- Email: [email protected]
Osteoblast response to the elastic strain of metallic support
ARTICLE IN PRESS
Journal of Biomechanics 40 (2007) 554–560 www.elsevier.com/locate/jbiomech www.JBiomech.com
Osteoblast response to the elastic strain of metallic support M. Lewandowska-Szumie"a,b,, K. Sikorskic, A. Szummerc, Z. Lewandowskid, W. Marczyn´skie a
Department of Biophysics and Human Physiology, Medical University of Warsaw, Cha!ubin´skiego 5, 02-004 Warsaw, Poland b Department of Transplantology and Central Tissue Bank, Centre of Biostructure Research, Medical University of Warsaw, Cha!ubin´skiego 5, 02-004 Warsaw, Poland c Faculty of Materials Science and Engineering, Warsaw University of Technology, Warsaw, Poland d Institute of Social Medicine, Department of Epidemiology, Medical University of Warsaw, Warsaw, Poland e Department of Traumatology & Orthopaedics, Central Hospital Military School of Medicine, Poland Accepted 20 February 2006
Abstract It is known that metallic elements of joint endoprostheses undergo elastic strain due to their mechanical function. This is one of the factors which may be responsible for the loosening of endoprostheses. Since mechanisms involved in it remain unclear, it seems valuable to verify if cells responsible for bone regeneration are affected by a strain of the implant. Our experiment examines the influence of elastic strain applied to Ti6Al4V samples on osteoblasts cultured on their surface in vitro. Human bone-derived cells are observed in contact with metallic plates. Titanium alloy was chosen as a support since it is one of the most commonly used materials for stems in joint endoprostheses. Cyclic elastic deformation of 0.1% was applied to the support once daily for 7 days. Two thousand cycles were applied each time. Samples which were not subject to strain served as control. After the observation period XTT assay was performed, alkaline phosphatase activity as well as osteocalcin concentration and nitric oxide secretion were determined and compared with the results obtained in the control group. It was found that the number of viable cells in the mechanically stimulated population was significantly higher than in control, while both alkaline phosphatase activity and osteocalcin concentration were significantly lower in the experimental group. Nitric oxide secretion was found in the culture which was subject to elastic strain, but not in the control. The possible clinical implication is that elastic strain of the metallic endoprostheses may influence osteoblasts which are in contact with the implant in vivo. r 2006 Elsevier Ltd. All rights reserved. Keywords: Osteoblast; Cell culture; Titanium; Mechanical stimulation; Aseptic loosening
1. Introduction Most biocompatibility studies of biomaterials are performed under static conditions. This is an important limitation of preclinical studies of candidate materials destined for load-bearing intrabone implants, since the evaluation of direct contact between biomaterial and Corresponding author. Department of Biophysics and Human Physiology, Medical University of Warsaw, Cha"ubin´skiego 5, 02-004 Warsaw, Poland. Tel./fax: +48 22 628 78 46. E-mail address: [email protected] (M. Lewandowska-Szumie").
0021-9290/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2006.02.012
bone tissue is strongly affected by mechanical circumstances at the bone-implant interface. The consequences of micromotion between endoprosthesis and host tissues have been experimentally observed in some studies (Aspenberg and Herbertsson, 1996; Aspenberg et al., 1992; Jasty et al., 1997; Søballe et al., 1992). However, not much attention has been paid to tissue reaction to elastic strain of the implant which undergoes mechanical stress. This, for example, applies to the contact of metallic stem of joint endoprosthesis with the host bone. The key problem with this type of reconstruction is the so-called aseptic loosening which leads to the necessity of reoperation. Although the presence of wear particles
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
near the implant is considered to be the main factor responsible for this complication (National Institute of Health (NIH) Consensus Statement, 1994;), the influence of mechanical stress of endoprosthesis on surrounding tissues as a factor involved in loosening is not out of question. Nevertheless, experimental verification is not easy—firstly, due to the difficulties in the proper simulation of mechanical conditions which exist in clinical application of endoprostheses and secondly, due to the high complexity of factors which accompany experimental implantation. Thanks to the great development of cell culture techniques, new experimental possibilities of investigating the relationship between mechanical factors and cellular functions in living organisms appear. The aim of the present study is to examine the influence of elastic strain applied to the metallic support on osteoblast-like cells cultured on its surface in vitro. Titanium alloy, one of the most commonly used materials for stem in joint endoprostheses, has been used in order to discuss the potential clinical implications of the results.
2. Materials and methods 2.1. Cells used in the experiment Human bone derived cells (HBDCs) were isolated from pieces of tissue removed at surgery, which would otherwise be discarded. Material from a spinous process (female, 15 years old) was used. Experiments were repeated with cells harvested from fibula (female, 69 years old). The procedure of HBDCs’ isolation was based on the protocols described by Gallagher et al. (1996) with modifications (Kudelska-Mazur et al., 2001). Briefly, soft tissue found on the extracted bone pieces was removed by scraping, bone was cut into small pieces and treated with collagenase (type XIS, SIGMA) for 24 h. The bone pieces were put into the culture medium containing DMEM (GIBCO BRL) enriched with 10% of heat-inactivated fetal calf serum (GIBCO BRL), Antibiotic–Antimycotic (GIBCO BRL), L-glutamine (GIBCO BRL) and 100 mM L-ascorbic acid 2phosphate (SIGMA), and left in humidified atmosphere (5% CO2, 95% humidity) in order to allow osteoblasts to migrate from bone explants onto the surface of culture flasks, 75 cm2 (NUNC). If the amount of material was sufficient, pieces of bone harvested from one patient were located in two different flasks, in order to obtain two populations of cells originating from the same donor. After the confluence has been reached, cells were detached by means of collagenase and trypsin digestion, and used in experiments. During the experiment cells were cultured in a medium described above,
supplemented (SIGMA).
with
555
1a,25-dihydroxycholecalciferol
2.2. Samples of biomaterial Cells were seeded on the surface of samples made of commercial Ti6Al4V alloy. Titanium alloy was chosen as a substrate since it is one of the most commonly used materials for stems in joint endoprostheses. Samples were prepared in the form of 0.1-mm thick, 14-mm long and 5-mm wide plates. They were autoclaved and used as a support for cells without any additional treatment. Particularly, they were not conditioned with any media, and cells were seeded on dry surfaces. Cells were seeded on the surface of titanium plates put on the bottom of 24-well culture dishes in the density of 47 000 cells/cm2. The next day substrate-strain procedure was started in the manner described below. 2.3. Method of stimulation—system for elastic strain of titanium plates Titanium plate samples with cells seeded on them were placed on two parallel TEFLONs supports at the distance of 10 mm, which were located at the bottom of a 24-well culture dish. The support was made of TEFLONs since it is known as a biocompatible material and thus it was expected not to influence the system when being put in contact with culture medium. As confirmed in the preliminary studies (unpublished data), even direct contact of cells with TEFLONs substrate does not affect any of cell features which were investigated in the experiment. Elastic strain of the titanium plates was generated by cyclic stressing of the centre of plates by TEFLONs stamps with conical end mounted to the cover of the culture dish (schematic illustration at Fig. 1—Sikorski et al., Polish Patent P 339 046). The cyclic motion of the stamps was performed by pneumatically operated system in which an elastic bag located on the upper side of a cover with TEFLONs stamps mounted on it, provoked bending of titanium samples. In a consequence cyclical elastic strain of titanium plates was achieved. The upper surface of the plate, on which the cells are seeded, is subjected to compressive strain,
Fig. 1. Schematic illustration of the experimental system.
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
556
which varied along the length of the sample, from ‘‘emax’’ under the stamp to zero over the sample support. The maximal amplitude of the strain is calculated on the basis of measurements of sample deflection under the stamp using the following expression (coming from the beam theory): emax ¼
6hf 100 ð%Þ, l2
where h is the thickness of the sample, f the deflection of the sample measured under the stamp, and l the distance between the sample supports. The strain amplitude of the cycle can be adjusted by changing the plate deflection, which depends on the length of the TEFLONs stamps. In our experiment the deflection was 0.25 mm, which makes the true strain in the surface layer of the plate under the stamp equal to 0.1%. Two thousand strain cycles per day were applied, similarly, as in the experiments of Kaspar et al. (2000) and Stanford et al. (1995), in which osteoblasts cultured on the surface of polymers were stretched. The procedure was carried on for 3 min (which makes the frequency of 11 Hz) and repeated once daily throughout 7 days. The process was carried out in a laminar flow in order to provide sterile conditions, thus, cells were staying up to 10 min outside the incubator each day. Cells cultured under identical conditions on the surface of titanium plates (including removing from incubator, placing titanium plates with cells on them on Teflon supports, and using the same volume of culture medium), but not subject to strain, served as a control. Both in the experimental and in the control group, the cells remained in the same culture medium during the whole experiment. The medium was not exchanged in order to collect osteocalcin which was determined at the end of the experiment. On the next day after the 7-daylong strain series has been completed, supernatants were harvested from both experimental and control cultures in order to measure osteocalcin and nitric oxide concentration, and viability test was performed. As recommended by Gallagher at al. (1996), and confirmed in our previous experiments, a significant increase of the osteocalcin concentration in response to 1a,25-dihydroxycholecalciferol, characteristic for the osteogenic phenotype, should be expected after that time. The next day alkaline phosphatase activity was determined in all cultures.
2.5. Viability assay In order to compare the number of living cells remaining in the culture after the applied handling, XTT assay was performed. This test is used in toxicology and is based on the ability of mitochondrial dehydrogenase enzymes in living cells to convert the XTT substrate (2.3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenyloamino)carboxyl]-2H-tetrazolium hydroxide) into a water-soluble formazan product. The final product of the reaction is measured in the ELISA reader at 450 nm. Although the result of this test is not equal to the number of cells, there are reports which confirm that the production of formazans is proportional to the number of living cells (Breault et al., 1995; Doyle et al., 1996; Sadler et al., 1996), which makes this method useful in studies basing on the comparison of the number of living cells among groups. The number of cells was calculated from the calibration curve for XTT assay, which was obtained in a parallel experiment. After the XTT test, fresh medium was added to the culture in order to keep cells alive till the next day when the alkaline phosphatase activity was determined. 2.6. Alkaline phosphatase activity determination Alkaline phosphatase (ALP) activity was chosen to label the influence of the applied procedure on osteoblast in culture since it may indicate whether the function of osteoblasts is disturbed by the micromotion. Also, it is one of the most important promoters of osteogenesis in vivo, and thus—an important factor in predicting an in vivo situation on the base of the observations performed in cell culture. Since the product of the XTT assay is water soluble and the cells are not injured by the assay, the cells can be subsequently studied by other means. Thus, ALP activity was determined on the next day. This combination of XTT and ALP activity assay is routinely used in our laboratory, since it allows us to measure ALP activity exactly in the same population in which cell viability is measured by XTT assay. As found in numerous preliminary studies performed in order to establish the procedure, the ALP test is only slightly and insignificantly affected by a previous XTT assay when the ALP test is performed on the next day (data not shown). ALP activity was measured by means of Alkaline Phosphatase SIGMA Diagnostic Kit. 2.7. Nitric oxide measurement
2.4. Osteocalcin concentration measurement Osteocalcin concentration was determined in cell culture supernatants with N-MIDTM Osteocalcin One Step ELISA test (Osteometer Biotech A/s). The results were calibrated and expressed in ng/ml.
Nitric oxide (NO) presence was detected in supernatants harvested from both mechanically stimulated and control culture. Griess reagent was used (Lowik et al., 1994; Mac Pherson et al., 1999), which reacts with nitric oxide giving the stable product of characteristic
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
colour (sodium nitrite) which is then measured in the ELISA reader at 540 nm. 2.8. Statistical analysis The experiments were carried out 3 times, i.e. the first, preliminary, experiment showed the relations, which were then confirmed in the next two experiments. The results obtained in those two experiments were taken into calculations. Cells from one donor were used for the first two experiments (the preliminary and the first one of those taken into calculation) and cells from another donor were used in the last one, to make sure that the found effects are not accidental or connected with the cells from a particular donor. Seven samples in the experimental group (cells cultured on the support subjected to strain) and seven samples from the control were investigated in each of the experiments taken to calculation (which makes 14 measurements in each group). Each measurement was done on three aliquots from each sample. Results were analyzed using Two-way ANOVA in which cell viability, alkaline phosphatase activity, osteocalcin level, and NO secretion were considered as dependent variables. Two factors were taken into account: firstly, strain effect (a group subject to strain and non-strained control), and secondly, donor effect (cells were isolated from tissue harvested from two patients). Such conservative approach was applied in order to estimate the influence of the source of the cells (donor) on the results.
557
of the substrate (136% of control). On the contrary, both ALP activity and osteocalcin concentration are significantly higher ðpo0:001Þ in control i.e. in culture which was carried out on unstrained titanium plates. ALP activity value in the population subject to strain was at the level of 56% of control, and osteocalcin concentration, at the level of 63% of control. The number of cells, calculated from the calibration curve for the XTT assay, as well as the osteocalcin level and alkaline phosphatase activity in the population exposed to the elastic strain of titanium support are shown together as a percentage of untreated control in Fig. 2. Nitric oxide was practically not detected in control, while in the mechanically stimulated group some amount of NO was found and determined as the mean concentration of NO 2 in the well, which was about 2.3 mM. All found relations between values found in the experimental group vs. the control, were confirmed in all experiments. The influence of the source of the cells (donor) on the results was found to be insignificant.
4. Discussion and conclusions The results of viability assay indicate that the number of cells cultured on the titanium that underwent elastic strain is enhanced. The effect is even stronger than shown by quantitative data if we take into account the 160% 140%
133%
Results of quantitative analysis are gathered in Table 1. All found relations were confirmed in three experiments. For calculation presented in the table results from two experiments (14 samples which were subject to strain and 14 control-not subject to strain) were taken. Values of ALP activity, osteocalcin and NO concentrations were normalised to the number of cells calculated from the calibration curve for the XTT assay. Absorbance, read in XTT assay, is significantly ðpo0:0001Þ higher in culture which was subject to strain
% of control
120%
3. Results
100% 80%
63%
56%
60% 40% 20% 0%
cell number
ALP
osteocalcin
Fig. 2. Number of cells calculated from the calibration curve for the XTT assay, alkaline phosphatase activity and osteocalcin concentration—expressed as percentage of control (based on values presented in Table 1).
Table 1 Results of quantitative analysis: absorbance obtained in XTT assay, number of cells calculated from the calibration curve for the XTT assay, values of ALP activity, osteocalcin and NO concentrations normalized to the number of cells i.e. divided by (number of cells 104); shown are mean values (MV) and standard errors (SE) XTT assay (absorbance)
Number of cells
ALP/104 cells (U/L)
Osteocalcin/104 cells (mg/ml)
4 NO 2 /10 cells (mM)
MV
SE
MV
SE
MV
SE
MV
SE
MV
SE
0.014 0.007
56845 42720
1040 500
8.763 15.721
1.598 2.281
39.900 63.820
4.760 6.080
0.402 0.050
0.038 0.028
Subject to strain 0.725 Control 0.534
ARTICLE IN PRESS 558
M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
fact that some cells in the experimental group must have been damaged due to the contact with a TEFLONs stem, which performs the strain of the substrate. 11% of the whole sample surface was affected by this. It was estimated after the whole experiment by measuring the ‘‘empty’’, area (i.e. devoid of cells) visible in the central part of the samples which underwent elastic strain. Some cells from this area must have been killed, and others were probably dislocated. It was not possible to take this factor into account quantitatively. Anyway, if this drawback was avoided, the effect of the enhanced cell number after application of the strain to the support, would probably be even more evident, since only viable cells were taken into calculation. The enhanced number of cells in the group, in which the cell support underwent elastic strain, may result from cell proliferation or preventing cell death (by necrosis or apoptosis). As can be judged on the basis of preliminary observation of cells in contact with Ti6Al4 V plates under static conditions, the system is very well tolerated by cells, and during the 8-day long culture no signs of apoptosis were observed (unpublished data). So, it can be postulated that the applied procedure promotes cell proliferation. This finding is in good correlation with the results obtained in experiments of Stanford et al. (1995) and Kaspar et al. (2000) in which the conditions that exist during physiological strain of long bones were imitated in osteoblast culture on silicone substrate subject to strain. Also, similarly to the papers cited above, in our experiment, the enhanced proliferation of cells is accompanied by a significantly lower expression of phenotype features in the experimental group as compared to control. Both osteocalcin level and alkaline phosphatase activity are slightly, but statistically significantly, lower when elastic strain of metallic support is applied. Enhanced proliferation of HBDCs due to the applied strain was observed by Holbein et al. (1995), who also noticed that the addition of serum obtained from patients who used Ilizarov device resulted in the increase of proliferation in vitro of SaOS–line cells. A significant decrease in alkaline phosphatase activity and increase in cell number, as reflected by DNA content, was observed in a culture of primary rat osteoblast-like cells in response to intermittent stretching (Winter et al., 2003). Similarities of our results to those obtained in experiments of others—mainly designed to explain cellular mechanisms responsible for bone remodelling under loading—confirm the reliability of the proposed experimental system, which is to imitate the situation which exists at the interface of bone and load-bearing metallic implant. In addition to the enhanced proliferation and diminished osteocalcin release as well as alkaline phosphatase activity, nitric oxide secretion—in response
to the elastic strain of titanium substrate—was found in our experiment. Pitsillides et al. (1995) reported an analogous effect in a culture of chicken embryonic bone cells exposed to stretching of polymeric support. On the contrary, Smalt et al. (1997) did not observe such effect on osteoblast culture on stretched polystyrene. Although in the culture of the same cells exposed to fluid shear, NO concentration in the culture medium was found to be increased. In experiments by S.Kapur, addition of L-NAME—the inhibitor of all types of NO synthases—blocked the shear-stress-induced osteoblast proliferation and differentiation (Kapur et al., 2003). The potential role of NO in bone remodelling is a new factor of interest. Its role as a mediator in the osteoblast–osteoclast communication is postulated (Lowik et al., 1994). Although the mechanisms by which nitric oxide influences bone remodelling still remain unclear, many factors support its role in bone physiology and pathology. Nitric oxide secretion in HBDC in response to elastic strain of implantable metallic support found in our experiments indicates that it also may play some role in bone remodelling in contact with load-bearing intrabone metallic implants. Similarly, diminishing markers of osteoblast phenotype, i.e., alkaline phosphatase activity and production of osteocalcin, found in our observation, can be interpreted from the point of view of a clinical situation. HBDCs are a population of osteogenic cells at various stages of differentiation. A significant increase in both alkaline phosphatase activity and osteocalcin production in response to 1,25-dihydroxy-vitamin D3 confirms their potency to differentiation in vitro. The lower level of ALP activity and osteocalcin concentration in culture exposed to elastic strain of support indicates that such mechanical handling may slow down the differentiation of HBDCs. A higher proliferation of less differentiated cells in a mechanically stimulated system seems comprehensible if we take into account that terminally differentiated osteoblasts do not multiply. Such interpretation of results leads to potential clinical implications. Elastic strain of the metallic part of endoprosthesis may influence the osteoblasts which are in the contact with the implant in vivo and should be taken into account as one of the potential factors responsible for loosening the endoprosthesis. This is the main conclusion from the reported findings. It remains an open question to which extent experiments performed in cell culture are a good basis for understanding phenomena which take place in vivo. There are many variables which influence the experimental system: among them, strain level, the number of cycles, frequency, duration of observation, and others. In our experiment, in which special attention was paid to the similarity of the applied conditions to an in vivo situation, the selected frequency of cyclic loading needs some comment.
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
The applied frequency, equal to 11 Hz (2000 cycles for 3 min), is relatively high. The frequency of walking during normal physical activity of a human is 1 Hz (Bergmann et al., 2001). However, frequency components above 20 Hz were found by Stergiou et al. (2002) in frequency domain analysis of the ground reaction forces during walking and there were observed as high as 75 Hz frequency components associated with impulsive loading at heel strike during barefoot walking (Harris et al., 1996). Besides, there is plenty in vitro evidence for a cellular component in the response of bone cells to high frequency stimuli (Kaspar et al., 2002; Kasra et al., 2003; Tanaka et al., 2003), supporting the choice of a frequency higher than 1 Hz. As shown by Kaspar et al. (2002) HBDCs’ proliferative response to cyclic strain at frequencies ranging from 0.1 to 30 Hz, did not significantly depend on the frequency when the cycle number was left constant. The cycle number in our experiment was similar to those applied by Kaspar et al. (2000) and Stanford et al. (1995). Altogether, because using the frequency of 11 Hz is time efficient (the cells spent only 3 min outside the incubator) such frequency was used in our experiment. The weak point of our system is the fact that some amounts of cells undergo destruction as a consequence of the direct contact of the surface covered by cells with the TEFLONs stamp. The influence of dead cells on any factor measured in a culture cannot be completely excluded. On the other hand, to the best of our knowledge there is only one other system which allows to observe the influence of stretching nonflexible support on cells in vitro (Di Palma et al., 2003). This system is based on a Dynacells device, especially designed to apply cyclic strains on rigid biomaterials. It does not require direct pushing of the support, but unfortunately it requires samples of a big area (discs of 90 mm diameter). The area of a singular sample is little less than a 75 cm2 culture flask. In consequence, the application of this method is practically limited to the observations of cell lines because it is difficult to obtain a sufficient number of isolated from bone explants osteoblasts, to cover the investigated surface in a reliable number of repetitions. This limitation is severe, especially that the results obtained in a culture of cell lines may be completely different from the results in a culture of HBDCs. To sum up, there still is a need to work on an in vitro system which would entirely correspond to the situation of osteoblasts in direct contact with load-bearing implants which undergo elastic strain. The model proposed in the presented paper is one of possible responses for such requirement, and the basic finding— that osteoblasts respond to elastic strain of titanium in direct contact through an enhanced proliferation accompanied by a diminished differentiation—can be considered in the assessment of the consequences of elastic strain of metallic endoprosthesis in vivo.
559
Acknowledgments This work was supported by the State Committee of Scientific Research (Grant no. 4. PO5A.065.11). References Aspenberg, P., Herbertsson, P., 1996. Periprosthetic bone resorption. Particles versus movement. Journal of Bone and Joint Surgery— British Volume 78, 641–646. Aspenberg, P., Goodman, S., Toksvig-Larsen, S., Ryd, L., Albrektsson, T., 1992. Intermittent micromotion inhibits bone ingrowth, Titanium implants in rabbits. Acta Orthopaedica Scandinavica 63, 141–145. Bergmann, G., Deuretzbacher, G., Heller, M., Graichen, F., Rohlmann, A., Strauss, J., Duda, G.N., 2001. Hip contact forces andgait patterns from routine activities. Journal of Biomechanics 34, 859–871. Breault, L.G., Schuster, G.S., Billman, M.A., Hanson, B.S., Kudryk, V.L., Pashley, D.H., Runner, R.R., McPherson, J.C., 1995. The effects of intracanal medicaments, fillers, and sealers on the attachment of human gingival fibroblasts to an exposed dentin surface free of a smear layer. Journal of Periodontology 66, 545–551. Di Palma, F., Douet, M., Boachon, C., Guignandon, A., Peyroche, S., Forest, B., Alexandre, C., Chamson, A., Rattner, A., 2003. Physiological strains induce differentiation in human osteoblasts cultured on orthopaedic biomaterial. Biomaterials 24, 3139–3151. Doyle, A., Griffiths, J.B., Newell, D.G. (Eds.), 1996. Core Techniques. Cell & Tissue Culture: Laboratory Procedures. Wiley, West Sussex, England, pp. 4B:6.1–4B:6.2. Gallagher, J.A., Gundle, R., Beresford, J.N., 1996. Isolation and culture of bone-forming cells (osteoblasts) from human bone. In: Jones, G.E. (Ed.), Methods in Molecular Medicine: Human Cell Culture Protocols. Humana Press Inc., Totowa, NJ, pp. 233–262. Harris, G.F., Acharya, K.R., Bachschmidt, R.A., 1996. Investigation of spectral content from discrete plantar areas during adult gait: an expansion of rehabilitation technology. IEEE Transactions on Rehabilitation Engineering 4, 360–374. Holbein, O., Neidlinger-Wilke, C., Suger, G., Kinzl, L., Claes, L., 1995. Ilizarov callus distraction produces systemic bone cell mitogens. Journal of Orthopaedic Research 13, 629–639. Jasty, M., Bragdon, C., Burke, D., O’Connor, D., Lowenstein, J., Harris, W.H., 1997. In vivo skeletal responses to porous-surfaced implants subjected to small induced motions. Journal of Bone and Joint Surgery—American Volume 79, 707–714. Kapur, S., Baylink, D.J., Lau, K.-H.W., 2003. Fluid flow shear stress stimulates human osteoblast proliferation and differentiation through multiple interacting and competing signal transduction pathways. Bone 32, 241–251. Kaspar, D., Seidl, W., Neidlinger-Wilke, C., Ignatius, A., Claes, L., 2000. Dynamic cell stretching increases human osteoblast proliferation and CICP synthesis but decreases osteocalcin synthesis and alkaline phosphatase activity. Journal of Biomechanics 33, 45–51. Kaspar, D., Seidl, W., Neidlinger-Wilke, C., Beck, A., Claes, L., Ignatius, A., 2002. Proliferation of human-derived osteoblast-like cells depends on the cycle number and frequency of uniaxial strain. Journal of Biomechanics 35, 873–880. Kasra, M., Goel, V., Martin, J., Wang, S.T., Choi, W., Buckwalter, J., 2003. Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells. Journal of Orthopaedic Research 21, 597–603. Kudelska-Mazur, D., Lewandowska-Szumie", M., Benke, G., Kowalski, M., Komender, J., 2001. Isolation of human primary osteoblsts for biomaterials investigation in vitro. Engineering of Biomaterials 15&16, 51–58.
ARTICLE IN PRESS 560
M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
Lowik, C.W., Nibbering, P.H., van de Ruit, M., Papapoulos, S.E., 1994. Inducible Production of Nitric Oxide in Osteoblast-like Cells and in Fetal Mouse Bone Resorption. Journal of Clinical Investigation 93, 1465–1472. Mac Pherson, H., Noble, B.S., Ralston, S.H., 1999. Expression and functional role of nitric oxide synthase isoforms in human osteoblast-like cells. Bone 24, 179–185. National Institute of Health Consensus Statement, 1994. Total Hip Replacement 12(5), 1–31; published also in The Journal of American Medical Associations 1995, 273, 1950–1956. Pitsillides, A.A., Rawlinson, S.C., Suswillo, R.F., Bourrin, S., Zaman, G., Lanyon, L.E., 1995. Mechanical strain-induced NO production by bone cells: a possible role in adaptive bone (re)modeling? Faseb Journal 9, 1614–1622. Sadler, G.P., Jones, D.L., Woodhead, S., Horgan, K., Wheeler, M.H., 1996. Effect of growth factors on growth of bovine parathyroid cells in serum-free medium. World Journal of Surgery 20, 822–829. Sikorski, K., Lewandowska-Szumie", M., Szummer, A., Polish Patent P 339 046 ‘‘Urza˛ dzenie do badania z˙ ywych komo´rek w kontakcie z biomateria"ami in vitro’’ (Apparatus for in vitro investigations of cells in contact with biomaterials).
Smalt, R., Mitchell, F.T., Howard, R.L., Chambers, T.J., 1997. Induction of NO and prostaglandin E2 in osteoblasts by wall-shear stress but not mechanical strain. American Journal of Physiology 273, 751–758. Søballe, K., Hansen, E.S., Brockstedt-Rasmussen, H., Jørgensen, P.H., Bu¨nger, C., 1992. Tissue ingrowth into titanium and hydroxyapatite-coated implants during stable and unstable conditions. Journal of Orthopaedic Research 10, 285–299. Stanford, C.M., Morcuende, J.A., Brand, R.A., 1995. Proliferative and phenotypic responses of bone-like cells to mechanical deformation. Journal of Orthopaedic Research 13, 664–670. Stergiou, N., Giakas, G., Byrne, J.B., Pomeroy, V., 2002. Frequency domain characteristics of ground reaction forces during walking of young and elderly females. Clinical Biomechanics (Bristol, Avon) 17, 615–617. Tanaka, S.M., Li, J., Duncan, R.L., Yokota, H., Burr, D.B., Turner, C.H., 2003. Effects of broad frequency vibration on cultured osteoblasts. Journal of Biomechanics 36, 73–80. Winter, L.C., Walboomers, X.F., Bumgardner, J.D., Jansen, J.A., 2003. Intermittent versus continuous stretching effects on osteoblast-like cells in vitro. Journal of Biomedical Materials ResearchApplied Biomaterials 67, 1269–1275.
Journal of Biomechanics 40 (2007) 554–560 www.elsevier.com/locate/jbiomech www.JBiomech.com
Osteoblast response to the elastic strain of metallic support M. Lewandowska-Szumie"a,b,, K. Sikorskic, A. Szummerc, Z. Lewandowskid, W. Marczyn´skie a
Department of Biophysics and Human Physiology, Medical University of Warsaw, Cha!ubin´skiego 5, 02-004 Warsaw, Poland b Department of Transplantology and Central Tissue Bank, Centre of Biostructure Research, Medical University of Warsaw, Cha!ubin´skiego 5, 02-004 Warsaw, Poland c Faculty of Materials Science and Engineering, Warsaw University of Technology, Warsaw, Poland d Institute of Social Medicine, Department of Epidemiology, Medical University of Warsaw, Warsaw, Poland e Department of Traumatology & Orthopaedics, Central Hospital Military School of Medicine, Poland Accepted 20 February 2006
Abstract It is known that metallic elements of joint endoprostheses undergo elastic strain due to their mechanical function. This is one of the factors which may be responsible for the loosening of endoprostheses. Since mechanisms involved in it remain unclear, it seems valuable to verify if cells responsible for bone regeneration are affected by a strain of the implant. Our experiment examines the influence of elastic strain applied to Ti6Al4V samples on osteoblasts cultured on their surface in vitro. Human bone-derived cells are observed in contact with metallic plates. Titanium alloy was chosen as a support since it is one of the most commonly used materials for stems in joint endoprostheses. Cyclic elastic deformation of 0.1% was applied to the support once daily for 7 days. Two thousand cycles were applied each time. Samples which were not subject to strain served as control. After the observation period XTT assay was performed, alkaline phosphatase activity as well as osteocalcin concentration and nitric oxide secretion were determined and compared with the results obtained in the control group. It was found that the number of viable cells in the mechanically stimulated population was significantly higher than in control, while both alkaline phosphatase activity and osteocalcin concentration were significantly lower in the experimental group. Nitric oxide secretion was found in the culture which was subject to elastic strain, but not in the control. The possible clinical implication is that elastic strain of the metallic endoprostheses may influence osteoblasts which are in contact with the implant in vivo. r 2006 Elsevier Ltd. All rights reserved. Keywords: Osteoblast; Cell culture; Titanium; Mechanical stimulation; Aseptic loosening
1. Introduction Most biocompatibility studies of biomaterials are performed under static conditions. This is an important limitation of preclinical studies of candidate materials destined for load-bearing intrabone implants, since the evaluation of direct contact between biomaterial and Corresponding author. Department of Biophysics and Human Physiology, Medical University of Warsaw, Cha"ubin´skiego 5, 02-004 Warsaw, Poland. Tel./fax: +48 22 628 78 46. E-mail address: [email protected] (M. Lewandowska-Szumie").
0021-9290/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2006.02.012
bone tissue is strongly affected by mechanical circumstances at the bone-implant interface. The consequences of micromotion between endoprosthesis and host tissues have been experimentally observed in some studies (Aspenberg and Herbertsson, 1996; Aspenberg et al., 1992; Jasty et al., 1997; Søballe et al., 1992). However, not much attention has been paid to tissue reaction to elastic strain of the implant which undergoes mechanical stress. This, for example, applies to the contact of metallic stem of joint endoprosthesis with the host bone. The key problem with this type of reconstruction is the so-called aseptic loosening which leads to the necessity of reoperation. Although the presence of wear particles
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
near the implant is considered to be the main factor responsible for this complication (National Institute of Health (NIH) Consensus Statement, 1994;), the influence of mechanical stress of endoprosthesis on surrounding tissues as a factor involved in loosening is not out of question. Nevertheless, experimental verification is not easy—firstly, due to the difficulties in the proper simulation of mechanical conditions which exist in clinical application of endoprostheses and secondly, due to the high complexity of factors which accompany experimental implantation. Thanks to the great development of cell culture techniques, new experimental possibilities of investigating the relationship between mechanical factors and cellular functions in living organisms appear. The aim of the present study is to examine the influence of elastic strain applied to the metallic support on osteoblast-like cells cultured on its surface in vitro. Titanium alloy, one of the most commonly used materials for stem in joint endoprostheses, has been used in order to discuss the potential clinical implications of the results.
2. Materials and methods 2.1. Cells used in the experiment Human bone derived cells (HBDCs) were isolated from pieces of tissue removed at surgery, which would otherwise be discarded. Material from a spinous process (female, 15 years old) was used. Experiments were repeated with cells harvested from fibula (female, 69 years old). The procedure of HBDCs’ isolation was based on the protocols described by Gallagher et al. (1996) with modifications (Kudelska-Mazur et al., 2001). Briefly, soft tissue found on the extracted bone pieces was removed by scraping, bone was cut into small pieces and treated with collagenase (type XIS, SIGMA) for 24 h. The bone pieces were put into the culture medium containing DMEM (GIBCO BRL) enriched with 10% of heat-inactivated fetal calf serum (GIBCO BRL), Antibiotic–Antimycotic (GIBCO BRL), L-glutamine (GIBCO BRL) and 100 mM L-ascorbic acid 2phosphate (SIGMA), and left in humidified atmosphere (5% CO2, 95% humidity) in order to allow osteoblasts to migrate from bone explants onto the surface of culture flasks, 75 cm2 (NUNC). If the amount of material was sufficient, pieces of bone harvested from one patient were located in two different flasks, in order to obtain two populations of cells originating from the same donor. After the confluence has been reached, cells were detached by means of collagenase and trypsin digestion, and used in experiments. During the experiment cells were cultured in a medium described above,
supplemented (SIGMA).
with
555
1a,25-dihydroxycholecalciferol
2.2. Samples of biomaterial Cells were seeded on the surface of samples made of commercial Ti6Al4V alloy. Titanium alloy was chosen as a substrate since it is one of the most commonly used materials for stems in joint endoprostheses. Samples were prepared in the form of 0.1-mm thick, 14-mm long and 5-mm wide plates. They were autoclaved and used as a support for cells without any additional treatment. Particularly, they were not conditioned with any media, and cells were seeded on dry surfaces. Cells were seeded on the surface of titanium plates put on the bottom of 24-well culture dishes in the density of 47 000 cells/cm2. The next day substrate-strain procedure was started in the manner described below. 2.3. Method of stimulation—system for elastic strain of titanium plates Titanium plate samples with cells seeded on them were placed on two parallel TEFLONs supports at the distance of 10 mm, which were located at the bottom of a 24-well culture dish. The support was made of TEFLONs since it is known as a biocompatible material and thus it was expected not to influence the system when being put in contact with culture medium. As confirmed in the preliminary studies (unpublished data), even direct contact of cells with TEFLONs substrate does not affect any of cell features which were investigated in the experiment. Elastic strain of the titanium plates was generated by cyclic stressing of the centre of plates by TEFLONs stamps with conical end mounted to the cover of the culture dish (schematic illustration at Fig. 1—Sikorski et al., Polish Patent P 339 046). The cyclic motion of the stamps was performed by pneumatically operated system in which an elastic bag located on the upper side of a cover with TEFLONs stamps mounted on it, provoked bending of titanium samples. In a consequence cyclical elastic strain of titanium plates was achieved. The upper surface of the plate, on which the cells are seeded, is subjected to compressive strain,
Fig. 1. Schematic illustration of the experimental system.
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
556
which varied along the length of the sample, from ‘‘emax’’ under the stamp to zero over the sample support. The maximal amplitude of the strain is calculated on the basis of measurements of sample deflection under the stamp using the following expression (coming from the beam theory): emax ¼
6hf 100 ð%Þ, l2
where h is the thickness of the sample, f the deflection of the sample measured under the stamp, and l the distance between the sample supports. The strain amplitude of the cycle can be adjusted by changing the plate deflection, which depends on the length of the TEFLONs stamps. In our experiment the deflection was 0.25 mm, which makes the true strain in the surface layer of the plate under the stamp equal to 0.1%. Two thousand strain cycles per day were applied, similarly, as in the experiments of Kaspar et al. (2000) and Stanford et al. (1995), in which osteoblasts cultured on the surface of polymers were stretched. The procedure was carried on for 3 min (which makes the frequency of 11 Hz) and repeated once daily throughout 7 days. The process was carried out in a laminar flow in order to provide sterile conditions, thus, cells were staying up to 10 min outside the incubator each day. Cells cultured under identical conditions on the surface of titanium plates (including removing from incubator, placing titanium plates with cells on them on Teflon supports, and using the same volume of culture medium), but not subject to strain, served as a control. Both in the experimental and in the control group, the cells remained in the same culture medium during the whole experiment. The medium was not exchanged in order to collect osteocalcin which was determined at the end of the experiment. On the next day after the 7-daylong strain series has been completed, supernatants were harvested from both experimental and control cultures in order to measure osteocalcin and nitric oxide concentration, and viability test was performed. As recommended by Gallagher at al. (1996), and confirmed in our previous experiments, a significant increase of the osteocalcin concentration in response to 1a,25-dihydroxycholecalciferol, characteristic for the osteogenic phenotype, should be expected after that time. The next day alkaline phosphatase activity was determined in all cultures.
2.5. Viability assay In order to compare the number of living cells remaining in the culture after the applied handling, XTT assay was performed. This test is used in toxicology and is based on the ability of mitochondrial dehydrogenase enzymes in living cells to convert the XTT substrate (2.3-bis(2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenyloamino)carboxyl]-2H-tetrazolium hydroxide) into a water-soluble formazan product. The final product of the reaction is measured in the ELISA reader at 450 nm. Although the result of this test is not equal to the number of cells, there are reports which confirm that the production of formazans is proportional to the number of living cells (Breault et al., 1995; Doyle et al., 1996; Sadler et al., 1996), which makes this method useful in studies basing on the comparison of the number of living cells among groups. The number of cells was calculated from the calibration curve for XTT assay, which was obtained in a parallel experiment. After the XTT test, fresh medium was added to the culture in order to keep cells alive till the next day when the alkaline phosphatase activity was determined. 2.6. Alkaline phosphatase activity determination Alkaline phosphatase (ALP) activity was chosen to label the influence of the applied procedure on osteoblast in culture since it may indicate whether the function of osteoblasts is disturbed by the micromotion. Also, it is one of the most important promoters of osteogenesis in vivo, and thus—an important factor in predicting an in vivo situation on the base of the observations performed in cell culture. Since the product of the XTT assay is water soluble and the cells are not injured by the assay, the cells can be subsequently studied by other means. Thus, ALP activity was determined on the next day. This combination of XTT and ALP activity assay is routinely used in our laboratory, since it allows us to measure ALP activity exactly in the same population in which cell viability is measured by XTT assay. As found in numerous preliminary studies performed in order to establish the procedure, the ALP test is only slightly and insignificantly affected by a previous XTT assay when the ALP test is performed on the next day (data not shown). ALP activity was measured by means of Alkaline Phosphatase SIGMA Diagnostic Kit. 2.7. Nitric oxide measurement
2.4. Osteocalcin concentration measurement Osteocalcin concentration was determined in cell culture supernatants with N-MIDTM Osteocalcin One Step ELISA test (Osteometer Biotech A/s). The results were calibrated and expressed in ng/ml.
Nitric oxide (NO) presence was detected in supernatants harvested from both mechanically stimulated and control culture. Griess reagent was used (Lowik et al., 1994; Mac Pherson et al., 1999), which reacts with nitric oxide giving the stable product of characteristic
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
colour (sodium nitrite) which is then measured in the ELISA reader at 540 nm. 2.8. Statistical analysis The experiments were carried out 3 times, i.e. the first, preliminary, experiment showed the relations, which were then confirmed in the next two experiments. The results obtained in those two experiments were taken into calculations. Cells from one donor were used for the first two experiments (the preliminary and the first one of those taken into calculation) and cells from another donor were used in the last one, to make sure that the found effects are not accidental or connected with the cells from a particular donor. Seven samples in the experimental group (cells cultured on the support subjected to strain) and seven samples from the control were investigated in each of the experiments taken to calculation (which makes 14 measurements in each group). Each measurement was done on three aliquots from each sample. Results were analyzed using Two-way ANOVA in which cell viability, alkaline phosphatase activity, osteocalcin level, and NO secretion were considered as dependent variables. Two factors were taken into account: firstly, strain effect (a group subject to strain and non-strained control), and secondly, donor effect (cells were isolated from tissue harvested from two patients). Such conservative approach was applied in order to estimate the influence of the source of the cells (donor) on the results.
557
of the substrate (136% of control). On the contrary, both ALP activity and osteocalcin concentration are significantly higher ðpo0:001Þ in control i.e. in culture which was carried out on unstrained titanium plates. ALP activity value in the population subject to strain was at the level of 56% of control, and osteocalcin concentration, at the level of 63% of control. The number of cells, calculated from the calibration curve for the XTT assay, as well as the osteocalcin level and alkaline phosphatase activity in the population exposed to the elastic strain of titanium support are shown together as a percentage of untreated control in Fig. 2. Nitric oxide was practically not detected in control, while in the mechanically stimulated group some amount of NO was found and determined as the mean concentration of NO 2 in the well, which was about 2.3 mM. All found relations between values found in the experimental group vs. the control, were confirmed in all experiments. The influence of the source of the cells (donor) on the results was found to be insignificant.
4. Discussion and conclusions The results of viability assay indicate that the number of cells cultured on the titanium that underwent elastic strain is enhanced. The effect is even stronger than shown by quantitative data if we take into account the 160% 140%
133%
Results of quantitative analysis are gathered in Table 1. All found relations were confirmed in three experiments. For calculation presented in the table results from two experiments (14 samples which were subject to strain and 14 control-not subject to strain) were taken. Values of ALP activity, osteocalcin and NO concentrations were normalised to the number of cells calculated from the calibration curve for the XTT assay. Absorbance, read in XTT assay, is significantly ðpo0:0001Þ higher in culture which was subject to strain
% of control
120%
3. Results
100% 80%
63%
56%
60% 40% 20% 0%
cell number
ALP
osteocalcin
Fig. 2. Number of cells calculated from the calibration curve for the XTT assay, alkaline phosphatase activity and osteocalcin concentration—expressed as percentage of control (based on values presented in Table 1).
Table 1 Results of quantitative analysis: absorbance obtained in XTT assay, number of cells calculated from the calibration curve for the XTT assay, values of ALP activity, osteocalcin and NO concentrations normalized to the number of cells i.e. divided by (number of cells 104); shown are mean values (MV) and standard errors (SE) XTT assay (absorbance)
Number of cells
ALP/104 cells (U/L)
Osteocalcin/104 cells (mg/ml)
4 NO 2 /10 cells (mM)
MV
SE
MV
SE
MV
SE
MV
SE
MV
SE
0.014 0.007
56845 42720
1040 500
8.763 15.721
1.598 2.281
39.900 63.820
4.760 6.080
0.402 0.050
0.038 0.028
Subject to strain 0.725 Control 0.534
ARTICLE IN PRESS 558
M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
fact that some cells in the experimental group must have been damaged due to the contact with a TEFLONs stem, which performs the strain of the substrate. 11% of the whole sample surface was affected by this. It was estimated after the whole experiment by measuring the ‘‘empty’’, area (i.e. devoid of cells) visible in the central part of the samples which underwent elastic strain. Some cells from this area must have been killed, and others were probably dislocated. It was not possible to take this factor into account quantitatively. Anyway, if this drawback was avoided, the effect of the enhanced cell number after application of the strain to the support, would probably be even more evident, since only viable cells were taken into calculation. The enhanced number of cells in the group, in which the cell support underwent elastic strain, may result from cell proliferation or preventing cell death (by necrosis or apoptosis). As can be judged on the basis of preliminary observation of cells in contact with Ti6Al4 V plates under static conditions, the system is very well tolerated by cells, and during the 8-day long culture no signs of apoptosis were observed (unpublished data). So, it can be postulated that the applied procedure promotes cell proliferation. This finding is in good correlation with the results obtained in experiments of Stanford et al. (1995) and Kaspar et al. (2000) in which the conditions that exist during physiological strain of long bones were imitated in osteoblast culture on silicone substrate subject to strain. Also, similarly to the papers cited above, in our experiment, the enhanced proliferation of cells is accompanied by a significantly lower expression of phenotype features in the experimental group as compared to control. Both osteocalcin level and alkaline phosphatase activity are slightly, but statistically significantly, lower when elastic strain of metallic support is applied. Enhanced proliferation of HBDCs due to the applied strain was observed by Holbein et al. (1995), who also noticed that the addition of serum obtained from patients who used Ilizarov device resulted in the increase of proliferation in vitro of SaOS–line cells. A significant decrease in alkaline phosphatase activity and increase in cell number, as reflected by DNA content, was observed in a culture of primary rat osteoblast-like cells in response to intermittent stretching (Winter et al., 2003). Similarities of our results to those obtained in experiments of others—mainly designed to explain cellular mechanisms responsible for bone remodelling under loading—confirm the reliability of the proposed experimental system, which is to imitate the situation which exists at the interface of bone and load-bearing metallic implant. In addition to the enhanced proliferation and diminished osteocalcin release as well as alkaline phosphatase activity, nitric oxide secretion—in response
to the elastic strain of titanium substrate—was found in our experiment. Pitsillides et al. (1995) reported an analogous effect in a culture of chicken embryonic bone cells exposed to stretching of polymeric support. On the contrary, Smalt et al. (1997) did not observe such effect on osteoblast culture on stretched polystyrene. Although in the culture of the same cells exposed to fluid shear, NO concentration in the culture medium was found to be increased. In experiments by S.Kapur, addition of L-NAME—the inhibitor of all types of NO synthases—blocked the shear-stress-induced osteoblast proliferation and differentiation (Kapur et al., 2003). The potential role of NO in bone remodelling is a new factor of interest. Its role as a mediator in the osteoblast–osteoclast communication is postulated (Lowik et al., 1994). Although the mechanisms by which nitric oxide influences bone remodelling still remain unclear, many factors support its role in bone physiology and pathology. Nitric oxide secretion in HBDC in response to elastic strain of implantable metallic support found in our experiments indicates that it also may play some role in bone remodelling in contact with load-bearing intrabone metallic implants. Similarly, diminishing markers of osteoblast phenotype, i.e., alkaline phosphatase activity and production of osteocalcin, found in our observation, can be interpreted from the point of view of a clinical situation. HBDCs are a population of osteogenic cells at various stages of differentiation. A significant increase in both alkaline phosphatase activity and osteocalcin production in response to 1,25-dihydroxy-vitamin D3 confirms their potency to differentiation in vitro. The lower level of ALP activity and osteocalcin concentration in culture exposed to elastic strain of support indicates that such mechanical handling may slow down the differentiation of HBDCs. A higher proliferation of less differentiated cells in a mechanically stimulated system seems comprehensible if we take into account that terminally differentiated osteoblasts do not multiply. Such interpretation of results leads to potential clinical implications. Elastic strain of the metallic part of endoprosthesis may influence the osteoblasts which are in the contact with the implant in vivo and should be taken into account as one of the potential factors responsible for loosening the endoprosthesis. This is the main conclusion from the reported findings. It remains an open question to which extent experiments performed in cell culture are a good basis for understanding phenomena which take place in vivo. There are many variables which influence the experimental system: among them, strain level, the number of cycles, frequency, duration of observation, and others. In our experiment, in which special attention was paid to the similarity of the applied conditions to an in vivo situation, the selected frequency of cyclic loading needs some comment.
ARTICLE IN PRESS M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
The applied frequency, equal to 11 Hz (2000 cycles for 3 min), is relatively high. The frequency of walking during normal physical activity of a human is 1 Hz (Bergmann et al., 2001). However, frequency components above 20 Hz were found by Stergiou et al. (2002) in frequency domain analysis of the ground reaction forces during walking and there were observed as high as 75 Hz frequency components associated with impulsive loading at heel strike during barefoot walking (Harris et al., 1996). Besides, there is plenty in vitro evidence for a cellular component in the response of bone cells to high frequency stimuli (Kaspar et al., 2002; Kasra et al., 2003; Tanaka et al., 2003), supporting the choice of a frequency higher than 1 Hz. As shown by Kaspar et al. (2002) HBDCs’ proliferative response to cyclic strain at frequencies ranging from 0.1 to 30 Hz, did not significantly depend on the frequency when the cycle number was left constant. The cycle number in our experiment was similar to those applied by Kaspar et al. (2000) and Stanford et al. (1995). Altogether, because using the frequency of 11 Hz is time efficient (the cells spent only 3 min outside the incubator) such frequency was used in our experiment. The weak point of our system is the fact that some amounts of cells undergo destruction as a consequence of the direct contact of the surface covered by cells with the TEFLONs stamp. The influence of dead cells on any factor measured in a culture cannot be completely excluded. On the other hand, to the best of our knowledge there is only one other system which allows to observe the influence of stretching nonflexible support on cells in vitro (Di Palma et al., 2003). This system is based on a Dynacells device, especially designed to apply cyclic strains on rigid biomaterials. It does not require direct pushing of the support, but unfortunately it requires samples of a big area (discs of 90 mm diameter). The area of a singular sample is little less than a 75 cm2 culture flask. In consequence, the application of this method is practically limited to the observations of cell lines because it is difficult to obtain a sufficient number of isolated from bone explants osteoblasts, to cover the investigated surface in a reliable number of repetitions. This limitation is severe, especially that the results obtained in a culture of cell lines may be completely different from the results in a culture of HBDCs. To sum up, there still is a need to work on an in vitro system which would entirely correspond to the situation of osteoblasts in direct contact with load-bearing implants which undergo elastic strain. The model proposed in the presented paper is one of possible responses for such requirement, and the basic finding— that osteoblasts respond to elastic strain of titanium in direct contact through an enhanced proliferation accompanied by a diminished differentiation—can be considered in the assessment of the consequences of elastic strain of metallic endoprosthesis in vivo.
559
Acknowledgments This work was supported by the State Committee of Scientific Research (Grant no. 4. PO5A.065.11). References Aspenberg, P., Herbertsson, P., 1996. Periprosthetic bone resorption. Particles versus movement. Journal of Bone and Joint Surgery— British Volume 78, 641–646. Aspenberg, P., Goodman, S., Toksvig-Larsen, S., Ryd, L., Albrektsson, T., 1992. Intermittent micromotion inhibits bone ingrowth, Titanium implants in rabbits. Acta Orthopaedica Scandinavica 63, 141–145. Bergmann, G., Deuretzbacher, G., Heller, M., Graichen, F., Rohlmann, A., Strauss, J., Duda, G.N., 2001. Hip contact forces andgait patterns from routine activities. Journal of Biomechanics 34, 859–871. Breault, L.G., Schuster, G.S., Billman, M.A., Hanson, B.S., Kudryk, V.L., Pashley, D.H., Runner, R.R., McPherson, J.C., 1995. The effects of intracanal medicaments, fillers, and sealers on the attachment of human gingival fibroblasts to an exposed dentin surface free of a smear layer. Journal of Periodontology 66, 545–551. Di Palma, F., Douet, M., Boachon, C., Guignandon, A., Peyroche, S., Forest, B., Alexandre, C., Chamson, A., Rattner, A., 2003. Physiological strains induce differentiation in human osteoblasts cultured on orthopaedic biomaterial. Biomaterials 24, 3139–3151. Doyle, A., Griffiths, J.B., Newell, D.G. (Eds.), 1996. Core Techniques. Cell & Tissue Culture: Laboratory Procedures. Wiley, West Sussex, England, pp. 4B:6.1–4B:6.2. Gallagher, J.A., Gundle, R., Beresford, J.N., 1996. Isolation and culture of bone-forming cells (osteoblasts) from human bone. In: Jones, G.E. (Ed.), Methods in Molecular Medicine: Human Cell Culture Protocols. Humana Press Inc., Totowa, NJ, pp. 233–262. Harris, G.F., Acharya, K.R., Bachschmidt, R.A., 1996. Investigation of spectral content from discrete plantar areas during adult gait: an expansion of rehabilitation technology. IEEE Transactions on Rehabilitation Engineering 4, 360–374. Holbein, O., Neidlinger-Wilke, C., Suger, G., Kinzl, L., Claes, L., 1995. Ilizarov callus distraction produces systemic bone cell mitogens. Journal of Orthopaedic Research 13, 629–639. Jasty, M., Bragdon, C., Burke, D., O’Connor, D., Lowenstein, J., Harris, W.H., 1997. In vivo skeletal responses to porous-surfaced implants subjected to small induced motions. Journal of Bone and Joint Surgery—American Volume 79, 707–714. Kapur, S., Baylink, D.J., Lau, K.-H.W., 2003. Fluid flow shear stress stimulates human osteoblast proliferation and differentiation through multiple interacting and competing signal transduction pathways. Bone 32, 241–251. Kaspar, D., Seidl, W., Neidlinger-Wilke, C., Ignatius, A., Claes, L., 2000. Dynamic cell stretching increases human osteoblast proliferation and CICP synthesis but decreases osteocalcin synthesis and alkaline phosphatase activity. Journal of Biomechanics 33, 45–51. Kaspar, D., Seidl, W., Neidlinger-Wilke, C., Beck, A., Claes, L., Ignatius, A., 2002. Proliferation of human-derived osteoblast-like cells depends on the cycle number and frequency of uniaxial strain. Journal of Biomechanics 35, 873–880. Kasra, M., Goel, V., Martin, J., Wang, S.T., Choi, W., Buckwalter, J., 2003. Effect of dynamic hydrostatic pressure on rabbit intervertebral disc cells. Journal of Orthopaedic Research 21, 597–603. Kudelska-Mazur, D., Lewandowska-Szumie", M., Benke, G., Kowalski, M., Komender, J., 2001. Isolation of human primary osteoblsts for biomaterials investigation in vitro. Engineering of Biomaterials 15&16, 51–58.
ARTICLE IN PRESS 560
M. Lewandowska-Szumie! et al. / Journal of Biomechanics 40 (2007) 554–560
Lowik, C.W., Nibbering, P.H., van de Ruit, M., Papapoulos, S.E., 1994. Inducible Production of Nitric Oxide in Osteoblast-like Cells and in Fetal Mouse Bone Resorption. Journal of Clinical Investigation 93, 1465–1472. Mac Pherson, H., Noble, B.S., Ralston, S.H., 1999. Expression and functional role of nitric oxide synthase isoforms in human osteoblast-like cells. Bone 24, 179–185. National Institute of Health Consensus Statement, 1994. Total Hip Replacement 12(5), 1–31; published also in The Journal of American Medical Associations 1995, 273, 1950–1956. Pitsillides, A.A., Rawlinson, S.C., Suswillo, R.F., Bourrin, S., Zaman, G., Lanyon, L.E., 1995. Mechanical strain-induced NO production by bone cells: a possible role in adaptive bone (re)modeling? Faseb Journal 9, 1614–1622. Sadler, G.P., Jones, D.L., Woodhead, S., Horgan, K., Wheeler, M.H., 1996. Effect of growth factors on growth of bovine parathyroid cells in serum-free medium. World Journal of Surgery 20, 822–829. Sikorski, K., Lewandowska-Szumie", M., Szummer, A., Polish Patent P 339 046 ‘‘Urza˛ dzenie do badania z˙ ywych komo´rek w kontakcie z biomateria"ami in vitro’’ (Apparatus for in vitro investigations of cells in contact with biomaterials).
Smalt, R., Mitchell, F.T., Howard, R.L., Chambers, T.J., 1997. Induction of NO and prostaglandin E2 in osteoblasts by wall-shear stress but not mechanical strain. American Journal of Physiology 273, 751–758. Søballe, K., Hansen, E.S., Brockstedt-Rasmussen, H., Jørgensen, P.H., Bu¨nger, C., 1992. Tissue ingrowth into titanium and hydroxyapatite-coated implants during stable and unstable conditions. Journal of Orthopaedic Research 10, 285–299. Stanford, C.M., Morcuende, J.A., Brand, R.A., 1995. Proliferative and phenotypic responses of bone-like cells to mechanical deformation. Journal of Orthopaedic Research 13, 664–670. Stergiou, N., Giakas, G., Byrne, J.B., Pomeroy, V., 2002. Frequency domain characteristics of ground reaction forces during walking of young and elderly females. Clinical Biomechanics (Bristol, Avon) 17, 615–617. Tanaka, S.M., Li, J., Duncan, R.L., Yokota, H., Burr, D.B., Turner, C.H., 2003. Effects of broad frequency vibration on cultured osteoblasts. Journal of Biomechanics 36, 73–80. Winter, L.C., Walboomers, X.F., Bumgardner, J.D., Jansen, J.A., 2003. Intermittent versus continuous stretching effects on osteoblast-like cells in vitro. Journal of Biomedical Materials ResearchApplied Biomaterials 67, 1269–1275.