Osteoclastic resorption of biomimetic calcium phosphate coatings in vitro

Osteoclastic resorption of biomimetic calcium phosphate coatings in vitro S. Leeuwenburgh,1 P. Layrolle,2 F. Barre`re2, J. de Bruijn,2 J. Schoonman,1 C.A. van Blitterswijk,2 K. de Groot2 1 Laboratory for Inorganic Chemistry, Delft University of Technology, Delft, The Netherlands 2 IsoTis BV, Prof. Bronkhorstlaan 10, 3723 MB Bilthoven, The Netherlands Received 12 June 2000; revised 25 January 2001; accepted 16 February 2001 Abstract: A new biomimetic method for coating metal implants enables the fast formation of dense and homogeneous calcium phosphate coatings. Titanium alloy (Ti6Al4V) disks were coated with a thin, carbonated, amorphous calcium phosphate (ACP) by immersion in a saturated solution of calcium, phosphate, magnesium, and carbonate. The ACPcoated disks then were processed further by incubation in calcium phosphate solutions to produce either crystalline carbonated apatite (CA) or octacalcium phosphate (OCP). The resorption behavior of these three biomimetic coatings was studied using osteoclast-enriched mouse bone-marrow cell cultures for 7 days. Cell-mediated degradation was ob-

served for both carbonated apatite and octacalcium phosphate coatings. Numerous resorption lacunae characteristic of osteoclastic resorption were found on carbonated apatite after cell culture. The results showed that carbonated apatite coatings are resorbed by osteoclasts in a manner consistent with normal osteoclastic resorption. Osteoclasts also degraded the octacalcium phosphate coatings but not by classical pit formation. © 2001 John Wiley & Sons, Inc. J Biomed Mater Res 56: 208–215, 2001

INTRODUCTION

(BMP) cannot be included into such a plasma-sprayed coating. Furthermore, these types of coatings have low resorption levels and are not easily degraded by cells. Finally, particle release and delamination are specific drawbacks for the plasma-spray technique.4 The crystallinity of plasma-sprayed coatings is not uniform as the coatings consist of crystalline and amorphous regions. If Ca-P material is released from these heterogeneous coatings, the resultant particles may initiate inflammation in surrounding tissues. Recently, research has focused on the development of new coating techniques that overcome the above-mentioned drawbacks. For instance, Ca-P coatings have been deposited onto metal implants using a biomimetic method.4–7 This biomimetic coating process involves the nucleation and growth of crystals on metal surfaces from calcium phosphate supersaturated solutions at ambient temperature. In this way a dense and homogeneous biomimetic coating can be formed with a uniform crystallinity and composition. Calcium phosphate bone substitutes generally are accepted to be biocompatible and osteoconductive when implanted in bone defects.8–10 Ideally, calcium phosphate biomaterials participate in the continuous process of bone remodeling. In such a case, a synthetic biomaterial is fully integrated into the human body.

In the past 20 years, there has been an increased interest in calcium phosphate (Ca-P) bioceramics because of their lack of adverse physiologic response and favorable osteoconductive properties. In order to improve long-term performance of orthopedic and dental implants, Ca-P coatings have been applied on metal prostheses. To date, the most successful method used to apply a calcium phosphate coating onto titanium alloys (Ti6Al4V) is the plasma-spraying technique.1 However, this technique has several drawbacks related to extremely high processing temperatures and line-of-sight application. For instance, phase changes in the Ca-P powder particles during the coating process are unpredictable because of the high temperature. Also, particularly promising phases such as carbonated apatite (CA), which is close to bone composition,2 and octacalcium phosphate (OCP), which may be involved in the early biomineralization process,3 cannot be deposited using plasma spraying. Biomolecules such as bone morphogenetic proteins Correspondence to: P. Layrolle; e-mail: [email protected] © 2001 John Wiley & Sons, Inc.

Key words: osteoclastic resorption; biomimetic; calcium phosphate; apatite; octacalcium phosphate

OSTEOCLASTIC RESORPTION OF BIOMIMETIC CA-P

Recently, Doi et al.11 investigated the osteoclastic response to various sintered calcium phosphates: hydroxyapatite (HA), ␣- and ␤-tricalcium phosphate (TCP), carbonate apatite (CA), tetracalcium phosphate (Te-CP), dicalcium phosphate dihydrate (DCPD), and octacalcium phosphate (OCP). Sintered carbonate apatite appeared to be the only substrate to be resorbed by osteoclasts in vitro. The evidence concerning osteoclast response to HA is controversial.12–15 De Bruijn et al.16 found that amorphous plasma-sprayed HA coatings are resorbed by osteoclasts whereas crystalline plasma-sprayed coatings do not reveal any characteristics of osteoclastic resorption. The crystal size of biomimetic coatings is smaller and the crystallinity is more comparable to bone mineral than to large and sintered hydroxyapatite particles produced by plasma spraying. This present study is based on the hypothesis that smaller Ca-P crystals would be resorbed by osteoclasts and that the physiologic CA crystals would exhibit the greatest resorbability. To test this hypothesis, the ability of osteoclasts to resorb three different biomimetic calcium phosphate coatings (amorphous calcium phosphate, crystalline carbonated apatite, and crystalline octacalcium phosphate) was examined in vitro.

MATERIALS AND METHODS Preparation of biomimetic coatings Biomimetic coatings were produced in a two-step procedure. First, the Ti6Al4V samples were soaked in a supersaturated Ca-P solution in order to deposit a thin, amorphous carbonated Ca-P film (ACP). This film acts as a seed surface for the subsequent growth of crystalline biomimetic Ca-P coatings. Second, these precoated samples were immersed in supersaturated Ca-P solutions, resulting in the formation of either carbonated apatite (CA) or octacalcium phosphate (OCP) coatings.

Formation of primary ACP coating Ti6Al4V disks (⭋ 20mm) were sandblasted using aluminum oxide particles. Ti6Al4V disks successively were cleaned ultrasonically in acetone, ethanol (70%), and demineralized water. The disks then were soaked for 24 h in a supersaturated Ca-P solution at 37°C (Table I), the pH of which was lowered to 6.0 ± 0.2 by dissolving CO2 until saturation. These mildly acidic conditions increase the solubility of the inorganic salts. The gradual release of CO2 from this solution and the consequent increase in pH enable the fast formation of biomimetic coatings in highly concentrated supersaturated Ca-P solutions. The ACP solution was prepared by dissolving reagent pure chemicals NaCl,

209

TABLE I Inorganic Composition (mM) of Human Blood Plasma (HBP), Simulated Body Fluid (SBF), and Supersaturated Ca-P Solutions for the Deposition of ACP, CA, and OCP Biomimetic Coatings

Na+ K+ Mg2+ Ca2+ Cl− HPO42− SO42− HCO3−

HBP

SBF

ACP

CA

OCP

142.0 5.0 1.5 2.5 103.0 1.0 0.5 27.0

142.0 5.0 1.5 2.5 147.8 1.0 0.5 4.2

733.5

733.5

140.4

7.5 12.5 720.0 5.0

1.5 12.5 720.0 5.0

3.1 142.9 1.86

21.0

10.0

HBP = human blood plasma; SBF = simulated body fluid; ACP = amorphous carbonated calcium phosphate; CA = carbonated apatite; OCP = octacalcium phosphate.

CaCl2 ⭈ 2H2O, MgCl2 ⭈ 6H2O, NaHCO3, and Na2HPO4 ⭈ 2H2O into demineralized water. The inorganic composition of the ACP solution is given in Table I.

Formation of biomimetic CA and OCP coatings CA coatings were deposited on top of the ACP-precoated Ti6Al4V samples in a highly concentrated Ca-P solution at 50°C for 3 h under the same mildly acidic conditions as described above. The inorganic composition of human blood plasma, conventional simulated body fluid (SBF), and the supersaturated Ca-P solutions used for the biomimetic CA and OCP coatings is given in Table I. The CA solution contained fewer inhibitors of crystal growth, such as magnesium and carbonate, than did the ACP solution. OCP coatings were obtained by immersing the pretreated samples in a buffered supersaturated Ca-P solution at 37°C for 3 h. NaCl, CaCl2 ⭈ 2H2O, and Na2HPO4 ⭈ 2H2O were dissolved into demineralized water. Crystal growth inhibitors (Mg2+ and HCO3−) were excluded from this solution while the ionic strength was kept constant by adding NaCl. The solution was buffered at physiologic pH of 7.40 at 37°C by adding a mixture of tris-hydroxymethylaminomethane (TRIS, 50 mM) and 1M of HCl. Finally, the solution was filtrated through a 0.22-␮m Millipore membrane.

Coating characterizations ACP, CA, and OCP coatings were characterized by X-ray diffraction (XRD, Rigaku, Miniflex). XRD patterns were collected within a range of 3° < 2␪ < 60°, with a step size of 0.02° and a scanning speed of 2.00°/min. Fourier transform infrared spectroscopy (FTIR, PerkinElmer, Spectrum 1000) was carried out in the range between 4000–400 cm−1. The different coatings also were observed by scanning electron microscopy (SEM; Jeol, JSM-35).

210

Osteoclastic mouse bone-marrow cell culture An osteoclast-enriched mouse bone-marrow cell culture was developed according to a modification of the method described by Gao and Yamaguchi.17 Young-adult nude mice, approximately 10–12 weeks old (strain: HsdCpb: NMRI-nu/+) were purchased from a certified supplier (Harlan Nederland B.V., Horst, The Netherlands). The nude mice were killed by cervical dislocation. Their femora were removed and dissected free of adhering tissues. The bone ends were cut off with scissors, and the marrow cavity was flushed with ␣-minimal essential medium (␣-MEM; Gibco Life Technologies, Breda, The Netherlands) by slowly injecting at one end of the bone using a sterile 25-G needle. The marrow cells from three or more animals were collected into a tube, centrifuged, and resuspended in 1 mL of ␣-MEM. Subsequently cells were seeded in each well of 12-well culture dishes (Nalge Nunc, Roskilde, Denmark) at a seeding density of 1.0 × 106 cells/mL (erythrocytes not included). The culture medium used was ␣-MEM supplemented with 10% fetal bovine serum (FBS; Gibco), 1% antibiotics (200 U/mL of penicillin G and 100 ␮g/mL of streptomycin) in the presence of freshly added 1␣,25-dihydroxyvitamin D3 (Sigma-Aldrich; Zwijndrecht, The Netherlands) at a concentration of 10 nM. 1␣,25-dihydroxyvitamin D3 is a resorbing agent well known to stimulate the formation of osteoclasts.18 Cells were cultured in a humidified atmosphere of 95% air, 5% CO2 at 37°C. Old medium was replaced after 3 days in culture by new medium with freshly added 1␣,25dihydroxyvitamin D3. After 7 days in culture, the cells were fixed overnight in a 37-wt % formaldehyde solution in water, stained for tartrate-resistant acid phosphatase (TRAP), and counterstained with methyl green.

Resorption assay of biomimetic calcium phosphate coatings Three biomimetic calcium thin films (ACP, CA, and OCP) were tested in vitro for their response to osteoclasts. Slices of dentine were used as a reference substrate material. The biomimetic coatings were applied to the sandblasted Ti6Al4V samples, as described above. A dentine block (elephant tooth) was cut into thin slices of approximately 180 ␮m using a diamond saw (Heathway, United Kingdom). The dentine slices and Ca-P-coated samples (n = 6) were sterilized in 70% ethanol, washed twice in phosphatebuffered saline (PBS), placed at the bottom of each well of a 12-well plate, and incubated in culture medium at 37°C. Mouse bone-marrow cells were seeded in each well at 1.0 × 106 cells/mL (2.0 mL/well) and cultured as described above. After 7 days in culture, the coated disks were washed three times with PBS and then treated with 0.001% pronase E (Sigma–Aldrich) and 0.02% EDTA for 10 min at 37°C. This treatment removed most of the nonadherent hemopoietic cells and stromal cells so that the osteoclasts could be observed in detail.19 The specimens then were fixed overnight in a 37-wt % formaldehyde solution, dehydrated through a graded ethanol series, and critical point dried from CO2 (Balzers, CPP 030). All samples were sputter-coated with

LEEUWENBURGH ET AL.

gold (Cressington Sputter Coated, 108A C1329) prior to examination with a scanning electron microscope (SEM; Jeol, model JSM-35). Degradation of calcium phosphates can be explained by the combined action of two processes20: solution-mediated physicochemical dissolution and cellmediated degradation as the result of the degrading activities of resorptive cells. In order to separate these two effects, some of the produced thin coatings were pre-incubated in ␣-MEM culture medium without cells for 7 days. In this way, differences in morphology between these reference coatings and the coatings that were used for the resorption assay could be attributed solely to the degradative cellular activities of resorptive cells.

RESULTS AND DISCUSSION Coating characterizations SEM pictures of the disks after preincubation in ␣-MEM for 7 days are shown in Figure 1. After immersion in ␣-MEM, the morphology of the coatings remained unchanged. After formation of a primary ACP layer, a thin amorphous coating was deposited onto the roughened Ti6Al4V substrate [Fig. 1(a)]. The roughness of the underlying substrate was clearly visible, indicating that the ACP film was very thin. The FTIR spectrum of the ACP layer [Fig. 2(a)] showed featureless phosphate and carbonate bands. Phosphate groups exhibited broad and single bands at 560 cm−1 [P-O deformation in PO4 (␯4)] and 1041 cm−1 [P-O stretching in PO4 and HPO4 (␯3)]. The carbonate bands also were observed at 1410 cm−1 and 1450 cm−1. The spectrum of the ACP thin film is characteristic of an amorphous calcium phosphate (ACP). The XRD surface pattern [Fig. 3(a)] showed a halo or bump located at approximately 2␪ = 30° characteristic for a Ca-P amorphous state. Biomimetic crystalline CA and OCP coatings [Fig. 1(b,c)] had a homogeneous morphology after 3 h of soaking, as observed by SEM. CA coatings consisted of small crystals with crystal sizes of approximately 100 nm, which is in the same range as the typical crystal size of bone mineral (20–100 nm). OCP is composed of larger flake-like crystals, with crystal sizes between 2 and 5 ␮m. Cracks were visible in the dense CA coatings as a result of the drying process after coating. The FTIR spectrum of CA [Fig. 2(b)] exhibited a broad 1041 cm−1 band with two shoulders at 1104 and 960 cm−1 [P-O stretching in PO4 and HPO4 (␯3)] and two quite sharp bands at 603 and 562 cm−1. Furthermore, the FTIR spectrum exhibited carbonate bands at 1494, 1476, and 1417 cm−1. The position of these carbonate bands indicated that carbonate groups replace


211

Figure 2. FTIR spectra of (A) amorphous carbonated calcium phosphate (ACP); (B) carbonated apatite (CA); and (C) octacalcium phosphate (OCP) coatings compared to (D) human bone.

of the CA coating [Fig. 3(b)]. The diffraction line at 2␪ = 25.9° corresponded to the overlapping of (2 0 1) and (0 0 2) plans and the diffraction line at 2␪ = 32.1° to the overlapping of (2 1 1), (1 1 2), and (3 0 0) plans. This overlapping of the diffraction lines is characteristic for tiny crystals where the crystallographic order is relevant over a short distance. The FTIR and XRD spectra of CA and bone [Figs. 2(b,d) and 3(b,d)] have many common features, indicating that the CA coating resembles the structure of bone mineral. The FTIR spectrum of OCP [Fig. 2(c)] exhibited sharp P-O bands at 1100, 1070, and 1023 cm−1 [P-O stretching in PO4 and HPO4 (␯3)] and HPO42− bands at 906 cm−1 and 852 cm−1 (P-O stretching in HPO4), characteristic for the OCP phase.23 The two sharp P-O bands at 560 cm−1 and 600 cm−1 [P-O deformation in PO4 (␯4)] exhibited shoulders at 624 cm−1 (H2O libation) and 526cm−1 [P-O twisting in HPO4 (␯4)]. These sharp bands indicated a highly crystallized OCP structure. The XRD pattern of this coating exhibited high, sharp diffraction lines at 2␪ = 4.7°, 2␪ = 16.8°, and 2␪ = 25.5°, corresponding respectively to (0 1 0), (−1 0 1),

Figure 1. SEM micrographs of (A) amorphous carbonated calcium phosphate (ACP); (B) carbonated apatite (CA); and (C) octacalcium phosphate coatings (OCP) preincubated in ␣-MEM culture medium for 7 days.

phosphate and hydroxyl groups in the hydroxyapatite lattice.21 Two main broad diffraction lines at 2␪ = 25.9° and at 2␪ = 32.1°, corresponding to an apatitic calcium phosphate phase,22 were observed in the XRD pattern

Figure 3. XRD spectra of (A) amorphous carbonated calcium phosphate (ACP); (B) carbonated apatite (CA); and (C) octacalcium phosphate (OCP) coatings compared to (D) human bone.

212

LEEUWENBURGH ET AL.

and (0 0 2) plans. The first diffraction line at 2␪ = 4.7° is typical of triclinic OCP crystals.24 This indicated that OCP was highly crystalline. However, the relative intensities of the OCP diffraction peaks presented mismatches with the typical OCP pattern, suggesting that the crystals had grown on the substrate following the favorite directions, (−1 0 1) and (0 0 2). The analyses have shown that reproducible, dense coatings can be obtained by means of a biomimetic method. Bone-like carbonated apatite (CA coatings) and octacalcium phosphate (OCP coatings) are easily deposited onto sandblasted Ti6Al4V substrates and consist of small crystals.

Mouse bone-marrow cell culture Addition of 1␣,25-dihydroxyvitamin D3 (10 nM) induced formation of numerous TRAP-positive multinucleated cells in mouse marrow cultured for 7 days (at near confluency). Staining intensity, size, shape, and nuclearity of the TRAP-positive cells differed considerably within the culture. Two types of TRAPpositive multinucleated cells were detected, one in which nuclei were located in the center of the cell [Fig. 4(a)] and a second in which nuclei were located in the periphery of the cell [Fig. 4(b)].

Resorption assay Figure 5 shows some typical SEM micrographs of the surfaces of the three calcium phosphate thin films after culturing osteoclast-enriched mouse bonemarrow cell cultures for 7 days on top of the coatings. The SEM micrograph of dentine, a natural resorbable substratum, also is included as a reference. The ACP thin film exhibited a very roughened morphology, a result of the roughness of the sandblasted Ti6Al4V substrate. Resorption lacunae characteristic of osteoclastic resorption were not detected on this very thin coating. It was almost impossible to distinguish resorption lacunae from substrate roughness [Fig. 5(a)]. In contrast, resorption of the homogeneous and flattened CA coating was very clear as characteristic scalloped resorption pits or lacunae remained [Fig. 5(b– d)]. The edges of the dome-shaped resorption pits were very sharp. The morphology of the CA coating in the center of the resorption lacunae showed some crystals whereas the edges of the resorption pit remained homogeneous and dense [Fig. 5(d)]. This difference in morphology can be attributed to the effect of acid dissolution below the ruffled border of the osteoclasts. This effect is more visible just beneath the ruffled border than it is in the zone where the osteo-

Figure 4. TRAP-positive multinucleated cells formed in mouse bone marrow cell cultures (original magnification ×200). Mouse marrow mononuclear cells were cultured in the presence of 10 nM of 1␣,25-dihydroxyvitamin D3 for 7 days. Cells then were fixed and stained for TRAP: (A) TRAPpositive multinucleated cells with the nuclei located in the center of the cell body; and (B) TRAP-positive multinucleated cells with the nuclei located in the periphery of the cell. Arrows point at nuclei.

clast attaches to the Ca-P-coating. Compared to the dense and homogeneous CA reference coatings, which were incubated in culture medium without cells for 7 days [Fig. 3(b)], the coating morphology of the CA coatings appeared uniformly affected as a result of the mouse marrow culture [Fig. 5(c)]. The OCP coating did not reveal any clear resorption pits, but the coating morphology clearly had been modified after culturing osteoclast-enriched mouse marrow cell cultures on top of the OCP coatings. Many coating irregularities and attacked areas were visible [Fig. 5(e,f)]. The resorption pattern of dentine was characteristic


213

Figure 5. Scanning electron micrographs (SEM) after 7 days of culturing mouse bone-marrow cells on biomimetic Ca-P coatings ACP, CA, and OCP: (A) roughened surface morphology of thin ACP coating; (B–D) osteoclastic resorption of CA coating with resorption lacunae (in C, note the attacked coating morphology); (D) at a higher magnification, the morphology of the dense resorption edge zone is clearly different, as a result of acid dissolution of calcium phosphate, from the open structure in the pit center; (E,F) attacked coating morphology of OCP coating; (G) osteoclastic resorption of dentine.

214

LEEUWENBURGH ET AL.

Figure 5. Continued.

for osteoclastic resorption, with numerous small resorption pits [Fig. 5(g)]. CA coatings were resorbed in the same way as dentine. Therefore, the pit-like resorption behavior of CA and dentine was indicative of osteoclastic resorption. On the contrary, no resorption lacunae could be observed for OCP and ACP coatings. ACP coatings seem to be too thin to enable detection of resorption pits, if there are any. OCP coatings clearly had been attacked by cells, but the resorption pattern was different compared to that of CA. The morphologic results of the present study clearly demonstrate that the biomimetic coatings are easily resorbed in vitro by osteoclasts. After 7 days of culturing, osteoclast-enriched mouse bone-marrow cell cultures, both biomimetic CA and OCP coatings, had an affected coating morphology. Moreover, numerous resorption lacunae were found at the surface of the CA coating, indicating osteoclast-like resorption. Thus the synthetic coatings were resorbed in the same way as naturally occurring mineral-like dentine. Consequently, a biomimetic coating could participate in the process of bone remodeling when implanted into the human body. This phenomenon corresponds to full integration into the human body of biomimetically coated implants. In contrast, conventional plasmasprayed coatings can delaminate and release large particles (±50 ␮m) that are not easily degraded by cells.16 Whereas numerous resorption lacunae were found on CA coatings, no pits were visible on the OCP coatings even though its coating morphology was attacked severely as a result of cell-mediated degradation. The most likely explanation could be the fact that CA coatings are composed of smaller crystals than are OCP coatings. This results in a dense and flattened CA coating compared to the more irregular OCP coating morphology. Logically, resorption pits are sharper and more clearly defined on such a smooth CA coating

than they are on irregular OCP coatings. Therefore it could be that the visible formation of sharply edged resorption lacunae is impossible on irregular OCP coatings even if osteoclasts actively resorb the coating. Another explanation for the absence of resorption pits on OCP coatings could be the fact that, in contrast to CA and dentine, OCP does not contain carbonate. It has been suggested11 that the presence of carbonate in calcium phosphates may play an important role in osteoclastic resorption. The underlying results in this study are in line with this suggestion since dentine and CA coatings were the only substrates that clearly revealed resorption pits. Although ACP contains carbonate, no resorption pits were detected since it was almost impossible to distinguish resorption lacunae from substrate roughness. Both CA and dentine contain carbonate, which is liberated into the sealed region during osteoclastic resorption. This released carbonate possibly might stimulate carbonic anhydrase enzymatic activity25 and thereby promote acid secretion by the osteoclasts. Some reprecipitation of carbonate containing apatite onto the surface of the OCP coating from the ␣-MEM could occur. Still, the amount of carbonate incorporated by reprecipitation is negligible with respect to the relatively high amounts of carbonate in dentine and CA coating. A third reason could be the higher solubility of OCP as compared to CA.26 Consequently the intracellular calcium levels probably will increase faster in osteoclasts resorbing OCP than in those resorbing CA. High intracellular calcium levels result in inhibition of osteoclastic bone resorption and osteoclastic detachment from bone surfaces.27 The resorption cycle of osteoclasts probably will be shorter for soluble calcium phosphates such as OCP than it will be for less soluble Ca-P ceramics such as CA. Therefore, one could assume that resorption of OCP would be a process of short resorption cycles whereas resorption of CA would be a continuous process resulting in long resorption cycles and large resorption lacunae. ACP is more soluble than crystalline OCP or CA. However, it is impossible to conclude anything about the resorption behavior of ACP from these experiments since it was almost impossible to distinguish resorption lacunae—if there were any—in the ACP coating from substrate roughness. The absence of resorption lacunae on OCP coatings as compared to the pit-like resorption process on CA coatings may be explained in several ways. The most likely explanation for this phenomenon is the fact that the surface morphology of irregular OCP coatings does not enable formation of sharp resorption lacunae although the coating can be attacked by osteoclasts. The absence of carbonate in OCP and the higher solubility of OCP coatings also could be important factors.


CONCLUSIONS A new biomimetic coating procedure enables the formation of dense and homogeneous Ca-P coatings. OCP and bone-like CA were deposited onto sandblasted Ti6Al4V disks. After preincubation for 7 days in ␣-MEM, neither release of particles nor morphologic changes could be observed for any of the biomimetic coatings. The coatings easily degraded in the presence of cells, as indicated by altered coating morphology, after culturing osteoclast-enriched mouse bone-marrow cell cultures for 7 days. Moreover, numerous resorption lacunae were observed on CA coatings, which is characteristic of osteoclastic resorption. Thus CA coating is degraded by osteoclasts in vitro and might participate in bone remodeling after implantation. Cells degraded OCP without exhibiting resorption lacunae. Resorption pits could not be observed on the thin ACP coating. The cell-mediated degradation of different biomimetic Ca-P coatings is related to parameters such as crystal size and chemical composition of the coating. The authors are grateful to John Tibbe and Peter-Paul Platenburg for their help with cell culture and to Martin Stigter for his assistance in preparing the biomimetic coatings. The dentine used in this study was kindly provided by Dr. Peter Nijweide from Leiden University, The Netherlands.

References 1.

de Groot K. Calcium phosphate coatings: Alternatives to plasma spray. Bioceramics 1998;11:41–43. 2. Rey C. Calcium phosphate biomaterials and bone mineral. Differences in composition, structure and properties. Biomaterials 1990;11:13–15. 3. Brown WE. Octacalcium phosphate and hydroxyapatite. Nature 1962;196:1048–1055. 4. Layrolle P, de Groot K, van Blitterswijk, CA. Biomimetic hydroxyapatite coating on Ti6Al4V induced by pre-calcification. Bioceramics 1998;11:465–468. 5. Barre`re F, Layrolle P, van Blitterswijk CA, de Groot K. Fast deposition of apatite coatings on implants. Mater Res Soc Symp Proc 2000;599:135–140. 6. Kokubo T, Kushitani H, Abe Y, Yamamuro T. Apatite coating on various substrates in simulated body fluids. Bioceramics 1990;2:235–242. 7. Barre`re F, Layrolle P, van Blitterswijk CA, de Groot K. Biomimetic calcium phosphate coatings on Ti6Al4V: A crystal growth study of octacalcium phosphate and inhibition by Mg2+ and HCO3−. Bone 1999;25:107S–111S. 8. Jarcho M. Biological aspects of calcium phosphates. Properties and applications. Dent Clin N Am 1986;30:25–47.

215

9. de Groot K. Macropore tissue ingrowth: A quantitative and qualitative study on hydroxyapatite ceramics. Biomaterials 1986;7:137–143. 10. Muller-Mai CM, Voigt C, Gross U. Incorporation and degradation of hydroxyapatite implants of different surface roughness and surface structure in bone. Scanning Microsc 1991;4: 501–511. 11. Doi Y, Iwanaga H, Shibutani T, Moriwaki Y, Iwayama Y. Osteoclastic responses to various calcium phosphates in cell cultures. J Biomed Mater Res 1999;47:424–433. 12. de Bruijn JD, Bovell YP, Davies JE, van Blitterswijk CA. Osteoclastic resorption of calcium phosphates is potentiated in postosteogenic culture conditions. J Biomed Mater Res 1994;28: 105–112. 13. Gomi K, Lowenberg B, Shapiro G, Davies JE. Resorption of sintered synthetic hydroxyapatite by osteoclasts in vitro. Biomaterials 1993;2:91–96. 14. Gilles JA, Carnes DL, Windeler AS. Development of an in vitro culture system for the study of osteoclast activity and function. J Endodont 1994;20:327–331. 15. Shimizu H, Sakamoto S, Sakamato M, Lee DD. The effect of substrate composition and condition on resorption by isolated osteoclasts. Bone and Miner 1989;6:261–275. 16. de Bruijn JD, Flach JS, Leenders H, van den Brink J, van Blitterswijk CA. Degradation and interface characteristics of plasma-sprayed hydroxyapatite coatings with different crystallinities. Bioceramics 1992;5:291–298. 17. Gao YH, Yamaguchi M. Inhibitory effect of genistein on osteoclast-like cell formation in mouse marrow cultures. Biochem Pharmacol 1999;58:767–772. 18. Blair HC. How the osteoclast degrades bone. BioEssays 1998; 20:837–846. 19. Tezuka K, Sato T, Kamioka H, Nijweide PJ, Tanaka K, Matsuo T, Ohta M, Kurihara N, Hakeda Y, Kumegawa M. Identification of osteopontin in isolated rabbit osteoclasts. Biochem Biophys Res Comm 1992;186:911–917. 20. de Bruijn JD, Davies JE, Klein CPAT, de Groot K, van Blitterswijk CA. Biological responses to calcium phosphate ceramics. In: Ducheyne P, Kokubo T, van Blitterswijk CA, editors. Bonebonding biomaterials. Leiderdorp, The Netherlands: Read Healthcare Communications; 1992. p 57–72. 21. Heughebaert JC. Contribution a l’e´tude de l’evolution des orthophosphates de calcium precipites amorphes en orthophosphates apatitiques [These d’Etat INP]. Toulouse, France; 1977. 22. Joint comity for powder diffraction standards: Hydroxylapatite 9-432; 1992. 23. Joint comity for powder diffraction standards: Octacalcium phosphate 26-1056; 1992. 24. Fowler BO, Markovic M, Brown WE. Octacalcium phosphate. III. Infrared and Raman vibrational spectra. Chem Mater 1993; 15:1417–1423. 25. Anderson RE, Jee WSS, Woodbury DM. Stimulation of carbonic anhydrase in osteoclasts by parathyroid hormone. Calcif Tissue Int 1985;37:646–650. 26. Zhang J, Nancollas GH. J Phys Chem 1992;96:5478–5483. 27. Yamada S, Heymann D, Bouler J-M, Daculsi G. Osteoclastic resorption of calcium phosphate ceramics with different hydroxyapatite/␤-tricalcium phosphate ratios. Biomaterials 1997; 18:1037–1041.