¹ Academic Centre for Dentistry Amsterdam (ACTA), Dept. of Oral Cell Biology, Amsterdam, The Netherlands.
² Flemish Institute for Technological Research (VITO), Environmental Division, Mol, Belgium.

ABSTRACT

Biocompatibility is an important factor in the development of orthopaedic implants as well as in the development of new tissue culture devices.

Polysulphone has been used for orthopaedic implants because of its mechanical properties, ease of sterilization, its molding capacity and its biocompatibility. Therefore polysulphone has been chosen as the prime material for the construction of tissue culture devices to be used for cultivation of osteogenic cells (pre-osteoblast-like MN7 cells and primary bone marrow fragments) as well as complete fetal long bone explants under space flight conditions.

Whereas polysulphone did not interfere with the proliferation in early stages of bone forming cells, we show that leacheble factors within the polysulphone polymer prevented the final steps of matrix formation as measured by collagen synthesis and matrix mineralization. These data argue against polysulphone is a material for orthopaedic implants.

Key words: polysulphone, bone, mineralization, orthopaedic implants, microgravity, tissue culture device.

This paper is submitted for publication.

Recently carbon fiber reinforced polysulphone has attracted some attention as a new implant material in tumor surgery of the spine,⁽¹⁾ internal fracture fixation of long bones⁽²⁾ or after auto-alloplastic tooth replantation.⁽³⁾ Apart from its improved mechanical properties as compared with metallic alloys, also the effects on a biological system seems to be favorable.⁽⁴⁾ Polysulphone is also promising from an engineering point of view, since it is translucent, very well machinable and can be sterilized by heat, gas, or X-rays.⁽⁵⁾ For these reasons polysulphone (PSU) was chosen as the prime material for constructing tissue culture devices. These devices are to be used in experiments to study the effect of microgravity on bone development and metabolism while in culture.

It has been shown that bone metabolism changes during space flight.^(6,7) It is not clear from these in vivo experiments, however, whether these changes are due to changes in systemic factors like hormone levels or body fluid shifts,⁽⁸⁾ or whether they are the result of direct effects of near weightlessness on bone cell metabolism. By using these automated devices for in vitro studies we might address this problem on a more basic level.

Intramembranous ossification is characterized by a number of successive steps. Initially quiescent bone precursor cells are recruited for proliferation (proliferation phase), whereafter they gradually acquire bone phenotypic markers (maturation phase). In order of their expression these main markers are: Collagen Type I, alkaline phosphatase, osteopontin, and osteocalcin.⁽⁹⁾ The final result of this differentiation process is the formation of nucleation sites and finally a fully mineralized bone matrix (calcification phase). In the present experiments studying the effect of PSU on osteogenic differentiation, cell number, collagen synthesis and calcification were chosen as markers for the proliferation, maturation and calcification phase, respectively.

To gain better insight in the effect of PSU on the different osteogenic differentiation phases, three distinct but complementary bone differentiation models have been tested in contact with PSU. In the first model osteoblast precursor cells in primary murine bone marrow fragments have been triggered to form a mineralized matrix de novo10⁾ The second model features a c-fos transformed cell line with pre-osteoblast like characteristics (MN7), in order to study its development towards a more mature osteoblastic phenotype.⁽¹¹⁾ In the last model, complete explants, 16 day old fetal mouse cartilaginous long bone rudiments (metatarsals), have been used to study endochondral and intramembranous calcification in vitro. Whereas in the former two models the recruitment and maturation of pre-osteoblasts towards mature osteoblasts could be studied, the latter model focusses on the final differentiation step in the formation of bone, i.e. cartilage and bone calcification.

Besides the variation in developmental stages, these osteogenic models have also been chosen to be used for future microgravity experiments, since it has already been demonstrated that osteoblast like cells ⁽¹²⁾ as well as metatarsal long bones⁽¹³⁾ are susceptible to changes in gravitational forces.

The results have demonstrated that parameters indicative for cell proliferation have not been affected by PSU. However, the maturation and mineralization steps have been shown to be significantly retarded in the presence of PSU in all tests performed. We discuss the possible implications of these results on the potential use of PSU as an implant material.

The establishment of bone marrow organ cultures has been described previously.⁽¹⁰⁾ Briefly, the femurs of three-month-old Balb/c inbred mice (VITO animal facility, Belgium) were dissected, cleaned and the ends separated from the midshaft. The marrow was carefully flushed from the midshaft with a 22-gauge needle connected to a syringe. The marrow was aseptically collected as marrow fragments without disrupting its original three dimensional structure. The marrow fragments were allowed to adhere to aseptic collagen supports (CollaTech Inc.) for 30 minutes, whereafter 1 ml of BGJb medium (Gibco) was added, supplemented with 10% fetal calf serum, 1% L-glutamine and 1% gentamicin (all derived from Gibco), 10 mM sodium--glycerophosphate and 50 g/ml L-ascorbic acid (Sigma). The cultures were incubated in a 5% CO₂ in air, 37°C, 100% relative humidity incubator. The tissue culture medium was changed every 3 days. Control experiments were performed in standard laboratory 24 wells plates.

The mineralization rate of bone marrow cultures was followed in time using ⁸⁵Sr (5 Ci/ml)(NEN) as a tracer for calcium.^{(10) 85}Sr gamma-radioactivity in thoroughly rinsed marrow fragments, was measured using a NaI(Tl) detector (Packard 5320).

Establishment of the MN7 preosteoblast-like cell line has been published elsewhere.⁽¹¹⁾ Adherent cell cultures were established by plating 10³ cells/well in 96 well plates (Gibco). Cell number was scored on day 3 with a particle counter (Coulter model ZF, Coulter Electronics). Medium and culture conditions were identical to those described for the bone marrow fragments.

The middle 3 cartilaginous metatarsal long bones in each foot of 16 day embryonic (ED16) Swiss mice were used. Rudiments (typical dimensions 1.00.3 mm) were aseptically harvested and cultured in bicarbonate buffered MEM without nucleosides supplemented with 50 mg/l gentamicin and 0.5% v/v fungizone (Gibco), 50 mg/l ascorbic acid, 300 mg/l glutamine (Merck), 0.2% w/v BSA Factor V and 1 mM Na--glycerophosphate (Sigma). Controls were always cultured in standard 24 wells tissue culture plates (300 l medium / well) (Greiner) and placed in a 5% CO₂ in air, 37C, 100% relative humidity incubator. ED16 metatarsal long bones are cartilaginous non-mineralized rudiments at dissection but will calcify while in culture. Using an inverted microscope, growth and mineralization were assessed as total length of the rudiment and the calcified zone (diaphysis), respectively. Metatarsals were contralateral paired, one bone receiving control and the other experimental treatment.

Prototype hand operated devices based on a design by Ubbels and CCM^* were constructed of PSU (Ensinger, Technische Kunststoffe, Nufringen, Germany).⁽¹⁴⁾ Each device held two culture compartments. The devices also contained reservoirs for fluid exchange in the culture compartment (3 per culture compartment), each of them containing 1.0 ml tissue culture medium or fixative. This allowed for two successive replenishment of the tissue culture medium and a final fixation of the biological samples. Each of the two culture compartments contained either one bone marrow fragment, or 4 metatarsal long bones in 1.0 ml of culture medium. The surface area contact of PSU with culture medium in the culture compartment was 550 mm². The culture compartment was covered with a gas permeable polyethylene foil to allow ample gas exchange (5% CO₂ in air) during cultivation. (Fig. 4.1)

^*CCM, Center for Construction and Mechanotronics, Nuenen, The Netherlands.

Collagen synthesis was measured by the incorporation of 3H-proline (Amersham, UK) into collagenase digestible protein (CDP).⁽¹⁵⁾ The collagenase (Worthington, New Jersey) was free of non-specific protease activity. The cell cultures were exposed to ³H-proline (1 Ci/ml) for the last 18 hours of culture and then washed three times with PBS in 35 mm petri dishes. Collagenase was added (0.1 g/ml in PBS) for 3 hours and the ³H was measured in a liquid scintillation counter.

Proliferation of bone marrow and MN7 cells was measured by the incorporation of tritiated thymidine. The thymidine was added to the culture medium 16 hrs before finishing the experiment. After culture, the cells were washed and treated with trypsin, harvested and collected onto a paper support (Glass fiber filter, Beckman). Subsequently, the filters were washed, punched out and the remaining radioactivity was counted in a liquid scintillation analyzer (Tri-Carb 1600 CA, Packard).

To study the cytotoxicity of PSU two methods were used:

This method for studying the cytotoxicity of plastics has been described earlier.⁽¹⁶⁾ Briefly, for this conditioned medium (CM-1) the PSU resin was fragmented with a lathe to curled shavings. Eight grams of these shaving were autoclaved in 40 ml of Milli-Q water. The extraction water was 1:1 mixed with 2 times concentrated complete culture medium. This CM-1 was used for culturing bone marrow fragments as well as long bone rudiments.

To study whether leacheble factors deriving from PSU affects metatarsal long bone development, a second conditioned medium, CM-2, was prepared by incubating plates of PSU (total surface area 330 mm²) in a standard 6 wells tissue culture plate (polystyrene, Greiner) in 2 ml complete culture medium (MEM + all additives) for 3 days at 37C in a 5% CO₂ incubator. Control media were without PSU plates. This CM-2 was subsequently used to culture ED16 metatarsals for 6 days.

For CM-1 and CM-2 cultures, both proliferation and differentiation characteristics were measured.

Some cell and tissue samples were fixed with formaldehyde and embedded in historesin for histological evaluation. 3 m sections were stained with either toluidine blue (0.2% + 0.2% borax, 90 sec.) for general histology and cell integrity, or alizarin red-S (1%, pH=5.5, 60 sec., room temp.) for mineral formation.

Changes in anion and cation compositions were measured in culture medium. Complete BGJb tissue culture medium, including all additives, was incubated in open PSU trays or control petri dishes for 3 days in a 37C, 5% CO₂ incubator. The ionic composition of the medium was determined with a plasma emission spectrometer (Jarrell-ASM, Atomcomp, model 750).

Statistics were calculated using the two tailed Student t-test, for paired or unpaired values when applicable. All data are expressed as means ± SEM.

Figure 4.1 shows the modified version of a completely automated tissue culture device, which was already successfully used on board the Space Shuttle and MASER sounding rocket missions.^(14,17) The modified device has been developed for studying osteogenic cells in near weightlessness. Hand-operated prototypes of these cell culture devices have been used for biocompatibility tests. In these prototypes the proliferation of cells in bone marrow fragments and 16 day old metatarsal long bones as measured by ³H-thymidine incorporation and total bone length, respectively, was suppressed as compared to standard laboratory tissue culture plastics (Figs. 4.2a and 4.2b). In the same experiments the final differentiation steps, i.e. collagen synthesis and calcification were greatly reduced or totally absent (Figs. 4.2c and 4.2d).

To rule out 'device effects' (i.e. geometrical and physicochemical interferences from the integrated device on the osteogenic cultures), experiments were performed in open culture trays made from PSU. The use of these open PSU trays resulted in large variation among the samples (Fig. 4.3). There is, however, a significant reduction of strontium uptake in bone marrow fragments cultured in PSU compared to controls. Similar results were obtained when the PSU CM-1 was used to cultivate osteogenic cells and tissues. A significant decrease has been observed in the acquisition of collagen and in mineralization, in the bone marrow fragments and metatarsal long bones, respectively, when cultured in CM-1 medium (Fig. 4.5c, 4.5d). In all three different bone cell differentiation models that we used, no significant effect of PSU has been seen on the initial proliferative stages of the cultures (Figs. 4.4, 4.5a and 5.5b).

Also, from histological sections (Figs. 4.6a and 4.6b), it is clear that when metatarsal long bone rudiments are cultured for 6 days in either control or PSU conditioned medium no cell death or necrosis can be detected. When stained for calcium by alizarin red-S, however, mineralization of the matrix is only present in control cultures (Fig. 4.6c).

After incubation of complete BGJb tissue culture medium in either multiwell plates or open PSU trays, no striking changes were found of either cation or anion concentrations as a result of adsorption or excretion of these two polymers measured by plasma emission spectrometry (Table 4.I). The control medium, incubated in multiwell plates, even showed a decreased calcium and phosphate content compared to untreated or PSU incubated medium. The PSU extract showed a significant increase in Ca and P ion content compared to fresh medium. Measurements have been partly biased by the large amounts of sodium. Also no increase in Cd, Co, Cu, Fe, Pb, Se, Zn or Mn contents were detected after incubation.

With the ageing population in the Western world and the expansion of human activities towards space, questions related to bone metabolism and its disorders have come to the foreground. Whereas brittleness of the bone caused by immobility, old age or metabolic disease requires the development of more and better bone implant materials, microgravity induced osteoporosis has to be understood before actual long term manned space flight can be planned. Both issues, at first sight unrelated, merge because they both rely on thorough biocompatibility testing. In both bone implant surgery as well as in the development of a tissue culture device, only those materials can be used which are biocompatible and which allow maximal differentiation of skeletal tissues and cells. The present experiments have been conducted in order to develop an automated cell culture unit to study microgravity induced changes in bone metabolism. A lesson can be learned as far as the use of test systems required for biocompatibility testing for implant materials in vitro is concerned.

Although, it is well known that different cell types and various stages of cell development react differently to drugs or bioeffector molecules, some biocompatibility experiments still rely on relatively undifferentiated cells such as fibroblasts.^(18,19,20) These in vitro tests do not always reflect the cell population at the in vivo application site of the biomaterials tested. For this in vitro study we used representatives of various stages in bone development therefore approaching the in vivo situation for orthopaedic implants as close as possible.

In the osteogenic models used, we have found no PSU induced retardation in proliferation of MN7 cells (Fig. 4.4), bone marrow fragments (Fig. 4.5a) or general growth (= total length) of ED16 metatarsal long bones (Fig. 4.5b). These results have been confirmed via histological examinations of bone marrow fragments (data not shown) and whole explants (Fig. 4.6a, 4.6b). In either case only minimal or no cell death has been observed (Figs. 4.6a and 4.6b). These observations are consistent with previous experiments studying the proliferation of fibroblasts in PSU conditioned medium.⁽⁴⁾ The same authors also showed that cell death due to PSU, as accessed by lactate dehydrogenase (LDH) concentration in the medium, did not occur.

In contrast with cell proliferation during the initial stages of cultivation, the acquisition of more mature phenotypic characteristics of osteogenesis (collagen synthesis and mineralization) were clearly shown to be inhibited by PSU (Figs. 4.2c, 4.2d, 4.3, 4.5c and 4.5d). Indirect effects, related to PSU, such as inhibition of collagen synthesis and calcification by depletion of the tissue culture medium of proteins and/or minerals required for mineralization have to be ruled out. Firstly Mandenius et al.⁽²¹⁾ showed that only minute amounts of proteins adsorb onto PSU surfaces under physiological conditions. Secondly our analysis of cat- and anions composition of tissue culture medium has shown no dramatic PSU induced changes during a three day incubation (Table 4.I). Moreover, since not only PSU devices, but also PSU extracts (CM-1 and CM-2) interfere with the osteogenic differentiation, leacheble components are the only likeable candidates for interference (Figs. 4.5c and 4.5d). Since these components have been shown not to effect cell proliferation they have to be considered non-cytotoxic. The inhibition of osteogenesis partly derives from decreased collagen synthesis and/or post excretion modulation of molecules or molecule complexes in the extracellular matrix. The mechanism by which PSU interferes with the differentiation of skeletal cells and their extracellular matrix has to be further investigated.

The experiments being discussed were performed with the aim to develop a biocompatible tissue culture device to be used for space flight experiments on bone development and metabolism, and were entirely performed in vitro. These data can contribute, however, to our understanding of in vivo effects seen with PSU containing implants. Although PSU was successfully used as bone implant material,⁽²²⁾ a thin layer of non-mineralized fibrous tissue at the bone-PSU implant interface has been reported by Knowles et al.,⁽²³⁾ which was not seen in hydroxyapatite / polyhydroxybutyrate implants. Leacheble factors from PSU could explain this phenomenon. These leacheble factors may have a profound impact on the final choice of PSU as implant material. Although, in vitro methods with fibroblasts may be suitable first indicators for a general biocompatibility testing of new materials, future studies should aim at mimicking the in vivo situation more closely. More differentiated skeletal cells and tissues and relevant parameters should be introduced, such as described by Puleo et.al or Angle et al.,^(24,25) when orthopaedic implants are considered.

From these biocompatibility tests it is also clear that the application of similar hardware already successfully used for previous space flight experiments with Xenopus amphibian eggs⁽¹⁴⁾ and human A431 epidermoid carcinoma cells⁽¹⁷⁾ does not necessarily imply favorable culture conditions for other biological systems. This PSU material can not be used as prime material for devices to study the various differentiation steps in skeletal tissues under space flight conditions.

With the use of bone marrow fragments, osteogenic cell lines or developing fetal mouse metatarsal long bones, the in vivo aspects of skeletal growth, maturation and calcification is simulated more closely. These three osteogenic models used in this study may be considered to be used for future biocompatibility tests for orthopaedic implants.

In order to allow autonomous cultivation of mammalian cells in space, a fully automated tissue culture device has been developed which has allowed successive refreshment of the spent culture medium and final fixation of cells. We have shown that when these devices are made of polysulphone, leacheble components inhibit the final steps of osteogenesis. In the context of the interest that exists for PSU as an implant material, the present data may urge for some caution.

1. C. Burri, L. Claes and O. Wörsdörfer "Osteosynthese an der Wirbelsäule mit individuell gearbeiteter Platte aus kohlenstoffaserverstärktem Polysulfon," Unfallchirurg, 89, 528-532 (1989).

2. L. Claes, "Kohlenstoffaserverstärktes Polysulfon: ein neuer Implantatwerkstoff," Biomed. Technik, 34(12), 315-319 (1989).

3. W. Foerster, W. Huttner and H. Kirschner, "Kohlenstoffaserverstärktes Polysulfon als Implantatmaterial; Werkstoffliche Eigenschaften und biologische Untersuchung," Dtsch Z Mund Kiefer GesichtsChir, 8, 437-440 (1984).

4. L.M. Wenz, K. Merritt, S.A. Brown, A. Moet and A.D. Steffee, "In vitro biocompatibility of polyetheretherketone and polysulfone composites," J. Biomed. Mater. Res., 24, 207-215 (1990).

5. A.K. van der Vegt, in Polymeres: from chain to resin, Delfsche Uitgevers Maatschappij, Delft, The Netherlands (in Dutch) 1991.

6. G.P. Vose, "Review of roentgenographic bone demineralization studies of the gemini space flights," Am. J. Roentgenol., Rad. Therapy & Nuclear Med., 121, 1-4 (1974).

7. E.R. Morey and D.J. Baylink, "Inhibition of bone formation during space flight," Science, 210, 1138-1141 (1978).

8. C.S. Leach and P.C. Rambaut, "Biochemical responses of skylab crewmen: an overview," in Biomed results from skylab, R.S. Johnston, L.F. Dietlein (ed.), NASA SP-377, NASA, Washington, D.C. U.S.A., 1977, pp. 204-216..

9. G.S. Stein, J.B. Lian and T.A. Owen, "Relationship of cell growth to the regulation of tissue-specific gene experession during osteoblast differentiation," FASEB J., 4, 3111-3123 (1990).

10. G.E.R. Schoeters, L. de Saint-Georges, R. Van Den Heuvel and O. Vanderborght, "Mineralization of adult mouse bone marrow in vitro," Cell Tissue Kinetics, 21, 363-374 (1988).

11. E. Mathieu, G. Schoeters, F. Vander Plaetse and J. Merregaert, "Establishment of an Osteogenic cell line derived from adult mouse bone marrow stroma by use of recombinant retrovirus," Calcif. Tissue Int., 50, 362-371 (1992).

12. M. Miwa, O. Kozawa, H. Tokuda, A. Kawakubo, M. Yoneda, Y. Oiso and K. Takatsuki. "Effects of hypergravity on proliferation and differentiation of osteoblast like cells," Bone and Mineral, 14, 15-25 (1991).

13. J.J.W.A. van Loon, J.P. Veldhuijzen and E.H. Burger, "Hypergravity and bone mineralization," in Proc. Fourth Europ. Symp. on Life Sci. Res. in Space, ESA SP-307, V. Davis (ed.), ESA Publication Division, ESTEC, Noordwijk, The Netherlands, 1990, pp. 393-396.

14. G.A. Ubbels, "The role of gravity in the establishment of the dorso/ventral axis in the amphibian embryo," in: Biorack on Spacelab D1, N. Longdon, V. David (ed.), ESA Publication Division, ESTEC, Noordwijk, The Netherlands, ESA SP-1091, 1988, pp. 147-155.

15. B. Peterkofsky and R. Diegelmann, "Use of mixture of proteinase- free collagenase for the specific assay of radioactive collagen in the presence of other proteins," Biochemistry, 10, 988-994 (1971).

16. D.E. Andersen and V. Nielsen, "In vitro screening for acute cytotoxicity on plastic materials," 5th Int. workshop on in vitro toxicology. Schloss Elmau, F.R.G., 2-4 (1988).

17. R.P. de Groot, P.J. Rijken, J. den Hertog, J. Boonstra, A.J. Verkleij, S.W. de Laat and W. Kruijer, "Microgravity decreases c-fos induction and serum response element activity," J. Cell Sci., 97, 33-38 (1990).

18. K. Iio, N. Minoura, S. Aiba, M. Nagura and M. Kodama, "Cell growth on poly(vinyl alcohol) hydrogel membranes containing biguanido groups," J. Biomed. Mater. Res., 28, 459-462 (1994).

19. J.C. Wataha, C.T. Hanks and R.G. Craig, "In vitro effects of metal ions on cellular metabolism and the correlelation between these effects and the uptake of the ions," J. Biomed. Mater. Res., 28, 427-433 (1994).

20. A. van Sliedregt, J.A. van Loon, J. van der Brink, K. de Groot and C.A. van Bitterswijk, "Evaluation of polylactide monomers in an in vitro biocompatibility assay," Biomaterials, 15(4), 251-256 (1994).

21. C.F. Mandenius and L. Ljunggren, "Ellipsometric studies of plasma protein adsorption on membrane polymers for blood purification," Biomaterials, 12, 369-373 (1991).

22. M. Spector, M.J. Michno, W.H. Smarook and G.T. Kwiatkowski, "A high-modules polymer for porous orthopedic implants: Biomechanical compatibility of porous implants," J. Biomed. Mater Res., 12, 665-677 (1978).

23. J.C. Knowles, G.W. Hastings, H. Ohta, S. Niwa and N. Boeree, "Development of a degradable composite for orthopaedic use: in vivo biomechanical and histological evaluation of two bioactive degradable composites based on the polyhydroxybutyrate polymer," Biomaterials, 13(8), 491-496 (1992).

24. D.A. Puleo, K.E. Preston, J.B. Shaffer and R. Bizios, "Examination of osteoblast-orthopaedic biomaterial interactions using molecular techniques," Biomaterials, 14(2) (1993).

25. C.R. Angle, D.J. Thomas and S.A. Swanson, "Osteotoxicity of cadmium and lead in HOS TE 85 and ROS 17/2.8 cells: relation to metallothionein induction and mitochondrial binding," BioMetals, 6, 179-184 (1993).

We like to thank Dr. M. Heppener for active participation in the preparations for these experiments and CCM, Nuenen, for providing the necessary test hardware. This work is partly supported by Dienst voor de Programmatie van het Wetenschapsbeleid (DPWB/SPPS) via PRODEX, the Space Research Organization of the Netherlands (SRON) grant MG-004, and by the Dutch Organization of Scientific Research (NWO) grant 900-541-133.