lilt Arch Occup Environ Health 1994:66:303-8. Frymoyer JW. Magnitude of the problem. In: Wiesel S, Weinstein JN, Herkowitz H, Dvorak J, Bell G, editors. The lumbar spine. Philadelphia: W.B. Sanders Company: 1996. p. 8-16. Gruber HE, Hanky Jr EN. Analysis of aging and degeneration of the human intervertebral disc. Comparison of surgical specimens with normal controls. Spine 1998:23:751-7. Guilak F. Ting-Beall HP, Baer AE. Trickey WR. Erickson GR, Setton LA. Viscoelastic properties of intervertebral disc cells. Identification of two biomechanically distinct cell populations. Spine 1999;24:2475-83. Guiot BH, Fessler RG. Molecular biology of degenerative disc disease. Neurosurgery 2000;47: 103440. Guo J, Jourdian GW, MacCallum DK. Culture and growth characteristics of chondrocytes encapsulated in alginate beads. Conn Tissue Res 1989;19:277-97. Handa T, Ishihara H. Ohshima H . Osada R, Tsuji H, Obata K. Effects of hydrostatic pressure on inatrix synthesis and matrix metalloproteinase production in the human lumbar intervertebral disc. Spine 1997:15:1085-91. Hascall VC, Calabro A, Midura RJ, Yanagishita M. Isolation and characterization of proteoglycans. Meth Enzymol I994;230:390417. Heinegard D, Paulsson M. Cartilage. Meth Enzymol 1987; 145:336-63. Homer HA. Roberts S, Bielby RC, Menage J, Evans H, Urban IP. Cells from different regions of the intervertebral disc: effect of culture system on matrix expression and cell phenotype. Spine 2OO2;27( 10):1018-28. Hutton WC. Elmer WA, Boden SD, Hyon S, Toribatake Y, Tornita K, et al. The effect of hydrostatic pressure on intervertebral disc metabolism. Spine 1999;24: 1507-15. Hutton WC, Toribatake Y, Elmer WA, Ganey T M , Tomita K, Whitesides TE. The effect of compressive force applied to the intervertebral disc in vivo. A study of proteoglycans and collagen. Spine 1998;23:2524~-37. latridis JC, Mente PL. Stokes IA, Aronsson DD. Alini M. Compression-induced changes in intervertebral disc properties in a rat tail model. Spine 1999;24:996-1002. lshihara H, McNally DS, Urban JP, Hall AC. Effects of hydrostatic pressure on matrix synthesis in different regions of the intervertebral disk. J Appl Physiol 1996:80:83946. Ishihara H , Warensjo K, Roberts S, Urban JP. Protoglycan synthesis in the intervertebral disc nucleus: the role of extracellular osrnolarity. Am J Physiol 1997;272:C1499-505. Lipman JM. Fluorophotometric quantitation of DNA in articular cartilage utilizing Hoechst 33258. Anal Biochem 1989;176:128-31. Lipson SJ, Muir H. 1980 Volvo award in basic science. Proteoglycans in experimental intervertebral disc degeneration. Spine I98 l;6: 194-210. Lotz JC, Chin J R . Intervertebral disc cell death is dependent on the magnitude and duration of spinal loading. Spine 2000; 25: 1477-83. Lotz JC, Colliou OK, Chin JR, Duncan NA, Liebenberg E. Compression-induced degeneration of the intervertebral disc: an in vivo mouse model and finite-element study. Spine 1998; 23: 2493-506. Lyons G, Eisenstein SM, Sweet MB. Biochemical changes in intervertebral disc degeneration. Biochim Biophys Acta 198 I ; 673(4):443-53. Mason RM, Palfrey AJ. Intervertebral disc degeneration in adult mice with hereditary kyphoscoliosis. J Orthop Res 1984;2:333 8.
M. Ktrsrcr et
til.
I Journrrl
of' Orthopurrlic Reseurdi
[25] Nakajima M. Welch DR, Belloni PN. Nicolson GL. Degradation of basement membrane type IV collagen and lung subendothelial matrix by rat mammary adenocarcinoma cell clones of differing metastatic potentials. Cancer Res 1987:47:486976. [26] Poiraudeau S, Monteiro 1. Anract P, Blanchard 0. Revel M, Corvol MT. Phenotypic characteristics of rabbit intervertebral disc cells. Comparison with cartilage cells from the same animals. Spine 1999;24:83744. [27] Pope MH. Magnusson M. Wilder DG. Kxppa Delta Award. Low back pain and whole body vibration. Clin Orthop 1998;354: 241 8. [28] Pope MH. Occupational hazards for low back pain. In: Wienstein JW. Gordon SL, editors. Low back pain. Illinois: AAOS: 1996.
21 (2003) 507 -603
603
[29] Randall J M , Matthews RT, Stiles MA. Resonant frequencies of standing humans. Ergonomics 1997;40:879-86. [30] Ranu HS. Measurement of pressures in the nucleus and within the annulus of the human spinal disc: due to extreme loading. Proc Inst Mech Eng 1991;205:141-6. [31] Sato K, Kikuchi S, Yonezawa T. In vivo intradiscal pressure measurement in healthy individuals and in patients with ongoing back problems. Spine 1999;24:2468-74. [32] Wilke HJ. Neef P. Caimi M, Hoogland T. Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755 62. [33] Zagorski J. Jakubowski R, Solecki L, Sadlo A, Kasperek W. Studies on the transmission of vibrations in human organism exposed to low-frequency whole-body vibration. Acta Physiol Pol I976:27:347-54.
Journal of Orthopaedic Research 21 (2003) 597-603
Journal of Orthopaedic Research www.elsevier.com/locate/orthres
Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells Mehran Kasra 'I
Vijay Goel ', James Martin ', Shea-Tien Wang d , Woosung Choi e, Joseph Buckwalter
Department of Mechanical, Aerospace and Biomedical Engineering, The Universitjj of Tennessee, Knoxville, TN 37969. USA Depurtment of Biomedical Ettgineering, University of Toledo, Toledo, OH 43606, USA D e y t m e n t of Orthopaedic Surgery, University of' Iowa Hospitals and Clinics. 1on.u City, I A 52242, USA Department of Truumatology & Orthopaedics, Veterans General Hospital-Taipei, Taipei, TaiMiun Department of Orthopaedic Surgery, Duejeon St. Mary's Hospitul, Catholic University of Koreu, Daejeon. South Korea
Received 17 May 2002; accepted 10 January 2003
Abstract
The pathogenesis of vibration-induced disorders of intervertebral disc at the cellular level is largely unknown. The objective of this study was to establish a method to investigate the ranges of constructive and destructive hydrostatic loading frequencies and amplitudes in preventing or inducing extracellular disc matrix degradation. Using a hydraulic chamber, normal rabbit intervertebral disc cells were tested under dynamic hydrostatic loading. Monolayer cultures of disc outer annulus cells and 3-dimensional (3-D) alginate cultures of disc nucleus pulposus cells were tested. Effects of different loading amplitudes (3-D culture, 0-3 MPa; monolayer, 0-1.7 MPa) and frequencies (1-20 Hz) on disc collagen and protein metabolism were investigated by measuring 3H-proline-labeled proteins associated with the cells in the extracellular matrix and release of 'H-proline-labeled molecules into culture medium. High frequency and high amplitude hydrostatic stress stimulated collagen synthesis in cultures of outer annulus cells whereas the lower amplitude and frequency hydrostatic stress had little effect. For the same loading duration and repetition, neither treatment significantly affected the relative amount of protein released from the cell layers, indicating that protein degradation and stability were unaffected. In the 3-D nucleus culture, higher amplitude and frequency increased synthesis rate and lowered degradation. In this case, loading amplitude had a stronger influence on cell response than that of loading frequency. Considering the ranges of loading amplitude and frequency used in this study, short-term application of high loading amplitudes and frequencies was beneficial in stimulation of protein synthesis and reduction of protein degradation. 0 2003 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. Kqwwrds: Dynamic; Hydrostatic pressure; Intervertebral disc; Cells; Rabbit
Introduction Low back pain is the primary cause of disability in active age group of the society, playing a major role in the medical, social and economic structure of industrial countries. The etiology of low back pain is still unclear and therapeutic approaches are often ineffective [5,28]. Human intervertebral discs undergo age-related degenerative changes that contribute to some of the most common causes of impairment and disability for middle aged and older persons, such as back pain [2]. Potential * Corresponding author. Tel.: + 1-865-974-7673; fax: + 1-865-9748564. E-mail address: [email protected] (M. Kasra).
causes of the age-related degeneration of intervertebral discs include declining nutrition, loss of viable cells, cell senescence, post-translational modification of matrix proteins, accumulation of degraded matrix molecules, and fatigue failure of the matrix. Degenerative changes include decreased diffusion, decreased cell viability, decreased proteoglycan synthesis, and alteration in collagen distribution [2,8]. In vivo intradiscal pressure measurement in healthy individuals and in patients with ongoing back problem showed that the intradiscal pressure in degenerated discs was significantly reduced compared with that of normal discs [31]. Changes in matrix compressive stress would inhibit disc cell metabolism throughout the disc, and could lead to progressive deterioration of the matrix. Surviving cells are
0736-0266/03/$ - see front matter 0 2003 Orthopaedic Research Society. Published by Elsevier Science Ltd. All rights reserved. doi: 10.10l6/S0736-0266(03)00027-5
598
M. Kusra et ul. I Jourtiul of' Ortl?opuc.dic Rrsrurch 21 (2003) 597403
not synthetically inactive but are, rather, producing inappropriate matrix products during aging and degeneration [6]. Disc degeneration is a multifactorial phenomenon. Both biomechanical and biologic factors have been implicated in cases of accelerated degeneration [ 1,23,27]. Mechanical factors seem to play an important role. The intervertebral disc is routinely subjected to compressive loads that alter with posture and muscle activity and can produce pressures >2 MPa in human lumbar disks in vivo [30-321. Intervertebral disc pressure varies from about 0.19 at rest up to 3 MPa, above which vertebral fracture may occur [30]. Intervertebral disc cell death is dependent on the magnitude and duration of spinal loading [21,22]. Cell death, in turn, has been associated with disc degeneration in humans [2,6]. Chronically applied compressive forces and immobilization caused changes in the biomechanical behavior and biochemical composition of rat tail intervertebral discs [15]. Therefore, maintenance of appropriate stress or hydrostatic pressure within the disc may be an important basis for strategies to alleviate disc degeneration and initiate disc repair. Different investigators have shown that hydrostatic pressure directly affects the synthesis of collagen and proteoglycan by the intervertebral disc cells [3,10,14,17]. These studies suggested that a physiologic level of hydrostatic pressure might act as an anabolic factor for stimulation of proteoglycan synthesis. This may be essential for maintaining the matrix of the disc. However abnormal pressures, higher or lower pressure than physiological levels, may have a catabolic effect such as reduction of proteoglycan synthesis rate. These studies investigated only the effect of quasi-static hydrostatic loads. During daily occupational activities intervertebral disc is exposed to dynamic oscillatory hydrostatic loads, characterized by wide frequency spectrum and variable amplitude. A physiologic level of amplitude and frequency of hydrostatic pressure may be essential for maintaining the matrix of the disc, while an abnormal amplitude and frequency of hydrostatic pressure may accelerate disc degeneration. Long-term occupational exposure to whole-body vibration increases the risk of disc degeneration and the consequent back pain [4,27, 29,331. The pathogenesis of vibration-induced disorders is still not completely clear and there is no effective treatment. Although the potential effects of vibrational stress on extracellular matrix (ECM) assembly and degradation are particularly relevant to the clinical findings of the vibration-induced disorders, the effects of vibrational loads on disc cells are largely unknown. Considering the wide range of amplitude and frequency spectrum of in vivo dynamic loading on the human spine, the objective of this study was to establish a method to investigate the duration and ranges of
constructive and destructive loading frequencies and amplitudes in preventing or inducing extracellular disc matrix (ECM) degradation. We also investigated which of loading frequency and loading amplitude has more influence in affecting disc cell response to vibration.
Materials and methods Both monolayer and 3-D culture systems were tested. Disc outer annulus cells were tested in a monolayer culture and disc nucleus cells in a 3-D alginate culture. The use of different culture formats for these two cell types is based on findings which show that in vivo conditions for annulus cells are well approximated in monolayer culture while in vivo conditions for nucleus cells are better reproduced in alginate culture [13]. Rabbit disc cells were used in this study [?0,26]. Effects of two different loading levels, low and high, were first investigated by testing the monolayer disc annulus culture. In this case the effect of different loading repetition (number of days) was also investigated. This was then followed by a separate experiment with a different protocol testing 3-D nucleus cell cultures at different loading levels of low, medium. and high frequencies and amplitudes. Cells cultured in tissue culture dishes (Costar, Corning. N Y ) were placed in a hydraulic chamber filled with culture medium. The load was imposed through the fluid medium by a piston. The piston chamber assembly was sealed tight, bleeding any air out of the system. The chamber-cell assembly was fixed in a servo-hydraulic MTS mechanical testing system (MTS Corp., Minnesota) and a haversine compressive cyclic load was applied by the machine actuator (Fig. I ) . The applied hydrostatic pressure was calculated by dividing the applied load by the piston cross-sectional area. Monoltiycw disc unnu1u.r c,rll.s
Cells were isolated by sequential digestion of adult rabbit (New Zealaiid White) annulus fibrosis with bovine testicular hyaluronidase (I600 plml. 60 min) (Sigma, St. Louis, MO) followed by collagenase
Fig. I . Piston-chamber assembly installed in an Instron servohydraulic mechanical testing system. A haversine compressive cyclic load was applied by the machine actuator on the piston. The piston transferred the load to the cells placed in the chamber filled with medium.
599
M. Kusru e l ul. I Journul u f Orlhopaedic Resrurch 21 (2003) 597-603
type IA + pronase E (0.5 mglml, 16 h) (Sigma). First passage cells were seeded in 35 mm culture dishes (600,000 cells/dish) in medium (Dulbecco's Modified Eagle MediumllO%t fetal calf serum) (Life Technologies, Rockville. MD) and incubated overnight before the first hydrostatic stress treatment. The first passage cells were divided in three groups (seven dishes in each group), group I was the control group, group I1 was exposed to a cyclic hydrostatic pressure of 0.3 MPa amplitude and 1 Hz frequency simulating a low level physiological loading condition, and group 111 was exposed to a preload pressure of 0.8 MPa and a cyclic haversine pressure of 1.7 MPa at 20 Hz frequency (maximum of 2.5 MPa) simulating a high level loading condition within an in vivo physiological level [30-321. The cell cultures of different groups were loaded in the hydraulic chamber (Fig. 1) daily for 30 min. After three days three dishes of each groups were removed for analyses and the remainder four dishes were kept loaded up to nine days. In this case, the effect of loading repetition (number of loading days) was observed.
tracted in 4 M guanidine buffer (4 M guanidine HCI, I.O'%] Triton X-100, I0 mM EDTA. 0.8 mg/ml benzamidine, 1.0 mM phenylmethylsulfonyl fluoride. 50 mM sodium acetate; pH 6.8) and the extracts were fractionated on Sephadex G-50 (Amersham Pharmacia, Piscataway, NJ) columns. The column procedure removes any free (unincorporated) 'H-proline from both the medium and cell extracts so that all CPM values reflect only the fraction of 'H-proline that was incorporated into protein [I 1,121. Column eluates were analyzed by liquid scintillation counting to determine counts per minute (CPM) and DNA was measured in the cell extract by fluoronietric assay [19]. CPM data were normalized to DNA (CPM/pg) to control for differences in cell number from culture to culture. The sum of the CPM in both medium and extract (CPM,,,,I) was calculated to determine the amount of collagens synthesized during the labeling period. The ratio ( R ) of CPM released to the medium (CPM,,,,,,,) to the total CPM, (CPM,,,d,,,,,,lCPM,,,,i), was then calculated to determine the amount of collagen degraded after loading.
Tlir.ee-~lir~irn,sio~~~~l disc nuc1c~u.scc4l.s
Stutisticul anulysrs
Rabbit disc nucleus cells were maintained in monolayer culture for up to one week after their isolation. The monolayer cultures were trypsinized and the cells counted. Alginate cultures were prepared essentially as described in the literature [9]. Trypsinized disc cells were pelleted by centrifugation and the cell pellets re-suspended at a concentration of 1.5 x loh cells per ml in a sterile solution consisting of I50 mM NaCl and I .2'%1(w:v) alginate (Kelco, San Diego, CA). Individual alginate beads containing -25,000 cells each were formed by dripping the suspension through a 20 gauge needle into a shallow dish containing 102 mM CaCl?. Freshly polymerized beads were washed in Hanks balanced salt solution and transferred to a 96 well tissue culture plate ( 1 bead per well of 5-mm diameter). 0.25 ml medium was added and the cultures were incubated at 37 "C in a humidified chamber with 5% C 0 2 . Alginate cultures were divided into 10 groups (four dishes in each group) and exposed to different cyclic hydrostatic loading conditions simulating normal and abnormal loading condition within in vivo physiological levels. The applied dynamic hydrostatic pressure had a static component (preload) plus a haversine dynamic component (amplitude) both increasing from low to high levels. Three levels of high (H), medium (M), and low (L) amplitudes (A) (preload + amplitude; HA = 2 + 3. MA = I + I .5, LA =0.3 + 0.75 MPa) and frequencies (F) ( H F = 2 0 , M F = 10, L F = 1 Hz) were chosen. Group I was exposed to the high amplitude and high frequency (HA. HF); group 2 (HA. MF); group 3 (HA, LF); group 4 (MA, HF); group 5 (MA. MF); group 6 (MA, LF); group 7 (LA, HF); group 8 (LA, MF); group 9 (LA, LF) and group 10 was the control group. The cell cultures of different groups were placed in the hydraulic chamber (Fig. I ) and loaded daily for 30 min. After three days dishes of each groups were removed l o r analyses.
For both monolayer and 3-D alginate experiments, analysis of variance (ANOVA) with a significance level of p < 0.05 was first performed to detect the existence of any significant differences among different groups. Tukey's multiple comparison test was then used to determine significant differences between any two groups. For the 3-D alginate experiment, Pearson's correlation test was performed to determine any significant correlations between the dependent variable, ratio (R). and the independent variable, loading amplitude ( A ) . When there was a significant correlation present (p < 0.05). linear regression technique was used to find a relation between the dependent and independent variables.
of col/u,yen i~rc/uhoii.snr
Loading effects on collagen synthesis and degradation were determined essentially as described [25]. In this metabolic radiolabeling procedure. cells are first incubated in medium containing 'H-proline (L-[2.3,4,S3H] proline, 102 Cilmmol, Amersham, Piscataway, NJ) which is incorporated into collagen and other proteins. 'H-prolinelabeled collagens are deposited outside the cells where they form an insoluble ECM. In our experiments cultures were incubated following treatment in labeling medium (culture medium containing 20 pCilml 'H-proline and 25 pg/ml ascorbate) (Sigma) on days 2 and 8 for 16 h. This allowed sufficient time for the 'H-proline to be incorporated into proteins and for the labeled proteins to be secreted and added to the ECM. Thus, after 16 h in labeling medium (on days 3 and 9) the cells had established an essentially insoluble ECM containing 'H-prolinelabeled collagens. The labeling medium was then replaced with fresh medium containing unlabeled proline. All proteins synthesized after the medium change were not labeled with 3H-proline. Three hours later the cells were subjected to different loading regimes. After loading the cultures were incubated for an additional 36 h to allow sufficient time for the matrix destabilizing effects of loading to result in the release of soluble. 'H-proline-labeled collagen peptides into the medium. Such peptides may include the products of collagen fibril degradation and incompletely processed procollagens. Medium and cells were then ex-
Results Monoluyer annulus cells The results are indicated as mean 31 standard deviation (SD). The data in Fig. 2 show the total amount of 'H-proline incorporated at three and nine days (CPM/ pg DNA). The histogram shows that incorporation in group 111 was -1.4-fold greater than that in groups I1 and I on both days (3 and 9). The differences between group 111 and groups I1 or I were significant on both three and nine days (p < 0.05) but the difference between groups I1 and I was not significant. Fig. 3 shows the
0
400j
Day 3
Day9
300
I
n
111
Treatment Fig. 2. Total 'H-proline incorporated by monolayer annulus cells under no loading (group I: control). low level loading (group 11: 0.3 MPa, 1 Hz), and high level loading (group 111: 1.7 MPa. 20 Hz). Incorporation was measured after three and nine days of loading.
M. Kasra c’t al. I Journnl qf Orthopaedic Reseurch 21 (2003) 597403
600
9 were respectively lower and higher than those of day 3 (p < 0.05).
Day3
_. I
Three-dimensional nucleus cells The significant interactions among the 10 (loading amplitude, frequency) condition pairings are illustrated in Fig. 4. Examination of the released to total collagen ratio ( R = CPM,edlu,/CPM,o,aI) indicated no significant differences between the control group and the three groups with 0.75 MPa amplitudes. There was a significant drop (p < 0.05) in the ratio for the 1.5 MPa amplitude loading conditions, but no significant differences at this load level across the three frequency values. There was a further drop (p < 0.05) in the ratio for the 3.0 MPa amplitude loading conditions, and the cultures demonstrated monotonic decreases with increased frequency. In this case, the “loading amplitude, frequency” conditions (3.0, 1) and (3.0,20) were significantly different from each other (p < 0.05). In general, the results indicated a significant increase in the CPM,,,,, and CPM,,,, (better synthesis) and a decrease in the ratio ( R ) (less degradation) by increasing loading amplitude or frequency. The effect of frequency was significant only at the highest loading amplitude (3.0 MPa), while the effect of loading amplitude was significant at all loading frequencies (Fig. 4). Comparing the data of the three loading amplitudes (0.75, 1.5, and 3.0 MPa) in three groups, a significant negative correlation was found between ratio ( R ) and loading amplitude ( A ) . Fig. 5
Y
I1
I
111
Treatment Fig. 3. Percentage of ’H-proline-labeled protein released to the medium (ratio of released to total incorporated) by monolayer annulus cells under no loading (group I: control), low level loading (group 11: 0.3 MPa, 1 Hz), and high level loading (group 111: 1.7 MPa, 20 Hz). Incorporation was measured after three and nine days of loading.
percentage of counts in the medium to the total incorporated (‘h released = 100 x CPM,ed,,,/CPMt,t,t) on days 3 and 9. Although the histogram indicates that this percentage increased slightly from group I to groups I1 or 111, these differences were not significant. Therefore, the variations of total incorporated (Fig. 2) and released (Fig. 3) ’H-proline with loading condition follow the same trend for both days 3 and 9. However, for the loading groups I1 and 111 as well as the control group I, the amounts of total ‘H-proline incorporated (Fig. 2) and the percentage of released ’H-proline (Fig. 3) of day
0.6
0,X.*
Groups with any of these symbols in common are significantly different(p < 0.05) I
5m
O 0.5
I I
I
X
X,O 7
I I
x,0
I
m -
I
0
0 m
I
0.4
T
-
r
I I I
Y
0
c
2 0.3 U a,
In
8
0.2
??
L
0
0 .c
0.1
2 0.0
Loading amplitude, Frequency (MPa, Hz) Fig. 4. ’H-proline-labeled protein released to the medium (ratio of released to total ‘H-proline incorporated) versus loading condition (loading amplitude, frequency). The figure shows the change of ratio with loading frequency for different loading amplitudes (3-D nucleus cells). Dash lines separate the groups of loading conditions of different amplitudes. In this case, the groups with any of the symbols “ x ” and ‘.o” in common are significantly different. The decrease of ratio with increasing amplitude is significant at all the frequencies (p < 0.05). The symbol “*” indicates significant differences within the groups with the same loading amplitude. The decrease of ratio with increasing frequency is significant 0, < 0.02) only at the highest amplitude (3 MPa).
M. Kusru rt ul. I Journal of Orthopurciic Rrsrurclz 21 (2003) 597403 c
ar
0.55
U I
2
050-
0
R = - 0 1 A + 0 52 (r = 0 93,p < 0 05)
0.45 I
r-"
0.40
-
.h.
4
0.35 -
- 0.30 G)
-g 0 ._ +
0.25 0.20 0.15
Fig. 5. Variation of ratio of released collagen ( R ) versus loading amplitude ( A ) (MPa) within the frequency range of 1-20 Hz and loading amplitude range of 0.75-3.0 MPd. The ratio decreases significantly by increasing the loading amplitude (p < 0.05).
shows the linear regression plot describing the ratio ( R ) as a linear function of loading amplitude ( A ) within the frequency range of 1-20 Hz and loading amplitude range of 0.75-3.0 MPa: R
=
-0.1A
+ 0.52
(Y = 0.93,
p < 0.05)
Discussion In this study we mainly focused on introducing a method to investigate the effect dynamic hydrostatic pressure on intervertebral disc cells over a wide range of amplitude and frequency. Using a hydraulic chamber and transfer of load through a fluid medium, the stiffness of the system could be much higher than that of the traditional way of load transfer through the air in a pressure vessel. This allowed us to impose a wider range of amplitudes and frequencies. To show the applicability of our method, we used the device to assess the effects of human in vivo loading amplitudes [30-321 on monolayer cultures of rabbit outer annulus cells and 3-D cultures of rabbit nucleus pulposis cells in two different experiments with different protocols. Rabbit intervertebral discs undergo progressive degeneration of the entire inner annulus and nucleus, similar to those found in degenerative human intervertebral discs [20]. However, it is not known if the responses of rabbit disc cells to loading are similar to the responses of human cells. We found that high frequency, high amplitude hydrostatic stress treatment stimulated 3H-proline incorporation in monolayer cultured rabbit outer annulus cells whereas low amplitude, low frequency hydrostatic stress had little effect (Fig. 2). In contrast, neither treatment significantly affected the relative amount of radiolabel released from the cell layers, indicating that protein degradation and stability was unaffected. Analyzing the cell cultures of day 9 versus day 3, we found a reduction in the total
60 I
amount of 3H-proline incorporated (protein synthesis) (Fig. 2), and an increase in the percentage of 'H-proline released (protein degradation) (Fig. 3). This may indicate simply the degenerative effect of number of culturing days in our monolayer culture system rather than that of loading repetition. Because repetition (number of days) affected not only the loading groups I1 and 111 but also the no loading control group I (Figs. 2 and 3). In the 3-D nucleus culture, high amplitude and frequency increased proteir, synthesis and lowered protein degradation. In this case, loading amplitude had a stronger influence on cell response than that of loading frequency (Figs. 4 and 5). Our results suggested a linear relationship between the ratio of released 3H-proline (protein degradation) and the loading amplitude (Fig. 5). However, this linear relationship may only be valid for the loading condition and culturing system used in this study. This relationship does not account for any singularity due to the existence of a critical frequency within or above the range of the frequencies applied in this study (1-20 Hz). An example of a critical frequency may be the in-vivo resonant frequency of the human spine (5-8 Hz) [27]. The cumulative effect of loading (number of loading days) on cell response was not investigated for the 3-D nucleus culture. Several studies have previously shown that application of non-physiological loads can lead to cellular changes and even loss of cells in the intervertebral disc and surrounding tissues [14,21,24]. Lotz and Chin reported that the degree of apoptosis correlated with the magnitude and duration of the applied stress [21]. However, the results of another study [16] found that the proteoglycan content of disc under sustained compressive load increased, indicating that compression here led to stimulation of matrix 4ynthesis by disc cells. In partial agreement with that study we found that short-term (3day) application of loads at physiologic amplitude stimulated protein synthesis in both monolayer and 3-D cultures. However, increasing the number of loading treatments to nine days inhibited protein synthesis and stimulated protein degradation in annulus cells. Therefore, loading history or number of culturing days appears to play an important role in disc cell responses. These findings are consistent with data showing that long-term load-induced changes to the disc cells' environment can lead to cell death [ 17,181. It has also been shown that the same load will likely have different effects on cells from different regions of the disc. Cells from the nucleus, annulus, and endplate are phenotypically distinct, show differences in metabolic response to load [17], and are surrounded by matrices with different mechanical properties. Highly significant differences were observed in the morphology, cytoskeletal arrangement, and biomechanical properties of the nucleus pulposus cells as compared with anulus fibrosus or transition zone cells [7]. The loading response
602
M . KUVU et ul. I Journul of Orthopueriir Researcli 21 (2003) 5 9 7 4 0 3
and shape of a cell from a specific region (for example, nucleus region) may be different if prepared in different cultures of monolayer or 3-D. Human nucleus cells are spindle-shaped in monolayer culture and they become rounded when seeded into 3-D alginate culture [3]. Furthermore, the in vivo loading of cells in the annulus region is not purely hydrostatic and has a higher shear stress component than that in the nucleus region. In our study we compared the responses of cells from different regions (nucleus and outer annulus), and different culturing methods (monolayer and 3-D alginate) under hydrostatic pressure. The low and high dynamic loading conditions were qualitatively similar in both cases, but we used higher loading amplitude in the analysis of the 3-D nucleus cell culture. In both cases, the high loading condition significantly stimulated protein synthesis. However, decreased protein degradation under the high loading condition was only observed in the 3-D nucleus cell culture. The higher loading amplitude used for the 3-D cultures than for the monolayer cultures may also account for the different responses of the two culture systems. Increasing the loading amplitude is expected to stimulate stronger cell responses. However, above a certain loading amplitude (3 MPa in this study), loading frequency likely becomes also significant (Fig. 4). Our findings indicate that short-term application of high loading amplitudes and frequencies is likely to be beneficial to the disc by stimulating protein synthesis and reducing protein degradation. In contrast, the cumulative effects of prolonged loading may be destructive as observed in in vivo conditions [4,27,29,33].This study is expected to shed new light on the mechanism of vibration-induced intervertebral disc disorders at a cellular level. It also proposes that vibration loading can likely be used as a preventive or a treatment tool if applied in a controlled manner. It also introduces a new method for applying hydrostatic pressure with a wide range of frequency and amplitude on cell cultures to determine destructive and constructive loading amplitudes and frequencies.
Acknowledgements This work was supported by the departments of Biomedical Engineering and Orthopedic Surgery at the University of Iowa.
References [I] Adams MA, Freeman BJ, Morrison HP, Nelson IW, Dolan P. Mechanical initiation of intervertebral disc degeneration. Spine 200025: 1625-36. [2] Buckwalter JA. Aging and degeneration of the human intervertebral disc. Spine 1995;20:1307 14.
Chiba K, Anderson GB, Masuda K, Momohara S, Williams JM, Thonar EJ. A new culture system to study the metabolism of the intervertebral disc in vitro. Spine 1998;23: 1821-7. Dupuis H. Medical and occupational preconditions for vibrationinduced spinal disorders: occupational disease no. 21 10 in Germany. lilt Arch Occup Environ Health 1994:66:303-8. Frymoyer JW. Magnitude of the problem. In: Wiesel S, Weinstein JN, Herkowitz H, Dvorak J, Bell G, editors. The lumbar spine. Philadelphia: W.B. Sanders Company: 1996. p. 8-16. Gruber HE, Hanky Jr EN. Analysis of aging and degeneration of the human intervertebral disc. Comparison of surgical specimens with normal controls. Spine 1998:23:751-7. Guilak F. Ting-Beall HP, Baer AE. Trickey WR. Erickson GR, Setton LA. Viscoelastic properties of intervertebral disc cells. Identification of two biomechanically distinct cell populations. Spine 1999;24:2475-83. Guiot BH, Fessler RG. Molecular biology of degenerative disc disease. Neurosurgery 2000;47: 103440. Guo J, Jourdian GW, MacCallum DK. Culture and growth characteristics of chondrocytes encapsulated in alginate beads. Conn Tissue Res 1989;19:277-97. Handa T, Ishihara H. Ohshima H . Osada R, Tsuji H, Obata K. Effects of hydrostatic pressure on inatrix synthesis and matrix metalloproteinase production in the human lumbar intervertebral disc. Spine 1997:15:1085-91. Hascall VC, Calabro A, Midura RJ, Yanagishita M. Isolation and characterization of proteoglycans. Meth Enzymol I994;230:390417. Heinegard D, Paulsson M. Cartilage. Meth Enzymol 1987; 145:336-63. Homer HA. Roberts S, Bielby RC, Menage J, Evans H, Urban IP. Cells from different regions of the intervertebral disc: effect of culture system on matrix expression and cell phenotype. Spine 2OO2;27( 10):1018-28. Hutton WC. Elmer WA, Boden SD, Hyon S, Toribatake Y, Tornita K, et al. The effect of hydrostatic pressure on intervertebral disc metabolism. Spine 1999;24: 1507-15. Hutton WC, Toribatake Y, Elmer WA, Ganey T M , Tomita K, Whitesides TE. The effect of compressive force applied to the intervertebral disc in vivo. A study of proteoglycans and collagen. Spine 1998;23:2524~-37. latridis JC, Mente PL. Stokes IA, Aronsson DD. Alini M. Compression-induced changes in intervertebral disc properties in a rat tail model. Spine 1999;24:996-1002. lshihara H, McNally DS, Urban JP, Hall AC. Effects of hydrostatic pressure on matrix synthesis in different regions of the intervertebral disk. J Appl Physiol 1996:80:83946. Ishihara H , Warensjo K, Roberts S, Urban JP. Protoglycan synthesis in the intervertebral disc nucleus: the role of extracellular osrnolarity. Am J Physiol 1997;272:C1499-505. Lipman JM. Fluorophotometric quantitation of DNA in articular cartilage utilizing Hoechst 33258. Anal Biochem 1989;176:128-31. Lipson SJ, Muir H. 1980 Volvo award in basic science. Proteoglycans in experimental intervertebral disc degeneration. Spine I98 l;6: 194-210. Lotz JC, Chin J R . Intervertebral disc cell death is dependent on the magnitude and duration of spinal loading. Spine 2000; 25: 1477-83. Lotz JC, Colliou OK, Chin JR, Duncan NA, Liebenberg E. Compression-induced degeneration of the intervertebral disc: an in vivo mouse model and finite-element study. Spine 1998; 23: 2493-506. Lyons G, Eisenstein SM, Sweet MB. Biochemical changes in intervertebral disc degeneration. Biochim Biophys Acta 198 I ; 673(4):443-53. Mason RM, Palfrey AJ. Intervertebral disc degeneration in adult mice with hereditary kyphoscoliosis. J Orthop Res 1984;2:333 8.
M. Ktrsrcr et
til.
I Journrrl
of' Orthopurrlic Reseurdi
[25] Nakajima M. Welch DR, Belloni PN. Nicolson GL. Degradation of basement membrane type IV collagen and lung subendothelial matrix by rat mammary adenocarcinoma cell clones of differing metastatic potentials. Cancer Res 1987:47:486976. [26] Poiraudeau S, Monteiro 1. Anract P, Blanchard 0. Revel M, Corvol MT. Phenotypic characteristics of rabbit intervertebral disc cells. Comparison with cartilage cells from the same animals. Spine 1999;24:83744. [27] Pope MH. Magnusson M. Wilder DG. Kxppa Delta Award. Low back pain and whole body vibration. Clin Orthop 1998;354: 241 8. [28] Pope MH. Occupational hazards for low back pain. In: Wienstein JW. Gordon SL, editors. Low back pain. Illinois: AAOS: 1996.
21 (2003) 507 -603
603
[29] Randall J M , Matthews RT, Stiles MA. Resonant frequencies of standing humans. Ergonomics 1997;40:879-86. [30] Ranu HS. Measurement of pressures in the nucleus and within the annulus of the human spinal disc: due to extreme loading. Proc Inst Mech Eng 1991;205:141-6. [31] Sato K, Kikuchi S, Yonezawa T. In vivo intradiscal pressure measurement in healthy individuals and in patients with ongoing back problems. Spine 1999;24:2468-74. [32] Wilke HJ. Neef P. Caimi M, Hoogland T. Claes LE. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine 1999;24:755 62. [33] Zagorski J. Jakubowski R, Solecki L, Sadlo A, Kasperek W. Studies on the transmission of vibrations in human organism exposed to low-frequency whole-body vibration. Acta Physiol Pol I976:27:347-54.