biocompatibility of echinoderm skeleton with mammalian cells in vitro: preliminary evidence

Biocompatibility of echinoderm skeleton with mammalian cells in vitro: Preliminary evidence

A. R. Fontaine and B. D. Hall Departineiit of Biology, Unizrersity of Victoria, Victoria, British Colunibia, Canada V8 W 2Y2

The physical and chemical properties of echinoderm skeleton are reviewed. A method is described for preparing cell-free, sterile echinoderm skeletal plates (ossicles) which were used as porous substrates for cell cultures. Ossicles of the starfish Pisasfrr ochraceus were evaluated as substrates for the culture of three mammalian cell lines. Each line grew vigorously on ossicles, and fibroblasts quickly infiltrated their porous microstructure. Echinoderm skeletal plates

provide a simple, convenient alternative to coverslips and porous membranes for SEM or correlated SEM/TEM studies of cell behavior, More importantly, the preliminary evidence for biocompatibility presented suggests that native echinoderm skeleton has potential use as a biomaterial and, because of its microstructure and relative solubility; deserves evaluation as a kind of biodegradable ceramic.

INTRODUCTION

The calcareous plates (ossicles) that constitute echinoderm endoskeleton have unusual properties.1-4 They are formed of magnesium-rich calcite secreted in crystallographic homogeneity and lack any appreciable amount of organic ground substance. Consequently the skeleton is a system of purely inorganic plates, each of which behaves optically as a single crystal without respect to its gross configuration. Nevertheless, ossicles probably are secreted as polycrystalline aggregate^.^ Their most significant morphological property is a fenestrate microstructure known to echinoderm biologists as the “stereome.” For all echinoderm species, every skeletal plate is permeated by a system of interconnecting pores (Figs. 1 and 2) that convert the ossicle into a lightweight inorganic framework retaining the strength of solid calcite.4 In life the pores contain a stroma of diverse kinds of cells, and abundant collagen in many places. Collagen fibers in the form of loops that suture ossicles to- gether or serve as tendons for muscle insertions pass through the pores and around the calcite trabeculae, which thus function as b o l l a r d ~ . ~ Pore mean diameter is ca. 20 pm (range 10-50 pm); trabecular cross-sectional diameters are similar. The proportion of organic to skeletal material within an ossicle is about equal.6 Topologically cchinoderm skeleton is a “periodic minimal surface,” the only example known to occur naturally. Such a surface divides space into two intermingled and multiply connected domains, maximizing their surface areas in ~ o n t a c t . ~

Journal of Biomedical Materials Research, Vol 15, 61-71 (1 981) IC) 1981 John Wiley & Sons, Inc 0021-9304 /8 l /OOl~-O06l$Ol 10

62 FONTAINE AND HALL

Figure 1. Ambulacral ossicle of the starfish Pisastcr or:hracrus prepared for seeding with a cell suspension. Its distinctive asymmetry is advantageous for orientation. By restricting use of forceps to the "handle" area (H), the potential for damaging seeded cells is minimized. Bar, 1 mm.

We have considered the potential use of cell-free, sterile echinoderm ossicles as a substratum for the growth of mammalian cells in culture. We reasoned that calcite ossicles should provide an acceptable surface; indeed the Ca ion content could conceivably enhance cellular attachment and movement. Be- cause of the lack of an appreciable organic matrix, there should be minimal toxic or immune reactions. The continuous porosity, the dimensions of the pores, and the surface texture suggest a natural material potentially attractive to cells, particularly fibroblasts. Given these considerations, we tested the growth of three mammalian cell lines on ossicles of the starfish, Pisaster och- rficcoils, a common species in local waters.

MATERIALS AND METHODS

Cleaning ossicles

Ossicles were cleaned free of cells and organic debris as follows: (1) Severed starfish arms dissected free of superficial tissues were immersed for up to 2.5 hr in distilled water in an ultrasonic cleaning bath; frequent water changes prevented over-heating. (2) Tissue dissolution was completed by immersion in 6%) sodium hypochlorite solution (commercial bleach is satisfactory) with constant slow rotation for up to 3.5 hr; frequent solution changes are necessary to yield thoroughly cleaned ossicles. (3) Finally, the loose ossicles were given exhaustive washes in distilled water, ending with double glass-distilled water. Cleanliness and degree of surface-etching were monitored by SEM.

ECHINODERM SKELETON 63

Figure 2. demonstrated by SEM; ambulacral ossicle of P. ociirncrus. Bar, 50 pm.

The porous microstructure characteristic of echinoderm skeleton

Culturing cells

Cells were cultured on ambulacral ossicles (Fig. 1). These were chosen because of their abundance, uniform size (6 X 2 X 1 mm) and distinctive shape that proved convenient for orientation and handling. The cells tested were serially propagated lines of Chinese hamster ovary fibroblasts, Puck's clone A (CHA); a locally isolated line of human skin fibroblasts (HSF); and human prostate cancer cells (HPC), Microbiological Associates MA-1 60. These lines are maintained at the University of Victoria stored in liquid nitrogen. Cells were cultured at 37°C in a complete tissue medium based on Eagles' minimal essential medium, with Earle's balanced salt solution, supplemented with 10% fetal calf serum, 2 mM L-glutamine, 50 mgiml kanamycin sulphate (and 1 mM sodium pyruvate for HSF and HPC); buffered at pH 7.2 with 0.5M tricine and 0.9% (w/v) sodium bicarbonate. Sterile ossicles (wet autoclaved) were presoaked in medium for 2 hr to ensure uniform wetting. One or two drops of cells in suspension (concentrations from 1 X 1O5to 5 X 106 cells/ml) were placed on each presoaked ossicle lying flat in a Falcon plastic Petri dish. The seeded ossicles were put in a high humidity incubator (37"C, 2 hr) to promote cell attachment; then enough medium was added to submerge them. In long term cultures the medium was changed weekly.

Electron microscopy

Cells grown on ossicles were fixed with glutaraldehyde-osmium tetroxide by standard methods. For SEM they were critical point-dried from liquid CO2 and sputter-coated with palladium-gold. For TEM, the decalcification method of Dietrich and Fontaine* w a s used before embedding in Epon 812.


Figure 3. Human skin fibroblasts, high density culture on a substrate of starfish ossicle shows growth in parallel arrays and multiple layers, 10 day culture. Bar, 75 pm.

RESULTS

All three cell lines grew vigorously for the 21 day duration of the culture trials. Judging by rate of increase, behavior, and adhesion, cells grew on ossicles at least as well as on glass or Falcon plastic. CHA (Fig. 5) showed an SEM morphology comparable to that described elsewhere.lo.ll Cell under- lapping eventually produced multiple-layering on the surface.12 Long processes from surface cells had penetrated throughout the stereome at least by day 3. CHA cell bodies had infiltrated the whole ossicle and established an interior population at least by day 7 (Fig. 6), but less vigorously than HSF.

Figure 4. Human prostate cancer cells with vigorous surface activity; 3 day culture. The ossicle surface (*), showing as etched prisms, was moderately roughened by the cleaning process. Bar, 9 pm.


Figure 5. Chinese hamster ovary fibroblasts, high density culture. Cells in multiple layers bridge a depression in the underlying ossicle; 13 day culture. Bar, 10 pm.

The SEM morphology of HSF (Fig. 3 ) is comparable to that of human fibroblasts of diverse origin^.'^^^^ HSF grew particularly vigorously. By day 10 extensive multiple-layering was evident, usually in crisscross arrays.' 2,1

Long, fine cell processes had penetrated much of the stereome within 24 hr. A substantial population of HSF was resident within the stereome at least by day 7. The interior populations of HSF and CHA increased throughout the trials but whether by proliferation of internal cells or by infiltration from without is unknown. Interior HSF showed the same surface features as ex- terior cells, but interior HSF were longer, thinner, and somewhat contorted

Figure 6. An ossicle seeded with Chinese hamster ovary fibroblasts has been fractured transversely. 'The superficial cell population is at the top; arrows point to fibroblasts that have infiltrated deeply; 7 day culture. Bar, 30 Pm.


Figure 7. Human skin fibroblast, the leading edge attached to a starfish ossicle, TEM. Microfilament bundles are obvious; some insert on probable attachment sites (arrows). Ruthenium red demonstrates the cell surface and also a fuzzy "ghost" that delineate< the surface of the decalcified ossicle. Bar, 2.0 pm.

as they followed the channels within the ossicle. TEM showed interior cells to be apparently healthy (Figs. 7 and 8).

The SEM morphology of HPC (Fig. 4) has not been previously described in the literature but the cells resemble some other transformed cell l i n e ~ . ~ ~ J ~ The surfaces of HPC were conspicuously active. Filopodia16 (probably retraction fibres in some instances) were particularly obvious. HPC showed characteristics of cultured epithelial rather than of mesench ymal cells.17 They retained a polygonal shape; never elongared as in CHA or HSF; did not extend ex- ploratory processes deep into the ossicle; did not attempt to infiltrate; and restricted themselves to forming a surface layer. Multiple layering occurred in older cultures.

Initially we were concerned to produce cleaned ossicles with trabecular surfaces as smooth and glasslike as possible on the assumption that roughened surfaces would be inimical to cell attachment, movement, and growth. The cleaning method described yields least surface etching consistent with scru- pulous elimination of native cells and organic debris. However, all cleaning methods tried (hypochlorite, NaOH and KOH, enzymes, detergents, ultraso-

Figure 8. The attached surface of the cell conforms closely to the rugosities of the starfish skeleton provoked by cleaning; trabecula (*) of decalcified ossicle. Bar, 2.6 pm.

Human skin fibroblast, TEM.


nication; singly and in combinations) produced surface erosion to some degree, as did even storage in sterile glacs-distilled water longer than 24 hr.

Prolonged etching by most agents typically reveals a pattern of pointed prisms (1.4 pm mean diameter, range 0.6-2.3) packed with their long axes in parallel throughout the entire ossicle (Fig. 9). Similar prisms have been seen in ossicles of juvenile sea urchins18 aiid in naturally eroded ossicles deposited in marine sediments.19 Prolonged hypochlorite cleaning gave a different etch pattern, round-ended prisms with similar dimensions and packing at the bottoms of deep, sharp-edged pits (Fig. 10). These etching patterns probably reflect an underlying microstructure based on polycrystalline aggregates.5

Our original concern apparently was mostly unfounded since all three lines grew well on trabecular surfaces moderately roughened by etching. In early cultures of CHA and HPC, before cell numbers obscured the substrate, cells appeared to be preferentially attached to slightly or moderately roughened rather than to smooth areas. This is supported by rough counts of the frequency of marginal and filopodial adhesions12 to the substratum. TEM showed that HSF cell surfaces in contact with the ossicle often conformed very closely to the sharp peak and valley texture of etched trabeculae (Fig. 8), sometimes associated with attachment sites (electron-dense plaques with microfilament bundles) (Fig. 7).20 Higher frequency of filopodia on etched rather than smooth surface suggests a substrate-exploring function and a possible mechanism for expression of substrate preference.16 However, there is a limit to cells’ tolerance of roughened surfaces. Severely eroded surfaces,

Figure 9. a pattern of pointed prisms. Bar, 10 pm,

Skeletal surface etched by prolonged exposure to KOH showing


Figure 10. o f surface-etching. Bar, 9 pm.

Sharp-edged pits containing prisms is another common pattern

particularly where there are obvious abrupt edges, are avoided rather than preferred. Any pattern of sharp-edged pits or grooves provoked by prolonged cleaning is conspicuously unfavorable. Cells apparently make a discrimina- tion between rough surface (acceptable) and sharp edge (barrier), possibly by a form of contact guidance or a response to substratum geometry imposed by restrictions of the locomotory mechanism.21

DISCUSSION

These results provide preliminary evidence for the biocompatibility of native echinoderm skeletal ossicles for mammalian cells cultured in 71itvo. As a minimum, the material deserves evaluation as a three-dimensional, porous substrate for behavioral studies of cells in culture. We have noted here observations pertinent to the influence o f substrate geometry,21,22 to contact guidance,21 to filopodial e ~ p l o r a t i o n , ~ , ' ~ to adhesivi ty,I2r2" and to growth in parallel arrays.I5 Incidentally the use of ossicles gives another technique for correlated SEM and TEM of cultured cells. The advantages for SEM are ease of handling the skeletal plates with their attached cells; for TEM the simplicity of removing the substrate by decalcification.8 Ossicles provide a simple alternative to coated coverslips, porous membranes or f i and some complex techniques presently in use.

The evidence for biocompatibility with mammalian cells indicates the potential of native echinoderm skeleton as a biomaterial in its own right. Its


microstructure, especially the ideal pore size (ca. 15-20 pm) for fibroblast ingrowth, has already attracted interest in echinoderm skeleton as a biomater- ia1.24 However, the native calcite was considered unsuitable for direct structural prosthetic applications because of its softness (Mohs scale = 3) and relatively high solubility. Instead its microstructure has been replicated in ceramic, metal, and polymer materials by a patented Replamineform process6 for testing as orthopaedicz5 or vascular Hydrothermal conver- sion of echinoderm calcite into a calcium phosphate mineral (whitlockite) potentially compatible with human hard tissues has been attempted.27

We suggest that the disadvantages of echinoderm skeleton for direct use as a structural prosthetic may be its virtues if i t is considered as a transient substrate that could stimulate the regeneration of cancellous bone. The use of echinoderm skeleton as a packing of small, loose ossicles or implanted as individual ossicles should stimualte fibroblast ingrowth, and thus initiate the sequence of events that leads to cancellous bone formation. We envisage the sequence as: infiltration by fibroblasts and osteoprogenitor mesenchyme, vascularization, dissolution of the ossicle, bony spicule formation. Preliminary trials with ossicles implanted as long bone plugs and beneath skull periosteum in rats support this conjecture. Similar events, at least initially, are characteristic of tissue ingrowth into porous biomaterials, though pore size is critical in determining the final r e ~ u l t . ~ ~ , ~ " Initial pore size, continuous porosity, and relative high solubility of calcite are the potentially advantageous features of echinoderm skeleton. If ossicle dissolution is rapid enough, the restrictions of initial pore size should disappear. Thus echinoderm skeleton may be a natural biodegradable ceramicz8 with potential use for hard tissue regeneration. As such, it deserves further evaluation.

We thank Alexa Kennedy, Lesley M'ood, and H. F. Dietrich for technical assistance, and N. Sherwood, G. Voss, and H. Van Netten for useful discussions and use of their facilities. The work was supported by a University of Victoria Faculty Research Grant.

References

1. K. M. Wilbur, "Recent Studies of Invertebrate Mineralization," in The Mech-anisrns of Minernliznfion in f he Inuertebrafes rind P I o n k , N. Watabe and K. M. Wilbur, Eds., Belle W. Baruch Library in Marine Science, No. 5, Univ. South Carolina Press, Columbia, 1976, pp. 79-108. K. M. Wilbur, "Mineral Regeneration in Echinoderms and Molluscs," in Hard Tissue Growth, Repuir rind Reminr~ralizntion, K . Elliott and D. Fitzsim- mons, Eds., Ciba Foundation Symposium 11, new series, Elsevier, Am- sterdam, 1973, pp. 7-33. D. Nichols and J . D. Currey, "The Secretion, Structure, and Strength of Echinoderm Calcite," in Cell Structurr, und its Intrrprrvtation, S. M. McGee- Russell and K. F. A. Ross, Eds., Edward Arnold, London, 1968, pp. 251- 261. J. Weber, R. Greer, R. Voight, E. White, and R. Roy, "Unusual Strength Properties of Echinoderm Calcite Related to Structure," 1. Illtrastrurt. Res.,

K. Okazaki and S. Inoui., "Crystal Property of the Larval Sea Urchin Spi- cule,'' Den. Grozotli Differ., 18, 413-434 (1976).

2.

3.

4.

26,355-366 (1969). 5.


6.

7 .

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

R . A. White, J. hr. Weber, and E. W. White, ”Replamineform: A New Process for Preparing Porous Ceramic, Metal, and Polymeric Prosthetic Materials,” Science, 176,922-924 (1972). C . Donnay and D. L. Pawson, ”X-ray Diffraction Studies of Echinoderm Plates,” Scicricc, 166, 1147-1150 (1969). H. F. Dirtrich and A. R. Fontaine, “A Decalcification Method for Ultra- structure of Echinoderm Tissues,” Stiiir: Trchriol., 50,351-354 (1975). J . W. Shdy, K . R. Porter, and T. C. Krueger, ”Motile Behavior and Topog- raphy of \hihole and Enuclrate Mammalian Cells on Modified Substrates,”

R . Wetzel, E. M. Kendig, G. M. Jones and K. K. Sanford, ”A Systematic Scanning Electron Microscope (SEM) Analysis of Mitotic Cell Populations in Monolayer Culture,” in Scariiiiug Electror: Microscopy/ IY78/ II, R. P. Recker and 0. Johari, Eds., Scanning Electron Microscopy Inc., AMF O’Hare, 1978,

R. W. Rubin and 1,. P. Everhart, ”The Effect of Cell-to-Cell Contact on the Surface Morphology of Chinese Hamster Ovary Cells,” I . Cell Biol., 57,

P. R. Bell, ”Locomotory Behavior, Contact Inhibition, and Pattern For- mation of 3T3 and Polyoma Virus-Transformed 3T3 Cells in Culture,” ]. Cell Biol., 74, 963-982 (1977). M. A. Conda, S . A. Aaronson, N . Ellmore, V. H. Zeve, and K. Nagashima, ”Ultrastructural Studies of Surface Features of Human Normal and Turnour Cells in Tissue Culture by Scanning and Transmission Electron Micros- copy,” ]. Nnil. Cauccr 117st., 56, 245-249 (1 976). K. R. Porter and V. G. Fonte, ”Observations on the Topography of Normal and Cancer Cells,” in Scairriirig Elccfroir Mrcroscopyi 2973, 0. Johari and F. Corvin, Eds., IIT Research Institute, Chicago, 1973, pp. 683-688. C. A. Erickson, ”Analysis of the Formation of Parallel Arrays by BHK Cells i i i Vitro,” €‘xp/. Ccil Rcs., 115: 303-315 (1978). G. Albrecht-Buehler, ”The Function of Filopodia in Spreading 3T3 Mouse Fibroblasts,” in Ccll Motility, Rook A, R. Goldman, T. Pollard and I . Rosen- baum, Eds., Cold Spring Harbor Conferences on Cell Proliferation, Cold Spring Harbor Laboratory, Cold Spring Harbor, 1976, Vol. 3, pp. 247- 265. J. Overton, “Response of Epithelial and Mesenchymal Cells to Culture on Basement Lamella Observed by Scanning Microscopy,” E x p l . Crll Res.,

J . S. Pearse and V. R. Pearse, ”Growth Zones in the Echinoid Skeleton,” Am. ZOO^., 15, 731-753 (1975). E. T. Alexandersson, “Etch Patterns on Calcareous Sediment Grains: Petrographic Evidence of Marine Dissolution o f Carbonate Minerals,” Sciericc., 189,47-48 (1975). R. D. Goldman, J. A. Schloss, and J. M . Starger, ”Organizational Changes of Actinlike Microfilarnents During Animal Cell Movement,” in Cell Motility, B C J U ~ A, G. Goldman, T. Pollard and J. Rosenbaum, Eds., Cold Spring Harbor Conferences on Cell Proliferation, 1976, Cold Spring Harbor Laboratory, Cold Spring Harbor, Vol. 3., pp. 217-245. G. A. Dunn and T. Ebendal, “Contact Guidance on Oriented Collagen Gels,” E x p l . C d l Rcs., 111, 475-479 (1978). Y. A. Rovenskp and I . L. Slavnaya, “Spreading of Fibroblast-like Cells on Grooved Surfaces,” Fx-pi. Cri l Res,, 84, 199-206 (1974). W. Nopanitaya, R. K. Charlton, R. L. Turchin, and J. W. Grisham, ”Ultra- structure of Cell Cultures on Polycarbonate Membranes,” Sfnirr Tcchrrol.,

J . N . Weber, E. W. White, and J. Lebiedzik, ”New Porous Biomaterials by Replication of Echinoderm Skeletal Microstructures,” Nnfirrc, 233,337-339 (1971).

Ex~7l. Crll Rcs., 105, 1-8 (1977).

pp. 1-9.

837-844 (1973).

105,313-323 (1977).

52,143-149 (1977).


25. R. T. Chiroff, R. A. White, E. W. White, J. N. Weber, and D. Roy, ”The Restoration of Articular Surfaces Overlying Replamineform Porous Biomaterials,” f. Bionrcd. Mater. Rrs., 11, 165-178 (1977). R. A. White, E. W. White, E. L. Hanson, R. F. Rohner, and W. R. Webb, “Preliminary Report: Evaluation of Tissue Ingrowth into Experimental Replamineform Vascular Prostheses,“ S u v g ~ ~ y , 79,229-232 (1976). D. M. Roy and S. K. Linnehan, ”Hydroxyapatite Formed from Coral Skeleton Carbonate by Hydrothermal Exchange,” Nature, 247, 220-222 (1974). H. U. Cameron, J . MacNab, and R. M. Pilliar, ”Evaluation of a Biodegrad- able Ceramic,” 1. Biomcd. Mafcr. Res., 11, 179-186 (1977). L. L. Hench, “Biomaterials,” Science, 208,826-831 (1980). M. Spector, S. L. Harmon, and A. Kreutner, ”Characteristics o f Tissue Growth into Proplast Polyethylene Implants in Bone,” 1. Biomcd. Mater. Res., 13,677-692 (1979).

26.

27.

28.

29. 30.

Received May 16,1980 Accepted July 11, 1980

biocompatibility of echinoderm skeleton with mammalian cells in vitro: preliminary evidence

Documents