In vitro Evaluation of Potential Calcium Phosphate

0 downloads 0 Views 587KB Size Report
Biphasic calcium phosphate (BCP) consists of a mix- ture of HA and -tricalcium phosphate ( -TCP).7 Due to the higher solubility of the -TCP component, the reac-.
TISSUE ENGINEERING Volume 12, Number 2, 2006 © Mary Ann Liebert, Inc.

In vitro Evaluation of Potential Calcium Phosphate Scaffolds for Tissue Engineering SANDRA SÁNCHEZ-SALCEDO,1 Dr. ISABEL IZQUIERDO-BARBA,1 Dr. DANIEL ARCOS,1,2 and Prof. MARIA VALLET-REGÍ1

ABSTRACT Nanocrystalline calcium phosphates are very interesting candidates as scaffolds for bone tissue engineering. These materials show excellent in vivo biocompatibility, cell proliferation, and resorption. In this work we have studied the osteoblast-like cell behavior seeded onto HA and BCP synthesized by controlled crystallization method and treated at different temperatures. In vitro cell attachment, proliferation, differentiation, spreading, and cytotoxicity tests have been carried out. The results can be explained as a function of the phase composition and microstructure. Under in vitro closed conditions, nanocrystalline HA depletes the calcium of the medium avoiding cell proliferation, whereas well-crystallized HA enhances high cell proliferation. On the other hand, nanocrystalline BCPs supply Ca2 to the medium due to the higher solubility of the -TCP component, allowing an excellent in vitro cellular response when osteoblast-like cells are seeded on it. These features make BCPs excellent candidates as scaffolds for bone tissue engineering.

INTRODUCTION

H

(HA) is one the most important calcium phosphate-based bioceramics used for bone filling and replacement.1,2 This compound shows a chemical composition and structure similar to the mineral phase of bone and teeth and has been considered for coating on metallic implants,3 porous ceramic that facilitates bone ingrowths,4 inorganic component in a ceramic-polymer composite,5 and granulate to fill small bone defects.6 Biphasic calcium phosphate (BCP) consists of a mixture of HA and -tricalcium phosphate (-TCP).7 Due to the higher solubility of the -TCP component, the reactivity increases with the -TCP/HA ratio. Therefore, the bioreactivity of these compounds can be controlled through the phase composition.8–10 Currently, BCP bioceramics are recommended for use as alternatives or additives to autogenous bone for orthopedic and dental applications. YDROXYAPATITE

The bioactive behavior and excellent biocompatibility make both HA and BCP very useful materials for bone tissue replacement.11–16 Nowadays, new perspectives in the field of bioactive materials have led to a shift in emphasis from the replacement of tissues to regeneration of tissues. For this purpose, tissue engineering is an excellent alternative, where repair is initiated in vitro and then implanted in the patient.17 Tissue-engineering techniques generally require the use of scaffolds, which serve as three-dimensional templates for initial cell attachment and subsequent tissue formation.18,19 In the case of bone regeneration, it may be desirable for calcium phosphate materials to degrade over time because they lose strength as they age. The ability of the scaffold to be metabolized by the body allows it to be gradually replaced by new cells to form functional bone tissue. For this reason, the synthesis of HA and BCP as nanocrystalline materials is a very attractive idea for application as degradable scaffolds.

1Departamento 2Centro

de Química Inorgánica y Bioinorgánica, Facultad de Farmacia, Universidad Complutense, Madrid, Spain. Nacional de Investigaciones Metal£rgicas, CENIM-CSIC, Madrid, Spain.

279

SÁNCHEZ-SALCEDO ET AL.

280

There are several methods for synthesizing HA.20–22 Among them, the controlled crystallization method allows this compound to be obtained as a nanocrystalline material.23 Subsequent thermal treatments lead to higher degrees of crystallinity, which can be controlled as a function of the temperature. Similarly, BCP materials can be obtained by controlling the Ca/P ratio of the precursor solutions. The aim of this work is to evaluate the in vitro osteoblastic cell response in a closed system, when the cells are in direct contact with calcium phosphate ceramics of different composition and microstructure. Since a critical aspect of scaffold design is to determine the appropriate composition, microstructure, porosity, and surface reactivity to allow a good in vitro cell proliferation, a wide study comprising these subjects has been made.

MATERIALS AND METHODS Synthesis of HA and BCPs Stoichiometric hydroxyapatite and two different calcium deficient apatites (CDA1 and CDA2) powders were prepared by aqueous precipitation reaction of Ca(NO3)2  4H2O and (NH4)2HPO4 (Aldrich, Steinheim, Germany), which were dropped simultaneously, leading to the following reaction: 10-x Ca(NO3)2  4H2O  6 (NH4)2HPO4  8 NH4OH  Ca10-x(HPO4)x(PO4)6-x(OH)2-X  20 NH4NO3  6 H2O (Eq. 1) with x  0, 0.58 and 0.73 for HA, CDA1, and CDA2, respectively. For these syntheses, 0.6 M solutions of (NH4)2HPO4 reacted with Ca(NO3)2  4H2O solutions of different concentration under air atmosphere. Table 1 presents the synthesis parameters used to obtain the “as-precipitated” powders. The as-precipitated powders were aged in the aqueous media for 15 h, washed with 4 L of hot deionized water, and liophilized, obtaining dried dispersed powders. The powders were pressed into disks with uniaxial pressure at 50 MPa followed by isostatic pressure at 250 MPa. In order to test the influence of the

TABLE 1.

CONDITIONS

FOR THE

SYNTHESIS

Sample

[Ca2] solution (M)

[HPO42] solution (M)

HA CDA1 CDA2

1 0.95 0.92

0.6 0.6 0.6

microstructure on the cellular behavior, the stoichiometric as precipitated HA was treated at three different temperatures: 700, 900, and 1100°C for 1 h. These samples are referred to in this paper as HA-700, HA-900, and HA-1100, respectively. In addition, CDA1 and CDA2 samples were treated at 900°C for 1 h, giving rise to samples BCP-45 and BCP-27, respectively. Infrared spectroscopy (FTIR) was carried out with a Nicollet Nexus spectrometer from 500 to 4000 cm1 using the KBr technique and operating in the transmittance mode. Powder X-ray diffraction (XRD) experiments were made with a Philips X’Pert diffractometer (Eindoven, The Netherlands), by using Cu K radiation. In order to determine the quantitative phase composition and microstructure of the samples, the XRD patterns were refined by the Rietveld method using FullProf software (shareware, see ref.).24 Previously reported structural data for HA25 and -TCP26 were used as the initial model for Rietveld refinements. Textural properties were determined by N2 adsorption porosimetry in a Micromeritics ASAP2010 analyzer (Norcross, GA). Surface morphology was studied by scanning electron microscopy (SEM) in a JEOL 6400 microscope (Tokyo, Japan). The mean grain size was calculated using the intercept method.27 Twenty straight lines were drawn on each micrograph with different directions. More than 300 grains were considered for each sample. The chemical composition was obtained by EDX spectroscopy during the surface observation.

Biocompatibility assays Before carrying out the cell culture assays, the samples were sterilized by heating at 180°C for 24 h. A human osteoblast-like cell line denoted HOS was used. This cell line, obtained through the European Collection of Cell Cultures (ECACC, no. 87070202), was derived from an osteosarcoma of an old Caucasian female. The cells were cultured in Dulbecco’s modified Eagle medium (DMEM) containing 2 mM glutamine, 100 U/mL penicillin, 100 g/mL streptomycin, and 10% fetal calf serum (FCS) at 37°C in a humidified atmosphere of 95% air and 5% CO2. Osteoblast-like cells were routinely subcultured by trypsinization.

OF THE

AS-PRECIPITATED HA, CDA1,

AND

CDA2

pH

Reaction temperature (°C)

Reaction time (h)

9 8.5 7.1

90 60 50

22 2 4

EVALUATION OF POTENTIAL CALCIUM PHOSPHATE SCAFFOLDS TABLE 2. PHASE COMPOSITION AND AVERAGE CRYSTALLITE SIZE OF HA AND BCP MATERIALS CALCULATED FROM THE RIETVELD REFINEMENT

HA-700 HA-900 HA-1100 BCP-45 BCP-27

HA (% wt)

-TCP (% wt)

Crystal size (nm)

100 100 100 45.5 26.9

0 0 0 54.5 73.1

37 107 263 104 101

Cell attachment kinetics. The cells were resuspended in complete medium supplemented with ascorbic acid (50 g/mL) and -glycerophosphate (10 mM). Cells were seeded at a seeding density of 25  103 cell/cm2 directly onto the samples, and incubated for 1, 3, and 6 h under standard conditions. Unattached cells were removed by decantation of the culture medium. The cells attached to the substrate surface were rinsed four times with phosphate buffered saline (PBS) and fixed with an acetonemethanol (1:1) mixture. Afterwards, fixed cells were rinsed with distilled water and dyed with 1 mL of crystal violet solution in 20% of methanol. The crystal violet fixed onto cells was released with 1 mL of 1% (w/v) sodium dodecylsulphate (SDS) and quantified by means of UV-VIS spectrometry at 570 nm (Unicam UV 500). Cell-spreading assay. The spreading degree and morphology of the osteoblast-like cells were visualized by SEM after 6 and 24 h. The attached cells were rinsed four times in PBS and fixed with 2.5% (v/v) glutaraldehyde

281

in 0.1 M phosphate buffer. Dehydration was performed with slow water replacement by a series of graded ethanol solutions with final dehydration in absolute ethanol before critical-point drying. The materials were mounted on stubs, gold plated in vacuum using a sputter coater (Balzers SCD 004 (Wiesbaden-Nordenstadt, Germany), and analyzed by SEM. Cell proliferation assay. For this particular assay, the cells were seeded onto the materials surface in 24-well plates at a seeding density of 5  103 cell/cm2 in supplemented complete medium and incubated under standard conditions. Cell proliferation determinations were performed by using the MTT (3-[4,5-dimethylthiazol-2yl]-2,5-diphenyltetrazolium bromide) assay at 1, 3, and 7 days after seeding. Cell differentiation assay: Alkaline phosphatase (ALP) activity. ALP activity was considered for assessing the expression of the osteoblast-like phenotype. ALP activity was measured using the Reddi and Huggins method28 based on the hydrolysis of p-nitrophenylphosphate to p-nitrophenol. HOS cells were seeded at 15  103 cell/cm3 directly onto samples in 24-well plates and incubated in standard conditions with supplemented complete culture medium. ALP activity was determined at 7, 15, and 21 days with a commercial kit (Spinreact, Sant Esteve de Bas, Spain). Cytotoxicity assay: Lactate deshidrogenase (LDH) activity. LDH activity released from the osteoblast-like cells was considered for cell injury measurement. The measurements were made at 7 days of seeding by using a commercially available kit (Spinreact).

FIG. 1. Plots of the average crystallite shape of HA-700, HA-900, and HA-1100. Projection within the plane formed by [100]* and [001]* directions.

SÁNCHEZ-SALCEDO ET AL.

282

Determination of calcium ions content in the culture medium In order to study the ionic exchange between the samples and the substrates, nonseeded substrates were soaked in culture medium. The calcium ion concentration was measured as a function of soaking time at 37°C by means of an ion selective electrode technique (Ilyte).

Statistics The results are expressed as the mean  standard deviation of 6 specimens split in two independent experiments (3 separate groups of samples in each one). The value for each specimen was the mean value of 4 different aliquots. The differences between the mean values across groups were analyzed by ANOVA test with a minimal significance of p  0.05.

RESULTS The phase composition and crystallite size of the samples were calculated by the XRD Rietveld analysis and are presented in Table 2. In the Rietveld refinement per-

FIG. 2.

formed with FullProf, we can reconstruct an average crystallite shape using the apparent sizes along the different directions. Figure 1 shows the average shape of the crystallites calculated in this way. The plots are displayed along [100]* and [001]* directions of the HA lattice structure. It can be seen that the coherent diffraction domains are shortened within the (001) planes. This effect is higher for HA-700 and decreases for HA-900 and HA-1100. The analysis of HA and -TCP phases in BCP samples showed very similar characteristics to HA-900. The samples surfaces were studied by FTIR spectroscopy (Fig. 2). The FTIR spectra obtained for the surface of HA-900 shows the characteristic absorption bands corresponding to a hydroxyapatite phase. The bands at 3570 and 631 cm1 can be assigned to the stretching and librational modes, respectively, of the hydroxyl groups. The intense bands at 1089, 1054, and 961 cm1 correspond to the stretching vibration mode of P-O, whereas the doublet at 603 and 567 cm1 is assigned to the bending mode of the same bond. No bands corresponding to residual reagents, such as nitrates, were detected in the spectra. Samples HA-700 and HA-1100 showed very similar spectra. The spectrum collected for the surface of BCP-45 shows the bands corresponding to a -TCP sam-

FTIR spectra obtained for the surface of the samples. HA-700 and HA-1100 show very similar spectra than HA-900.

EVALUATION OF POTENTIAL CALCIUM PHOSPHATE SCAFFOLDS ple, in addition to the bands assigned to a hydroxyapatite phase. In this case the absorption bands of the hydroxyapatite phase are less intense than in HA-900, particularly the bands assigned to the hydroxyl groups. The spectra obtained for the surface of BCP-27 shows the same bands of BCP-45, but the bands corresponding to -TCP are more intense. On the contrary, the bands corresponding to O-H bonds are even less intense than in BCP-45. Figure 3 shows the scanning electron micrographs obtained for HA-700, HA-900, and HA-1100. HA-700 and HA-900 show the characteristic surfaces of nonsintered materials. The surfaces are formed by small rounded grains of 0.2 m in size, together with some aggregates that leave intergranular pores. Samples BCP-45 and BCP-27 presented very similar surfaces to HA-900 under SEM observations. The micrograph corresponding to HA-1100 shows larger grain size (between 0.5 and 1 m) compared with the others. Besides, the grains have polygonal shapes. This observation points out that the thermal

283

treatment at 1100°C is enough to start up the sintering process. The EDX spectra (Fig. 3, insets) were obtained during the SEM observation. These spectra show that the chemical composition at the surface of the samples involves only Ca, P, and O atoms. The atomic percentages of each sample (Table 3) were very similar to the theoretical ones. Table 4 presents the surface area and pore volume data obtained by N2 adsorption porosimetry. The textural parameters decrease as a function of the thermal treatment for HA samples. This fact is in agreement with the surface changes observed by SEM, as well as with the crystal size increase determined by XRD. The BCP samples (both treated at 900°C) show almost identical textural parameters to HA-900, pointing out that the presence of -TCP is not critical for the textural features of these samples. Figure 4 depicts the results obtained from the biocompatibility assays when the osteoblasts are seeded onto the samples’ surfaces. Cell attachment results (Fig. 4A)

FIG. 3. SEM micrographs of HAs samples treated at different temperatures. BCP samples (treated at 900°C) show very similar surfaces to HA-900.

SÁNCHEZ-SALCEDO ET AL.

284 TABLE 3.

CHEMICAL COMPOSITION

AND

EXPERIMENTAL CA/P RATIO CALCULATED P (% atoms)

O (% atoms)

Ca/P

21.64 21.40 22.83 22.08 19.59

12.99 12.72 13.57 13.74 13.03

65.36 65.88 63.61 64.18 67.39

1.67 1.68 1.68 1.61 1.50

indicate that the osteoblasts attach to the substrates in all cases. Sample HA-700 exhibits significantly lower adhesion values (p  0.05) than the other substrates, although this adhesion is always higher than in the plastic control (p  0.05). Cell proliferation tests (Fig. 4B) show that, under the studied in vitro and closed conditions, osteoblast-like cells proliferate onto HA-1100 and BCPs samples, but not onto HA-700 and HA-900. In order to measure the osteoblastic phenotype expression over the different samples, ALP tests were carried out after 7, 15, and 21 days. Figure 4C shows the data obtained at 7 days, which remained constant at 15 and 21 days. These results show higher ALP levels in samples HA-1100, BCP-45, and BCP-27. When comparing HA-700 with HA-900, significantly higher ALP values (p  0.05) were found for HA-900. The cytotoxicity of the samples was determined by means of LDH tests after one week of culture (Fig. 4D). In order to estimate the number of injured cells for each sample, 10,000 cells were quenched from room temperature to 84°C and the LDH level was measured. This test showed LDH values of 15 U/L, which can be considered an estimation of the LDH amount released to the medium when 10,000 cells are destroyed. LDH level in sample HA-700 is almost negligible. This sample does not lead to cytolysis, as could be assumed from the proliferation results. The cells do not proliferate, but they are not destroyed either. This fact indicates that osteoblasts seeded onto this sample underwent a cytostatic effect, but we cannot state that the lack of proliferation is due to the cytotoxicity of the substrate. TABLE 4. TEXTURAL PROPERTIES OF THE HYDROXYAPATITES TREATED AT DIFFERENT TEMPERATURES AND BIPHASIC MATERIALS

HA-700 HA-900 HA-1100 BCP-45 BCP-27

EDX SPECTROSCOPY

Ca (% atoms) HA-700 HA-900 HA-1100 BCP-45 BCP-27

Sample

BY

SBET (m2/g)

Pore vol. (cm3)

22.73 5.08 1.71 5.05 4.15

0.285 0.016 0.007 0.022 0.014

HA-900 showed a slightly higher LDH level, whereas HA-1100 showed a level similar to that produced by 10,000 injured cells. LDH levels for BCP samples are significantly lower than for HA-1100. Figure 5 shows the SEM micrographs after 6 h of culture. Differences of cell spreading and morphology can be observed for the different substrates. HA-700 samples show few round, attached cells with few signs of spreading. In samples HA-900 and HA-1100 (Fig. 5B and C), some flattened cells with leading edges can be observed. The surfaces of BCP-45 and BCP-27 (Fig. 5D and E) appear highly colonized by well-flattened cells with numerous filopodia and lamellipodia, showing connected anchoring elements. Figure 6 shows the micrographs after 24 h of culture. HA-700 (Fig. 6A) shows a noncolonized surface and only a few rounded and isolated cells can be observed. The situation in HA-900 is similar after 24 h to that at 6 h. Only a few spread cells can be observed at the surface, pointing out that proliferation activity did not start after this period. HA-1100 (Fig. 6C) and BCPs samples (Fig. 6D and 6E) show a very different scenario. The surfaces are almost covered by extended cells. A layer of cells is almost fully developed onto the surface, indicating that proliferation has occurred after 24 h of culture. In order to determine the effect of the samples on the environment, the Ca2 content of the culture medium was determined as a function of time, with nonseeded substrates (without cells) soaked in complete medium. Figure 7 shows the results obtained with this test. It can be seen that the Ca2 content almost disappears from the medium in contact with HA-700 after 24 h. Similarly, 80% of the Ca2 content is removed from the culture medium when HA-900 is soaked for 24 h. Afterward, the Ca2 content slowly decreases, resulting in a Ca2 depleted medium at the end of the test. HA-1100 shows a different behavior. After being soaked for 24 h, only 20% of Ca2 is removed from the medium, keeping between 80 and 40% of the initial amount throughout the test. BCP samples show an intermediate behavior between HA-700/HA-900 and HA-1100 samples. The calcium content decreases 60 and 50% after 24 h for BCP-45 and BCP-27, respectively, that is, the higher the HA content, the higher the Ca2 decrease. However, the Ca2 con-

EVALUATION OF POTENTIAL CALCIUM PHOSPHATE SCAFFOLDS

285

FIG. 4. Biocompatibility cell culture tests. (A) Cell attachment at 1, 3, and 6 h of culture. *Significantly different in respect to the rest of the series (p  0.05). (B) Cell proliferation at 1, 3, and 7 h of culture. *Significantly different in respect to the rest of the series (p  0.05). (C) ALP relative level at 7 days of culture. *The two series are significantly different (p  0.05). (D) LDH activity at 7 days of culture. Bar labeled “test” corresponds to the LDH measured after lysing 10,000 cells. *The three series are significantly different (p  0.05). §The two series are significantly different in respect to HA-1100.

centrations remain constant after 3 days and increase slightly from 3 days to the end of the test. This fact can be due to the partial dissolution of the more soluble -TCP from the substrate to the environment.

DISCUSSION This work describes the relationship between the chemical, textural, and microstructural characteristics of calcium phosphates and their in vitro behavior. For this purpose we have synthesized different stoichiometric HA by changing the subsequent thermal treatment. In this way, we can compare CaP with the same chemical composition, having different textural and microstructural features (HA-700, HA-900, and HA-1100). On the other hand, we have also obtained CaP with very similar microstructural and surface properties, having different phase compositions by changing the initial Ca/P ratio (HA-900, BCP-45, and BCP-27). The FTIR spectra, EDX spectroscopy, and quantitative phase analysis indicate

that the phases and chemical composition can be easily controlled with the synthesis method proposed in this work. The lower the Ca/P ratio the higher the -TCP content after a thermal treatment of 900°C. On the other hand, increasing the thermal treatment leads to larger crystallites and lower surface area in stoichiometric HA. The detailed microstructural study of our samples shows that not only the size but also the average morphology of the coherent diffraction domains can be modified by the thermal treatment. The microstructure analysis indicates that the growth perpendicular to the [001] direction is more difficult at lower temperatures, resulting in more needle-shaped crystallites. HA-700 is a nanocrystalline compound with an average crystallite size of around 37 nm. The crystallites are needle shaped, thus very similar to the biological apatite of bone. HA-900 shows an average crystallite size three times higher than HA-700. The crystallites grow more along the [100] direction with the thermal treatment, partially loosing the needle-shape morphology. HA-1100 shows a microstructure formed by larger crystallites (263 nm), with

286

FIG. 5. SEM micrographs of 6 h cultured surfaces. Cell spreading differences between HA-700 and the other samples can be clearly observed. BCPs samples show the higher spreading degree at this point.

SÁNCHEZ-SALCEDO ET AL. an almost isotropic growth along the different directions. These results show that not only the crystal size, but also the crystal morphology can by partially tailored with this synthesis method. By controlling all the features described above, we can evaluate the behavior of these compounds as potential scaffolds for bone tissue engineering. Osteoblast/material interaction depends on the characteristics of the material surface. In the first stage, cells in contact with a surface will attach, adhere, and spread. Thereafter, the quality of this adhesion will influence their ability to proliferate and differentiate. This first phase depends on the adsorption of adhesion proteins, such as vitronectin and fibronectin.29 In this work, the osteoblast-like cells attach all the samples better than the plastic control. This result is in agreement with the strong affinity of vitronectin with ceramic substrates,30 although subsequent cell proliferation is much lower in HA-700 and HA-900 than in the control. Additionally, it is observed that cell attachment and spreading are lower in HA-700 compared to the other samples. This fact could be due to the weak cohesion between surface particles, leading to grain debridement. Samples HA-900, HA-1100, BCP-45, and BCP-27 show similar cell attachment between them, and they are twice higher than in HA-700. These data point out that the higher intergranular cohesion (clearly reflected in the decrease of textural parameters) promotes the initial cell attachment. The cell proliferation results show that the proliferation activity of the cells seeded onto HA-700 is very low from the first day, and it is null after 7 days. Cells seeded onto HA-900 show slightly higher values, but the proliferation is also very low. SEM micrographs evidenced that osteoblast-like cells did not colonize the HA-900 surface. On the contrary, HA-1100 shows high cell proliferation levels, which is in agreement with the fully covered surface observed by SEM. In general, the lack of cell proliferation can be due to several factors: lack of cell attachment, changes on environment composition, material cytotoxicity, etc. Lower initial cell attachment was observed for HA-700. This fact would have a negative influence on the subsequent cell proliferation. It should be noted that cells highly proliferate on plastic control. Since HA-700 and HA-900 showed better cell attachment than plastic control; other factors must influence the cell proliferation in these substrates. Several authors have suggested that the low osteoblast-like cell proliferation can be attributed to ingestion of HA particles.31,32 The presence of HA particles and its intracellular solubilization were hypothesized to adversely affect calcium and phosphate homoeostatic mechanisms and to modify the mechanical regulators of DNA synthesis without any expression of cytotoxic effect.33 The behavior of the cells seeded on HA-700 and HA-900 fit into these hypotheses. The small crystallite

EVALUATION OF POTENTIAL CALCIUM PHOSPHATE SCAFFOLDS

FIG. 6. SEM micrographs of the different CaP showing the osteoblast-like cell extension after 24 h.

287

and grain sizes would facilitate the ingestion of particles decreasing the cell proliferation. This effect is also reflected in the significantly lower ALP production in HA-700. Kozawa et al. have reported on the decrease of ALP activity attributed to the phagocytosis of HA particulates by the cells.34 Ong et al. observed a significantly lower ALP activity and 1.25 (OH2) vitamin D3-stimulated osteocalcin production with low sintered HA.33 The very small crystallite size in this sample did not only influence the cell proliferation (as in HA-900) but also the phenotype expression. However, this mechanism cannot explain all the experimental results obtained in this work. Both BCP-45 and BCP-27 samples show very similar features to HA-900 but osteoblast-like cell proliferation is significantly higher. HA-900, BCP-45, and BCP-27 were treated at 900°C, and their microstructure and surface characteristics are very similar. However, the BCP samples notably increase the osteoblast proliferation compared to HA-900, indicating that -TCP presence enhances the cell proliferation. The chemical changes undergone by the culture medium also contribute to explain the cytostatic effect observed on HA-700 and HA-900. The medium is depleted in Ca2 after 24 h in contact with the samples. It is well known that the very small crystallite size, higher surface area, and porosity lead to the calcium (and phosphate) uptake to grow and nucleate new apatite crystals.35,36 In this way, the material consumes the Ca2 and phosphate available for cells, avoiding the proliferation and leading to a cytostatic effect. HA-700 and HA-900 are not cytotoxic. The very low LDH values demonstrate that very few cells are lysed when seeded on these samples. Therefore the low cell proliferation on HA-700 and HA-900 could be understood as a cytostatic effect due to calcium depletion and, in the case of HA-700, the low cell attachment. The LDH level for HA-1100 is much higher than that in HA-700 and HA-900. This amount of injured cells can be explained by the high and fast proliferation on this sample. The cells fully colonize the surface after 24 h of culture, and many osteoblasts could be lysed when detached, due to the limited surface available. BCP-45 and BCP-27 samples show a very interesting behavior. Both samples were constituted by a HA phase and a second more soluble -TCP phase. These samples differ in the phase composition percentage. From the textural and microstructural point of view, BCPs samples are very similar to HA-900. These results point out that the presence of -TCP does not modify these features, which are mainly dependent on the thermal treatment. Cell attachment is similar in BCPs and HA-900 samples, which is in agreement with the similar textural features of the samples. However, the cell proliferation on BCP samples is much higher than it is in HA-900. This result

SÁNCHEZ-SALCEDO ET AL.

288

FIG. 7.

Evolution of Ca2 content of complete medium as a function of time when nonseeded CaP are soaked.

could be explained by comparing the culture media evolution of HA-900 and BCPs samples. The Ca2 content decreases from 1.2 to 0.25 mM after 24 h in contact with HA-900. Thereafter the Ca2 keeps on decreasing until almost total depletion after 72 h. In BCP samples, the presence of a more soluble -TCP phase leads to a lower Ca2 decrease and avoids the Ca2 depletion. Consequently, the cell proliferation occurs on the surface. LDH levels are significantly lower in BCP samples (6 U  L1) than in HA-1100 (16 UL1) (p  0.05), indicating that the number of lysed cells is lower in BCP although the cell proliferation is very similar. This difference could be due to the multilayered growth of the cells on BCP sample (see Fig. 8), which could avoid in part the cell detachment and subsequent lysis (multilayered growth was not observed in HA-1100). We must take into account that these experiments were carried out in a closed system, so any modification in the culture media can lead to the misinterpretation of the cytocompatibility results. For instance, our in vitro cell proliferation results should not be extrapolated to an in vivo model, since the living medium (open system) can buffer the ion variations, allowing the cell proliferation on nanocrystalline HAs. However, the results obtained in this work provide an interesting insight in the field of ceramic scaffolds for tissue engineering. One of the most common strategies in tissue engineering is to place cells within ceramic matrices. The cells attached to matrices are subsequently implanted and become incorporated into the body. In this strategy, a critical aspect of ceramic design is to deter-

mine the appropriate composition, microstructure, pore size, porosity, and surface chemistry to match the specific biologic and metabolic requirements of tissues and disease conditions. High in vitro cell proliferation is very desirable prior to in vivo implantation. In this sense BCP samples show excellent properties for this purpose. BCPs can be obtained as nanocrystalline compounds, which could facilitate in vivo bioactivity and resorption of the scaffolds. On the other hand, BCP guarantees in vitro cell

FIG. 8. Detail of the multilayered cell growth onto a BCP-45 surface after 24 h culture.

EVALUATION OF POTENTIAL CALCIUM PHOSPHATE SCAFFOLDS proliferation due to the Ca2 and phosphate supplies into a closed in vitro system.

ACKNOWLEDGMENTS Financial support of CICYT Spain through research project MAT2005-01486 is acknowledged. D. Arcos is grateful to MEC for the financial support through “Ramón y Cajal” postdoctoral grant. We also thank A. Rodriguez (Electron Microscopy Center, Universidad Complutense) and F. Conde (C.A.I. X-ray Diffraction, Universidad Complutense) for their valuable technical assistance.

REFERENCES 1. Hench, L.L. Bioceramics: from concept to clinic. J. Am. Ceram. Soc. 74, 1487, 1991. 2. Vallet-Regí, M. Ceramics for medical applications. J. Chem. Soc. Dalton 2, 97, 2001. 3. Gross K.A., and Berndt C.C. Thermal processing of hydroxyapatite for coating production. J. Biomed. Mater. Res. 39, 580, 1998. 4. Kobayashi, T., Shingaki, S., Nakajima, T., and Hanada, K. Chin augmentation with porous hydroxyapatite blocks. J. Long-Term Effects Med. Impl. 3, 283, 1993. 5. Bonfield, W., Grynpas, M.D., Tuly, A.E., Bowman J., and Abram, J. Hydroxyapatite reinforced polyethylene-a mechanically compatible implant material for bone replacement. Biomaterials 2, 185, 1981. 6. Sari, A., Yavuzer, R., Ayhan, S., Tuncer, S., Latifoglu, O., Atabay, K., and Celebi, M.C. Hard tissue augmentation of the mandibular region with hydroxyapatite granules. J. Craniofac. Surg. 14, 919, 2004. 7. Daculsi, G. Biphasic calcium phosphate concept applied to artificial bone, implant coating and injectable bone substitute. Biomaterials 19, 1473, 1998. 8. Yamada, S., Heyman, D., Bouler J.M., and Daculsi, G. Osteoclastic resorption of calcium phosphate ceramics with different hydroxyapatite/-tricalcium phosphate ratios. Biomaterials 18, 1037, 1997. 9. Bouler, J.M., LeGeros, R.Z., and Daculsi, G. J. Biphasic calcium phosphates: influence of three synthesis parameters on the HA/-TCP ratio. J. Biomed. Mater. Res. 51, 680, 2000. 10. LeGeros, R.Z., Lin, S., Rohanizadeh, R., Mijares D., and LeGeros, J.P. Biphasic calcium phosphate bioceramics: preparation, properties and applications. J. Mater. Sci. Mater. Med. 14, 201, 2003. 11. Wilke, A., Orth, J., Lomb, M., Fuhrmann, R., Kienapfel, H., Griss, P., and Franke, R.P. Biocompatibility analysis of different biomaterials in human bone marrow cell cultures. J. Biomed. Mater. Res. 40, 301, 1998. 12. Hott, M., Noel, B., Bernache-Assolant, D., Rey, C., and Marie, P.J. Proliferation and differentiation of human trabecular osteoblastic cells on hydroxyapatite. J. Biomed. Mater. Res. 37, 508, 1997.

289

13. Malard, O., Bouler, J.M., Guicheux, J., Heymann, D., Pilet, P., Coquard, C., and Daculsi, G. Influence of biphasic calcium phosphate granulometry on bone ingrowth, ceramic resorption, and inflammatory reactions: preliminary in vitro and in vivo study. J. Biomed. Mater. Res. 46, 103, 1999. 14. Silva, S.N., Pereira, M.M., Goes, A.M., and Leite, M.F. Effect of biphasic calcium phosphate on human macrophage functions in vitro. J. Biomed. Mater. Res. 65A, 475, 2003. 15. Toquet, J., Rohanizadeh, R., Guicheux, J., Couillaud, S., Passuti, N., Daculsi, G., and Heymann, D. Osteogenic potential in vitro of human bone marrow cells cultured on macroporous biphasic calcium phosphate ceramic. J. Biomed. Mater. Res. 44, 98, 1999. 16. Gauthier, O., Bouler, J.M., Weiss, P., Bosco, J., Daculsi G., and Aguado, E. Kinetic study of bone ingrowth and ceramic resorption associated with the implantation of different injectable calcium-phosphate bone substitutes. J. Biomed. Mater. Res. 47, 28, 1999. 17. Jones, J.R., and Hench, L.L. Biomedical materials for new millennium: perspective on the future. Mater. Sci. Tech. 17, 891, 2001. 18. Langer, R., and Vacanti, J.P. Tissue engineering. Science 260, 920, 1993. 19. Hutmacher, D.W., Schantz, T., Zein, I., Ng, K.W., Teoh S.H., and Tan, K.C. Mechanical properties and cell cultural response of polycaprolactone scaffolds designed and fabricated via fused deposition modelling. J. Biomed. Mater. Res. 55, 203, 2001. 20. Elliot, J.C. Structure and chemistry of the apatites and other calcium orthophosphates. In: Studies in Inorganic Chemistry. Amsterdam: Elsevier, 1994. 21. LeGeros, R.Z. Calcium phosphates in oral biology and medicine. Monographs in Oral Science, vol. 15. Basel: Karger, 1991. 22. Suchanec, W., and Yoshimura, M. Processing and properties of hydorxyapatite-based biomaterials for use as hard tissue replacement implants. J. Mater. Res. 13, 94, 1998. 23. Rodríguez-Lorenzo, L.M, and Vallet-Regí, M. Controlled crystallisation of calcium phosphate apatites. Chem. Mater. 12, 2460, 2000. 24. Rodríguez-Carvajal, J. Recent advances in magnetic structure determination by neutron powder diffraction. Physica B. 192, 55, 1993. (For a more recent version see RodríguezCarvajal, J. Recent developments of the program FULLPROF, in commission on powder diffraction [IUCr]. Newsletter 26, 12, 2001, available at http://journals.iucr. org/iucr-top/comm/cpd/Newsletters/ or ftp://ftp.cea.fr/pub/ llb/divers/fullprof.2k.) 25. Kay, M.I, Young, R.A, and Posner, A.S. Crystal structure of hydroxyapatite. Nature 204, 1050, 1964. 26. Yashima, M., Sakai, A., Kmiyama, T., and Hoshikawa A. Crystal structure analysis of -tricalcium phosphate Ca3(PO4)2 by neutron powder diffraction. J. Solid State Chem. 175, 272, 2003. 27. ASTM E 384-84. Standard Test Method for Microhardness of Materials. Philadelphia: ASTM Committee on Standards, 1984. 28. Reddi, A.H., and Huggins, C.B. Citrate and alkaline phosphatase during transformation of fibroblast by the matrix

SÁNCHEZ-SALCEDO ET AL.

290

29. 30.

31.

32.

33.

34.

and minerals of bone. Proc. Soc. Exp. Biol. Med. 140, 807, 1972. Anselme, K. Osteoblast adhesion on biomaterials. Biomaterials 21, 667, 2000. Zreiqat, H., Standard, O.C., Gengenbach, T., Steele, J.G., and Howlett, C.R. The role of surface characteristics in the initial adhesion of human bone-derived cells on ceramics. Cells Mater. 6, 45, 1996. Alliot-Licht, B., Gregoire, M., Orly, I., and Menanteau, J. Cellular activity of osteoblasts in the presence of hydroxyapatite: an in vitro experiment. Biomaterials 12, 752, 1991. Orly, I., Grégoire, M., Menanteau, J., and Dard, M. Effects of synthetic calcium phosphates on the 3H-thymidine incorporation and alkaline phosphatase activity of human fibroblasts in culture. J. Biomed. Mater. Res. 23, 1433, 1989. Ong, J.L., Hoppe, C.A., Cardenas, H.L., Cavin, R., Carnes, D.L., Sogal, A., and Raikar, G.N. Osteoblast precursor cell activity on HA surfaces of different treatments. J. Biomed. Mater. Res. 39, 176, 1998. Kozawa, O., Takatsuki, K., Kotake, K., Yoneda, M., Oiso, Y., and Saito, H. Possible involvement of protein kinase C

in proliferation and differentiation of osteoblast-like cells. Fed. Eur. Biochem. Soc. 243, 183, 1989. 35. Frayssisnet, P., Rouquet, N., Fages, J., Durand, M., Vidalain, P.O., and Bonel, G. The influence of sintering temperature on the proliferation of fibroblastic cells in contact with HA-bioceramics. J. Biomed. Mater. Res. 35, 337, 1997. 36. Radin, S.R., and Ducheyne, P. The effect of calcium phosphate ceramic composition and structure on in vitro behaviour. II. Precipitation. J. Biomed. Mater. Res. 27, 35, 1993.

Address reprint requests to: Prof. Maria Vallet-Regí Departamento de Química Inorgánica y Bioinorgánica Facultad de Farmacia Universidad Complutense Plaza Ramón y Cajal s/n 28040 Madrid, Spain E-mail: [email protected]

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.