WO1994012624A1

WO1994012624A1 - Crystal carriers for cell culture

Info

Publication number: WO1994012624A1
Application number: PCT/US1993/011518
Authority: WO
Inventors: Benjamin Geiger; Lia Addadi; Dorit Hanein
Original assignee: Yeda Research And Development Co. Ltd.; Rycus, Avigail
Priority date: 1992-11-29
Filing date: 1993-11-29
Publication date: 1994-06-09
Also published as: IL103919A0

Abstract

Carrier materials are provided for the cultivation of anchorage-dependent cells, which materials comprise biocompatible crystals that allow normal cell growth and development. Examples of crystals are calcite, calcium (R,R)- and (S,S)-tartrate tetrahydrate, brushite and gypsum crystals. The cells, both human and non-human animal cells, can be cultivated using said carriers in a medium containing serum or in serum-free medium.

Description

CRYSTAL CARRIERS FOR CELL CULTURE

Field and Background of Invention

The present invention relates generally to the culture of cells, and more particularly to a carrier suitable for cultivation of anchorage-dependent cells. The carriers of the invention comprise biocompatible crystals whose surfaces may serve as substrates for cell adhesion.

Many biologically active substances are nowadays produced by cell culture, both for research and commercial purposes . Most cells used in cell culture are anchorage- dependent and require a solid substrate to attach to in order to proliferate, metabolize nutrients and produce biomolecules .

In order to provide for growth of adherent cells on the fixed surface of a substrate, the substrate must have a surface nature such as to permit the cells satisfactorily to adhere thereto. Adhesion to exogenous surfaces has profound effects on the structure and behavior of cells, especially of anchorage- dependent cells .

Cell-substrate adhesion is commonly viewed as a molecularly complex and diversified multistage process, which includes cell attachment and spreading, focal adhesion formation, extracellular matrix deposition and rearrangement. In general, cell attachment to and spreading on surfaces may occur via two distinct but highly interdependent mechanisms; specific binding meadiated by adhesive extracellular matrix (ECM) proteins and direct association with the underlying solid surface. The extracellular proteins that together form the various extracellular matrices include fibronectin, laminin, vitronectin, collagen and different proteoglycans . These proteins may bind both to external substrates and to cell surface receptors, mostly of the integrin superfamily. Those two binding processes, may however be interdependent, since the composition of the substrate-attached protein* layer and the conformation of the adsorbed proteins may be affected by the adsorption to the surface. Thus, cell adhesion and growth of anchorage-dependent cells, may depend directly or indirectly on the nature of the surface to which ECM-proteins adsorb. Although much information is available on the identity and structure of "adhesive proteins", only limited information exists on the precise surface chemistry of adhesion-promoting substrates, which either form direct associations with the plasma membrane or promote ECM- and receptor-mediated cell adhesion.

In order to achieve high levels of cell density which are necessary for upscaling the cell culturing process, and to increase the surface-to-volume ratio for economy on operational space, several substrates for cultivation of cells have been proposed, such as spherical carrier beads, known as microcarriers, polymeric sponge matrices, multiple tubings, hollow fibers or packing material such as glass beads placed in columns through which the cell culture medium is circulated. European Patent No. EP 289666 discloses a packing material element for use in bulk cultivation of anchorage-dependent cells, which element comprises a length of plastics material, e.g., polystyrene, formed by twisting a plane sheet or ribbon about a longitudinal axis .

The most common substrates used for cell culture consist of a hydrophobic polymer with hydrophilic, polar and ionic groups such as hydroxyl, carbonyl, carboxylate and sometimes sulfate groups (Maroudas, 1977; Van ache et al. , 1987). The structural organization and distribution of these groups on the surfaces, is, however, unknown.

Characterization of the properties of synthetic substrates that possess adhesive potential is mostly based on chemical modification of polymeric surfaces. These include modulation of hydrophobicity/hydrophilicity (Maroudas, 1977; van Wachem^" et al. , 1985; Pratt et al. , 1989; Dekker et al . , 1991), charge density (van Wachem et al. , 1987; Curtis and McMurray, 1986), physical and chemical heterogeneity (Lydon et al., 1985), and defor ability (Guidry and Grinnell, 1987) . It is, however, noteworthy that such definitions are macroscopic in nature, insufficiently defined to describe the structural properties of the interface accurately and that most surfaces used for cell culture are highly heterogeneous at the molecular level (Curtis and Clark, 1990).

While the absence of uniform surface organization on common substrates such as glass and tissue culture plastic is evident, the significance of this heterogeneity for cell adhesion is not clear. One possibility is that, since adhesive cellular interactions occur via structures such as focal and close contacts that characteristically have dimensions of several square micrometers, the adhesivity of the substrate might not be significantly affected by angstroms-scale fluctuations in surface structure. Alternatively, surface recognition by cells may depend on highly specific molecular conformations and distributions, and thus chemically equivalent but structurally distinct surfaces may differ in their adhesion potential. To distinguish between these possibilities, it appears mandatory to examine cell adhesion to surfaces whose composition and structure are fully defined and which may be experimentally modulated.

Crystals expose to the environment characteristic faces whose structures correspond to that of the bulk, limited by planes corresponding to specific crystallographic directions and defined by the crystal morphology. Crystal surfaces can be regarded as being composed of "active sites" of well-defined and well-known structures, that interact stereospecifically with molecules or macromolecules in solution, at a level of recognition comparable to that of enzyme-substrate or antibody- antigen interactions (Weissbuch et al. , 1991; Weiner and Addadi, 1991; Ka et al., 1992). The advantage of these surfaces lies in their highly ordered, repetitive arrangements and the knowledge available on their structure at the atomic scale, which offers accurate means to define structural and chemical parameters. Crystals of different chemical composition can be chosen among a wide variety of known structures . The modulation of the interactions with the same molecules, exposed at different faces of the same crystal in different orientations, adds further to this variety. Summary of the Invention

It has now been found in accordance to the present invention that crystals are suitable substrates for cell adhesion, supporting normal cell growth and development.

The present invention thus relates to a carrier for the cultivation of anchorage-dependent cells which comprises biocompatible crystals being non-toxic to the cells and allowing normal cell growth and development.

Examples of crystals that may be used as carriers according to the invention are calcite (CaCO,), calcium (R,R)-tartrate tetrahydrate [Ca⁺² (-OOC-CH(OH)-CH(OH)-COO) -] 4H₂0, calcium (S,S)-tartrate tetrahydrate, brushite (CaHP0..2H₂0) and gypsum (CaSO..2H.-0) . These crystals present a diversity of well-characterized surfaces, ranging from highly ionic (calcite) to moderately charged and hydrophilic (tartrate), with different stereochemical characteristics and crystalline water distribution. Major differences in the adhesive response of the cells to these crystal surfaces were found, these differences being particularly intriguing in the tartrate crystals, where the cells distinguish between two chemically equivalent faces, pointing at the fine specificity of adhesive interactions.

The invention further relates to a process for the cultivation of anchorage-dependent cells using biocompatible crystals as carrier substrates . The cultured cells may be used for production of biomolecules, e.g., compounds of commercial or scientific value produced by cells transfected with the gene coding therefor, or for implants, such as of vascular grafts of endothelial cells, skin grafts of cultured keratinocytes, etc.

Brief Description of the Drawings

Fig. la shows the number of A6 cells attached to calcite crystals following different incubation times, and under different culture conditions. Complete medium (DMEM +S, empty squares); serum-free medium (DMEM, filled circles); complete medium, after 24 hrs conditioning of the crystals in complete medium (DMEM Cond. , filled squares). The error bars represent the standard deviation of the data points . Cell counts were

2 performed on each crystal face and normalized to mm area.

Fig. lb is a histogram of cells attached to calcite crystals at different incubation times, in complete medium, in the absence (DMEM + S, black columns) and in the presence (DMEM + S/RGD, lined columns) of 50 μg/ml GRGDS peptide. Cell counting procedure as in la.

Fig. 2 shows the number of cells attached to the {011} faces (empty squares) and {101} faces (filled circles) of calcium (R,R)-tartrate tetrahydrate crystals, following different incubation times, in complete medium. Error bars and counts are as described for Fig. 1.

Figs. 3a and 3b show the number of cells attached to the {011} and {101} faces, respectively, of calcium (R,R)-tartrate tetrahydrate crystals at different incubation times and under different incubation conditions: in complete medium (DMEM + S, empty squares); in serum-free medium (DMEM, filled circles); and in complete medium after 24 hrs conditioning of the crystals in complete medium (DMEM Cond. , filled squares).

Figs. 4a and 4b show histograms of cells attached to the {011} and {101} faces, respectively, of calcium (R,R)-tartrate tetrahydrate crystals, at different incubation times (10 and 60 minutes) and culture conditions: complete medium (DMEM + S, filled columns); complete medium + 50μg/ml GRGDS peptide (DMEM + S/RGD, bold striped columns); complete medium after conditioning of the crystals for 24 hrs in complete medium (DMEM + S Cond., dotted columns); and complete medium after conditioning of the crystals for 24 hrs in complete medium + 50μg/ml GRGDS peptide (DMEM + Ξ Cond./RGD, striped columns).

Detailed Description of the Invention

While screening a variety of crystals to serve as substrates for cell growth, it was found that there is a high variability between crystals. Some of the crystal systems were apparently toxic to the cells while others were compatible with cell growth or even stimulated it (data not shown) . Thus selected crystals, in addition to their capacity to serve as defined adhesion substrates, may also serve as useful substrates for cultivation of cells .

A biocompatible crystal according to the invention will allow cell attachment to its surfaces, cell spreading and normal cell growth and development. For this purpose, the chemical components of the crystal in the soluble form should be non-toxic to the cells. In addition, the crystal should have at least one face type that allows normal cell growth and development (for definition of "crystal face", see Example 2 below). All faces of the crystals may be biocompatible, i.e., they all allow adhesion, spreading and normal cell growth and development, or some of the crystal faces are biocompatible and others are not, allowing neither adhesion nor spreading of the cells or allowing only cell adhesion but no spreading and growth of the cells.

Any biocompatible crystals as herein defined that allow normal cell growth and development are encompassed by the present invention. Examples of such crystals are calcite, calcium (R,R)- and (S,S)- tartrate tetrahydrate, brushite, gypsum, etc. It may be found that a material will be suitable for a certain type of cell but not for other cells.

Examples of anchorage-dependent cells that may be cultivated using crystals as carriers are human cells, e.g., fibroblasts, endothelial cells, keratinocytes, etc. , animal cells such as A6 kidney cells derived from Xenopus laevis, insect cells and mammalian cells, e.g., Chinese hamster ovary (CHO) and baby hamster kidney (BHK) cells.

For large scale cell culture, properly selected crystals are produced in suitable amounts. The crystals should be of uniform dimensions offering optimal conditions for cell growth, i.e., they should be large enough to provide space for adhesion and spreading, and small enough to obtain large surface to volume ratio. These are necessary conditions for high cell production. In one embodiment of the invention, the crystals will be used as micrσcarriers, suspended in fermenters or reactors containing the cell culture medium.

In the method of the invention, the anchorage-dependent cells may be cultivated in serum-containing medium as well as in a low serum or serum-free medium. The medium will be preferably saturated with respect to the particular crystal used, meaning that the chemical components of the crystal in their soluble form are present in the medium in a concentration that prevents dissolution of the crystals.

The crystal carriers of the invention present several advantages, namely, they are chemically defined substrates; allow efficient cell growth; they make possible the recruitment of serum and cell growth factors; they support cell growth in low or no serum in the medium; it may be possible to dissolve the crystal substrate after cell growth, if desired, thus obtaining intact cells and not contaminated by solid support material, by appropriately lowering the degree of saturation of the medium; the crystals may be found suitable for the growth of cells otherwise difficult to grow, e.g., keratinocytes, endothelial cells, and they enable recovery of cells without the use of proteolytic enzymes that are routinely used for the recovery of adherent cells following their removal from solid polymeric substrates.

According to the present invention, the mode of cell adhesion to calcite and calcium (R,R)-tartrate tetrahydrate crystals has been characterized. It is shown that calcite binds A6 Xenopus laevis epithelial cells rapidly and efficiently, most likely via surface-adsorbed proteins. Surface topology had only a limited effect on the adhesive interactions. Calcium (R,R)-tartrate tetrahydrate crystals exhibit two chemically equivalent, yet structurally distinct, faces that differ mainly in the surface distribution of their lattice water molecules and charges. However, despite the gross similarity between the two faces, striking differences were noted in their adhesive behavior. One of the face types denoted as {Oil}, was highly adhesive for A6 cells. A masssive attachment of A6 cells to {011} faces was observed shortly (10 minutes) after plating, followed by spreading and cell death. This binding is presumably independent of exogenous proteins as it is not affected by addition of RGD-peptides or the presence of serum in the medium. In contrast, cell adhesion to the other surface of tartrate, denoted as {101}, was relatively slow (>24 hrs) and apparently mediated by RGD-containing protein(s), since the addhesion was promoted by serum proteins and it was RGD- inhibitable. It was shown that the {101} faces of tartrate crystals could be "conditioned" by long (24 hrs) incubation with serum-containing medium, after which they become highly adhesive.

The results presented here thus indicate that crystal surfaces may serve as excellent, structurally defined, substrates for cell adhesion, that cell binding may occur directly or via RGD-containing proteins and that cell adhesion may be dramatically modulated by variations in surface structure.

The invention will now be illustrated by the following non-limiting examples .

EXAMPLES Materials and Methods

( a) Crystallization experiments : For each crystal system optimal conditions for crystallization from aqueous solution were determined, ensuring that the crystals were well formed, homogeneous and reproducible with respect to morphology. All crystallization experiments were carried out at room temperature.

(i) Calcite : Crystallization was induced by slow diffusion (over a period of 3 days) of ammonium carbonate vapor into cell culture dishes (Nunc, 3.5 cm) containing 2 ml of 7.5mM CaCl₂.2H~0 (pH 6.5) in a closed desiccator (Addadi and Weiner, 1985). The crystals are typically 100-200 μ_m in size.

(ii) Calcium (R,R)-tartrate tetrahydrate : A 10 ml sample of a solution of 3.2 mM sodium hydrogen tartrate was mixed with 10 ml of a solution containing 4.4 mM CaCl-.2HpO at pH 6.5 and aliquoted in 3.5 cm cell culture dishes (Falcon, USA) (Addadi and Weiner, 1985). Crystals of typically 1 mm in size form within one day. Crystal morphology was determined by X-ray diffraction on a Nonius CAD-4 diffracto eter (Addadi et al., 1982). In the scanning electron microscope, the tartrate face types can be further directly distinguished by the distinct morphology of the etch-pits on the different faces, and their different sensitivity to the electron beam.

(iii) Calcium (S ,S)-tartrate tetrahydrate: A 30 ml sample of a solution of 4 mM potassium hydrogen tartrate was mixed with 30 ml of 4.3 mM CaCl₂.2H₂0 at pH 6.5 and aliquoted in 3.5 cm cell culture dishes (Falcon, USA). Typically, crystals of about 1 mm size form within one day.

(b) Cell culture : The A6 kidney cell line derived from Xenopus laevis (ATCC, U.S.A.) was cultured, at 28° C, in 85% DMEM, supplemented with 8.5% fetal calf serum ("complete medium") in a humidified atmosphere of 5% CO,, in air. The cells were seeded in saturated complete medium (see below), on the same culture dishes in which the crystals were previously grown, still attached to the dish. To avoid crystals dissolution, all cell culture medium, fixation and washing solutions were saturated with respect to the particular crystals used (saturated complete medium).

( c ) Electron microscopy ( E1 ) : (i) Transmission EM (TEM)- A6 cells seeded on calcite crystals were fixed for 30 min in 2% glutaraldehyde and 3% paraformaldehyde in 0.1M cacodylate buffer, pH 7.2, containing 5 mM CaCl- (The presence of CaCl- in the fixation buffer prevented dissolution of the crystals). The dishes were rinsed four times and embedded in a thin layer of 1.7% agar to prevent the collapse of the cell monolayer grown on the crystals . The calcite crystals were dissolved at pH 4 and 4°C during three days. After complete dissolution of the crystals, cells were postfixed with 1% osmium tetraoxide, 0.5% potassium ferrocyanide and 0.5% potassium dichromate in 0.1M cacodylate buffer, pH 7.2. The cells were stained en bloc with 2% aqueous uranyl acetate, followed by ethanol dehydration. The dishes were embedded in t-Epon 812 (Tuosimis, Maryland, USA). Sections of 500-700°A were cut using a diamond knife (Diatome, Switzerland) , and examined using a Philips CM-12 transmission electron microscope at an accelerating voltage of 100 kV.

The preparation of the cells seeded on tartrate crystals for the TEM followed the same procedure as described above, except that fixation buffer was saturated with respect to tartrate and the crystals were dissolved in 0.2M EGTA, pH 7.2.

(ii) Scanning EM (SEM) : All the solutions were saturated with respect to the particular crystal used. Fixation was carried out as described above for the TEM, followed by ethanol dehydration using Microporours Spec Cap (Spi Supplies, USA) and critical point drying (Autosamdri-810, Tuosimis, Maryland, USA) with CO-. The crystals were placed on carbon- coated stubs (Spi Supplies, USA) and sputter coated with gold for 12 minutes at 6 mA (S150 Edwards, USA) . The specimens were examined at an accelerating voltage of 15-20 kV using a JEOL 6400 scanning electron microscope.

(d) Cell counting : The number of cells attached to the surface was directly counted for each crystal during SEM

2 observations and the cell density per mm area was calculated.

(e ) Synthetic peptide inhibition : Adhesion to surfaces was also tested in culture medium supplemented with a synthetic peptide containing the Gly-Arg-Gly-Asp-Ser sequence (RGD peptide) . A control peptide containing the Gly-Arg-Gly-Glu-Ser (RGE) sequence was also examined. Both peptides were synthesized at the peptide synthesis unit of the Weizmann Institute of Science, Rehovot, Israel and purified by high pressure liquid chromatography.

Example 1

Choice of cells and crystals for adhesion assays

A series of preliminary experiments with cultured A6, BHK and CHO cells indicated that calcite, calcium (R,R)- and (S,S)- tartrate tetrahydrate, brushite and gypsum crystals provide compatible surfaces upon which the cells may grow. A series of additional crystals was examined and found unsuitable for adhesion. These included calcium fumarate trihydrate and calcium creatine monohydrate crystals, which had toxic effects.

A6 cells were chosen in a set of experiments because they are optimally cultured at 28° C, a temperature that is lower than that commonly used (37°C) for mammalian cell culture. These conditions prevent drastic temperature changes during sample handling, which may cause crystal dissolution or reprecipitation. CHO and BHK cells were cultured at 37° C. Routinely, cell growth was performed in the presence of 8.5% fetal calf serum, saturated with respect to the particular crystal used (saturated complete medium). The salt concentrations after equilibration with the crystals did not significantly exceed the normal values in complete medium

+2 +2

(calcite, 1.44 mM Ca , 1.44 mM C0₂; tartrate, 2 mM Ca ,

2 mM tartrate) . Light microscopy revealed that cell attachment, spreading and growth in calcite or tartrate-saturated complete medium, is indistinguishable from that observed in non- saturated complete medium (data not shown) .

Example 2

Adhesion and spreading of A6 cells on calcite crystals

Calcite is a polymorph of calcium carbonate, stable at physiological conditions. The rhombohedral crystals (100-200 μ.m in size) are delimited by six equivalent faces, the {104} faces characterized by a tight packing of calcium and carbonate ions emerging at an oblique angle to the surface (Lippmann, 1973). Crystal faces are described by a set of indices (hkl) that unequivocally define the orientation of the face relative to the crystallographic axes a., b and c. of the structure. The notation {hkl} eg. {104} or {Oil}, {101} includes a set of identical, symmetry-related faces.

A6 cells were seeded in complete medium on the culture dishes where the calcite crystals were grown. This procedure allows the crystals to stay attached to the surface of the culture dish during the experiments (which is not a necessary requirement for cell growth) . The adhesion of A6 cells to the {104} faces of calcite was monitored after 1, 24 and 72 hours of incubation directly by optical microscopy and scanning electron microscopy. One hour after seeding only a small percentage of cells were adsorbed to the crystal surfaces and adhesion proceeded for several hours. The cells were initially moderately spread and were detected on essentially all crystal faces, irrespective of their orientation vis-a-vis the plane of the culture dish. Further incubation resulted in extensive spreading of the A6 cells accompanied by the development of a typical rough- dorsal surface. After 24 hours of incubation, an extracellular matrix layer deposited on the bare crystal surface was observed. By 72 hours the crystals were mostly covered by a confluent monolayer of flat epithelial cells . The monolayer preserves the characteristic angles between adjacent crystal faces. Macroscopic surface roughness neither favored cell attachment and spreading, nor prevented it.

The SEM observations were confirmed by TEM analysis of A6 cells attached to calcite crystals, following 72 hours of incubation. In a cross section of a whole calcite crystal (relative to the crystal face) attached to the culture dish, tight adhesion to the substrate was observed, including the formation of distinct focal contacts associated with cytoplasmic actin-connecting filaments . Ultrathin sections revealed the general morphology of the monolayer, still preserving the overall crystal morphology. Furthermore, the cells in the monolayer displayed a typical epithelial morphology with characteristic intercellular junctions and often spread over crystal edges.

At their ventral surface, cells were apparently associated with a layer of ECM which could be visualized by SEM, either between spread cells or by examination of the undersurface of cells that occasionally detached. Cross- sections examined by TEM showed a distinct and nearly continuous "serum line", running along the ventral cell membrane. These lines were also continuous across the crystal/plate interfacial junction.

To determine whether adhesion to calcite surfaces is mediated by specific Gly-Asp-Arg (RGD)-containing adhesion proteins such as fibronectin or vitronectin, 50 μg/ml Gly-Arg- Gly-Asp-Ser (GRGDS) synthetic peptide were added to the culture medium. As shown in Fig. lb, this treatment resulted in an essentially complete inhibition of the attachment when examined up to 1 hr after plating. The RGD peptide is known to be degraded upon longer incubation in culture. In contrast, the RGE peptide, added under the same conditions, had no apparent effect.

To distinguish further between the contributions of exogenous proteins (supplied in the medium) and endogenous proteins (secreted by the cells) to the adhesion process, cell spreading on calcite crystals was examined in serum-free medium (Fig. la). Under these conditions, the initial attachment of cells following 1 hr of incubation, was slightly higher than that observed in the experiments using complete medium. On the other hand, after further incubation (> 6 hrs), the adhesion of A6 cells to the calcite surface was lower than that obtained in complete medium.

"Conditioning" of the calcite crystals by incubation with complete medium for 24 hrs before the seeding of cells, slightly inhibits the short-term adhesion of cells to the crystals (Fig. lb) .

Example 3

Adhesion and spreading of A6 cells on calcium (R.R)- tartrate tetrahydrate crystals - Calcium (R,R)-tartrate crystallizes in a structure containing four molecules of water per molecule of tartrate as a part of the crystal lattice. The structure is thus characterized by a net of combined ionic interactions and hydrogen bonds (Ambady, 1968) . The large prismatic crystals (1 mm size) are delimited by two different face types, denoted {011} and {101}. As the two face types delimit the same crystal structure from different directions, by definition they are chemically equivalent but differ in their structure and stereochemical organization. The orientation of the carboxylate and hydroxyl groups of the tartaric acid, as well as the organization and distribution of calcium and water molecules on these surfaces, are different. The cell adhesion processes on the two face types were observed to be drastically different, and will be thus described separately.

3.1 The adhesive behavior of the -fOll} faces. The attachment of cells to the {011} faces can be described as an exponentially decaying function, with an average of 1350

2 cells/mm attached to the faces initially (within 10 min) ,

2 sharply declining to an average of 47 cells/mm observed at

72 hrs (Fig. 2). Shortly after plating, a massive attachment of cells to {011} faces was observed. SEM and TEM analysis of cell adhesion to the crystals following 10 minutes of incubation, indicated that the adherent cells were initially spherical and attached to the {011} surface through a foot-like structure. In contrast to cells plated on calcite, cells attached to the

{011} faces remained spherical during the first several hours and only after about 6 hours was moderate spreading on the surface apparent. During this period the overall morphology was normal, yet upon longer incubation, progressive cell death was observed on these faces and at time points of 24-72 hrs living cells were rarely detected on it, leaving mainly patches of cell debris and/or extracellular matrix. By TEM, a discontinuous serum line was observed.

Addition of the GRGDS peptide to the growth medium had only a marginal effect on the cell attachment to the {011} faces (Fig. 4a) . Similarly, there was no apparent effect of the

GRGDS peptide on cell adhesion to these faces in the

"conditioning" experiments.

3.2 Adhesive behavior of the {101} faces. The attachment of cells to the {101} faces can be described as an

2 exponential function, with an average of 10 cells/mm attached to the faces initially (at 1 hr after seeding), up to 2 an average of 1500 cells/mm observed at 72 hrs. In contrast to the rapid adhesion of A6 cells to {011} faces, attachment and spreading on the {101} faces proceeded slowly (Fig. 2) . Up to 24 hrs after seeding, only a small number of cells were detected on these faces. However, after further incubation, the

{101} faces gradually become coated with well-spread cells, possessing morphology similar to that of cells growing on calcite, or on tissue culture dishes. It appears that these cells migrate to the {101} faces from the dish surface and/or from the border of {101} faces. TEM analysis revealed a continuous "serum line", focal contacts and well-developed cytoskeletal organization.

The adhesion mechanism to the {101} faces was further investigated by serum-conditioning of the crystals before cell plating. It was found that adhesion of A6 cells to conditioned crystal {101} faces started shortly after seeding (10 min)

(Fig. 3b) and within 24 hrs these faces were covered by a nearly confluent layer of well-spread cells with normal morphology. This is in contrast to the slow adhesion kinetics described for the same crystal faces in either complete medium or serum-free medium. Addition of 50 μg/ml GRGDS peptide to cells plated on conditioned crystals resulted in a complete inhibition of cell adhesion to the {101} faces. These results indicate that the adsorption and organization of RGD-containing adhesion proteins on {101} faces is essential for cell attachment.

Example 4

Adhesion and spreading of A6. CHO and BHK cells on other crystals - A6, CHO and BHK cells were seeded in complete medium or low serum or serum-free medium on culture dishes where brushite, gypsum or calcium (S,S)-tartrate tetrahydrate crystals were grown, under the same conditions as in Examples 1 to 3. For CHO and BHK cells, the temperature of the culture was 37° C. In all experiments, the crystals were found to be compatible with the cells and allowed normal cell growth and development. REFERENCES

Addadi, L., Berkovitch-Yellin, Z., Do b, N., Gati, E., Lahav,

M. and Leizerowitz, L. (1982) Nature 29.6:21-26. Addadi, L. and Weiner, S. (1985) Proc . Natl. Acad. Sci. USA

£1:4110-4114. Ambady, G.K. (1968) Acta Crystallogr. sec. B 24:1548-1557. Curtis, A.S.G. and Clark, P. (1990) Critc. Rev. Biocomp.

5_:343-363. Curtis, A.S.G. and McMurray, H. (1986) J. Cell Sci. £6.:25-33. Dekker, A., Reitsma, K. , Beugeling, T., Bantjes, A., Feijen, J. and van Aken, W.G. (1991) Biomaterials 12:130-138. Guidry, C. and Grinnell, F. (1987) J. Cell Biol. 104:1987-1997. Kam, M. , Perl-Treves, D., Caspi, D. and Addadi, L. (1992)

FASEB £:2608-2613. Lippmann, F. (1973) Sedimentary Carbonate Minerals

(Springer, Berlin), pp. 6-16. Lydon, M.J., Minett, T.W. and Tighe, B.J. (1985) Biomaterials

.6:396-402. Maroudas, N.G. (1977) J. Cell Physiol. 90:511. Pratt, K.J., Williams, S.K. and Jarrel, B.E. (1989)

J. Bio ed. Mat. Res. .23:1131-1147. van Wachem, P.B., Beugeling, T., Feijen, J., Bantjes, A.,

Det ers, J.P. and van Aken, W.G. (1985) Biomaterials

.6:403-408. van Wachen, P.B., Hogt, A.H., Beugeling, T., Feijen, J. ,

Bantjes, A., Detmers, J.P. and van Aken, W.G. (1987)

Biomaterials 8_.:323-328. Weiner, S. and Addadi, L. (1991) Trends Bioche . Sci. 16:252- 256. Weissbuch, I., Addadi, L., Lahav, M. and Leiserowitz, L. (1991)

Science 253: 637-645.

Claims

1. A carrier material for the cultivation of anchorage- dependent cells, said material comprising biocompatible crystals that allow normal cell growth and development.

2. A carrier material according _\ to claim 1 wherein the crystals have at least one face type that allows normal cell growth and development.

3. A carrier material according to claim 2 comprising a crystal system selected from the group consisting of calcite, calcium (R,R)-tartrate tetrahydrate, calcium (S,S)-tartrate tetrahydrate, brushite, and gypsum crystals.

4. A method for the cultivation of anchorage-dependent cells which comprises cultivating said cells in low serum or serum-free medium using as carrier biocompatible crystals according to claim 1.

5. A method for the cultivation of anchorage-dependent cells which comprises cultivating said cells in a serum- containing medium using as carrier biocompatible crystals according to claim 1.

6. A method according to claim 4 or 5 wherein the biocompatible crystals are selected from the group consisting of calcite, calcium (R,R)-tartrate tetrahydrate, calcium (S,S)- tartrate tetrahydrate, brushite and gypsum crystals.

7. A method according to any one of claims 4 to 6 wherein the cells are human cells.

8. A method according to claim 7 wherein the cells are human fibroblast, endothelial or keratinocyte cells.

9. A method according to any one of claims 4 to 6 wherein the cells are non-human animal cells.

10. A method according to claim 9 wherein the cells are A6 kidney cells derived from Xenopus laevis.

11. A method according to claim 9 wherein the cells are non- human mammalian cells.

12. A method according to claim 11 for the cultivation of BHK and CHO cells.

13. The use of biocompatible crystals in which at least one surface thereof serves as substrate for cell adhesion and allows normal cell growth and development, as carriers for the cultivation of anchorage-dependent cells.