Cell Response of Calcium Phosphate Based Ceramics , a Bone Substitute Material

Centro de Ciências Naturais e Humanas – CCNH, Universidade Federal do ABC – UFABC, Rua Santa Adélia, 166, Bangu, CEP 09210-170, Santo André, SP, Brazil Centro de Engenharia e Ciências Sociais Aplicadas – CECS, Universidade Federal do ABC –UFABC, Rua Santa Adélia, 166, Bangu, CEP 09210-170, Santo André, SP, Brazil Centro de Ciência e Tecnologia de Materiais – CCTM, Instituto de Pesquisas Energéticas e Nucleares – IPEN, Av. Professor Lineu Prestes, 2242, Cidade Universitária, CEP 05508-000, São Paulo, SP, Brazil Departamento de Dentística, Faculdade de Odontologia – FO, Universidade de São Paulo – USP, Av. Professor Lineu Prestes, 2227, Cidade Universitária, CEP 05508-900, São Paulo, SP, Brazil


Introduction
Bone grafting is the most common form of regenerative therapy and has been used for almost 100 years in attempts to stimulate healing of bone defects.Alloplastic bone graft materials are synthetic, inorganic, biocompatible, and bioactive.These bone substitutes are believed to promote healing of bone defects through osteoconduction.
Progress in the field of biomaterials like bioglass, glass-ceramics and calcium phosphate ceramics (such as hydroxyapatite (HAp), α-tricalcium phophate (α-TCP), β-tricalcium phophate (β-TCP) and biphasic mixtures (BCPs)) has been significant in the last few decades [1][2][3][4] .The in vitro and in vivo behavior of these ceramics can be influenced by the Ca/P ratio, raw material purity, processing variables (such as conformation technique and sintering conditions) and test parameters 5 .The Ca/P ratio can result in a material that is resorbable or stable in the human body environment, through a dynamic process involving degradation, reabsorption, phase transformation and cell adhesion 1,[6][7][8] .Often, the best results following extended duration experiments were obtained with biphasic ceramics, considered to have good bioactive properties 7,9 .However, calcium phosphate ceramics with a different Ca/P ratio can respond differently in short term tests.Data reported in the literature are results of usually long duration tests and seldom consider responses during the initial stages.
Dissolution of calcium phosphate ceramics degrades its properties affecting formation of the calcium phosphate layer, cell anchorage and the morphological changes in in vitro tests 10 .This is particularly important when the kinetics of the dissolution process of material with varying composition is still unknown.To obtain improved understanding of the interaction between calcium phosphate with different compositions and the DMEM media, the dissolution behavior was determined without any cell approach 11 .However, since calcium phosphate ceramics do not exhibit inert behavior, its characterization through cell culture experiments was carried out 12,13 .
Besides bioactivity or bioinertness 14,15 , other aspects of material-tissue interface interactions are important in the context of biological performance of a biomaterial [16][17][18] .The composition, purity and roughness of calcium phosphate ceramics affect physical and chemical characteristics, and these in turn influence the biological behavior of these materials.To provide more information to clinicians who wish to use bioceramics as bone substitutive material, this investigation was carried out to determine the biological response of different calcium phosphate ceramics.The in vitro response of FMM1 fibroblasts on high purity calcium phosphate ceramics containing different Ca/P ratio was determined.

Characterization of raw material powders
Commercially available powders of HAp (20-2039 Strem Chemicals, Newburyport, MA, USA) and β-TCP (21218, Fluka, Buchs, SG, Switzerland) were used.The specific surface area of these powders was determined by the Brunnauer Emmet Teller (BET) method using a Nova 1200 BET Surface Area Analyzer (model 3.11, Quantachrome, Boynton Beach, Florida, USA) and the particle size distribution as well as the mean particle sizes were determined using a laser equipment (Cilas1064, 104, Orléans, France).

Sample preparation
The physical, chemical and biological properties of calcium phosphate ceramics were determined using disks (6.5 mm in diameter and ~5 mm high).The disks (50 in number) were prepared as follows: the BCPs powders were mechanically mixed (Turbula-type T2C system, Shatz Willy Abachofen Maschinen, Switzerland) for 2 hours with alumina spheres (~2 mm diameter) and deionized water.The five different compositions of calcium phosphate powders (Table 1) were uniaxially pressed (100 MPa), then cold isostatically pressed (200 MPa) and sintered under different conditions (Table 1), depending on TCP:HAp weight ratio 19 .Ten samples of each composition were prepared to permit statistical analysis of the results.The theoretical densities were estimated using the rule of mixtures 20 , assuming 3.07 g. m -3 for β-TCP and 3.16 g.cm -3 for HAp as their densities.

Characterization of Physical and chemical properties
The sintered calcium phosphate ceramic samples were characterized in terms of their density, surface roughness, phase composition and microstructure, as described below.
The densities of the samples were determined using the Archimedes method 21 .The final densities were expressed as percentage of the theoretical densities.
The surface roughness was measured with a Mitutoyo Surftest 211 portable rugosimeter (Mytutoyo, Japan).The mean surface roughness 'Ra' was estimated using 15 measurements, considering five along three different parallel lines.

Biological response
The biological response of the calcium phosphate bioceramics with different Ca/P ratios was determined by counting cultured cells attached to the top of the bioceramic disks.The cells were cultured and plated on the top of the disks, then, cell adhesion was assessed after 1 day and cell proliferation from 1 to 3 days.

Cell culture
The cells were cultured as previously described 22,23 .Briefly, the FMM1 fibroblasts, a human gingival cell line, were used.These cells were cultured in Dulbecco's modified Eagle medium (DMEM), supplemented with 10% fetal bovine serum (FBS, Cultilab, Campinas, SP, Brazil) and 1% antimycotic-antibiotic solution (10,000 U of penicillin, 10 mg of streptomycin, and 25 mg of amphotericin B per mL in 0.9% sodium chloride; Sigma Chemical Company, St Louis, MO).The cells were kept in an incubator at 37 °C and humidified 5% CO 2 atmosphere.Cultures were supplied with fresh medium every other day.Cells between the 10 th and the 14 th passages were used in all experimental procedures.The cell culture procedures were done under laminar flow and the tests followed the OSO10993-5 24 .

Cell adhesion and proliferation assays
Sterile ceramic disks of all the experimental groups (n = 45) were placed in the bottom of 24-wells cell culture plates (Corning Costar, Cambridge, MA) and covered with DMEM.Then, 10 5 cells were seeded on the top of each disk and the cell culture plates were incubated at 37 °C in a 5% CO 2 atmosphere.One, 2 and 3 days after seeding, three disks of each experimental group were prepared for SEM examination.

Scanning electron microscopy (SEM)
A SEM was used to observe the surface and to evaluate cell adhesion as well as growth.To observe the surface All the samples were fixed with 2.5% glutaraldehyde in 0.1 M phosphate buffer solution (pH 7.4) overnight at 4 °C for further SEM studies.Then, the samples were post fixed in 1% osmium tetroxide in 0.1 M phosphate buffer solution (pH = 7.4).The samples were then dehydrated in ethanol and chemically dried in hexa methyl disilazane (HMDS, Electron Microscopy Sciences, Fort Washington, PA).The samples were then gold sputter coated before SEM analysis.

Cell counting
Cell counting was done with SEM micrographs of three defined areas of each specimen taken from the same distance (10 mm) and at the same magnification (500X).The cells were counted using the Omage Pro Plus computational image software 25 .The counting provided data to obtain cell growth curves.

Statistical analysis
The number of cells counted in the scanning electron micrographs, obtained in triplicate, is presented as mean ± standard error of the mean.These data were compared by analysis of variance (ANOVA) complemented by the Tukey's test.The level of significance was 5% (p ≤ 0.05).

Characterization of raw material powders
The mean diameters of 10%, 50% and 90% volume (d 90 , d 50 , d 10 , respectively), mean diameter size and specific surface area of the HAp and β-TCP powders are shown in Table 2.The particle size of both powders indicated a multimodal distribution and was narrower for HAp.The HAp and β-TCP powders had similar d 50 , (2.83 and 2.67 mm, respectively).The mean diameters of HAp and β-TCP were 3.6 mm and 4 mm, respectively.

Physical and chemical characterization
The final density and roughness of calcium phosphate samples are shown in Figure 1.The final densities were about 93 ± 3% of their theoretical densities.The densities of calcium phosphate samples increased slightly with HAp content in the mixture.All samples showed less than 10% residual porosity.
The roughness of pure samples, namely 100 HAp and TCP were higher, with 1.75 and 1.55 µm, respectively.The roughness of 50 HA samples was lower and ∼1 mm.The 25 HA and 75 HA samples had intermediate roughness.Overall, calcium phosphate samples with lower densities had higher surface roughness (R 2 = 0.98).
Data from XRD analysis (Figure 2a, b) showed the presence of either HAp or β-TCP in single phase sintered ceramics.The 50 HA sample showed crystalline phases of both HAp and β-TCP.On the detail of the 30-35° range XRD (Figure 2b), α-TCP phase was not identified in all the samples.The other biphasic sintered ceramics also showed the two crystalline phases of calcium phosphate.
Typical SEM micrographs of surfaces of samples of calcium phosphate with different amounts of HAp and taken prior to the in vitro tests are shown in Figure 3.The 100 HA sample showed higher level of densification than the other mixtures.Samples with more than 50% β-TCP phase (Figure 3d, e) revealed grain growth.Residual porosity was also observed in these samples.

Biological response
Cell adhesion can be observed in the scanning electron micrographs of fibroblasts attached to ceramic substrates after 1 day of seeding (Figures 4 and 5).The overall morphology of the fibroblasts was similar in all groups showing stellate or fusiform aspects.Moreover, adherence of FMM1 cells was observed in samples from all groups.The cell monolayers were loose after one day of seeding.The high magnification SEM micrographs (Figure 5) reveal the overall cell morphology and cell extensions on the ceramic sample surface.
Cell proliferation can be observed by comparing the different scanning electron micrographs with fibroblasts attached to the ceramic substrates after 2 and 3 days of seeding (Figure 6 and 7).The cell monolayer was denser after 3 days (Figure 6 f-j).There were no marked differences in the distribution of cells on the top surface of samples from the different groups until 2 days after seeding.After longer times the cell density was higher and the cells had  more extensions.At the end of the study period (3 days), the samples made of TCP were completely covered by a cell monolayer (Figure 6f), whereas samples from the other groups still revealed areas devoid of cells.
Figure 8 shows graphically the number of FMM1 cells adhered to and proliferated on ceramic sample surfaces as a function of time (1, 2 and 3 days).The number of cells increased significantly, independent of the experimental  group (p ≤ 0.05) from the beginning till the end of the test.On day one, there was a significant increase in the number of adhered cells to the top of TCP and 25HA disks, compared with that adhered to 100 HA disks (p < 0.01).Disks with more than 50% of HAp revealed significantly fewer cells than TCP (p < 0.05).After three days, the TCP group revealed many more cells than all other groups (p < 0.05).Amongst the other groups, 25HA showed more cells than all other groups, except the TCP (p < 0.05).

Discussion
Search for bone substitute materials is an ongoing challenge.The composition and topography of these materials is of importance to determine its biological response.A common clinical situation that periodontists encounter is fixing of bone defects and this often requires the use of bone substitute materials.These materials include calcium phosphate based ceramics such as hydroxyapatite (HAp) and β-tricalcium phosphate (β-TCP).
Characterization of raw calcium phosphate powders, including particle size distribution, was done because this information is essential to define the processing conditions to obtain crystalline ceramics of HAp, β-TCP or a mixture of these, without phase transformation and with desirable densities.Samples of these ceramics were sintered under different conditions in achieve similar final densities.The  higher temperature used to sinter our calcium phosphate ceramics with more than 50% β-TCP phase lead to grain growth, even though the raw materials had similar particle size distribution.Moreover, the final densities that were attained (about 93% of the theoretical densities) gave the samples adequate microporosity to drain fluid, enabling thus, the dissolution/precipitation process 1 .Since physical and chemical properties of biomaterial surfaces play an important role in osteointegration 10,26 , initially the physical characteristics of the calcium ceramic materials were determined, focusing on composition, surface roughness and the microstructure.
The composition and surface roughness of calcium phosphate based ceramic are important to predict interfacial behavior at the material/tissue interface and its interaction with the biological environment 27 .Ot has been shown that smooth surfaces are suitable for soft tissues, both for anchoring and growth of fibroblasts 28,29 , However, roughened calcium phosphate implant surfaces have better bone deposition compared with polished surfaces 30 .Surface roughness influences cell alignment 31,32 , orientation and migration.These aspects are important in several bone formation stages, including adhesion, proliferation, differentiation, synthesis of bone matrix, maturation and calcification of tissue on the material's surface 27,[33][34][35] .Another important factor affecting initial cell adhesion is the presence of specific phases, as several authors have suggested that adhesion takes place only after the dissolution process 1,36 .TCP samples were more soluble, had higher surface roughness and the lowest density.These features lead to faster cell growth compared to samples containing HAp.On fact, higher amounts of HAp in the ceramic led to higher density, lower roughness and decreased solubility.Consequently, the number of cells attached to the surface of these ceramics was always lesser than those on TCP samples.
Thus, the higher biocompatibility of the TCP samples could be attributed not only to its physical characteristics but also to its dissolution behavior.The β-TCP dissolved faster, compared to Hap 1,2 and this aspect could have induced micro-roughness by diffusion of surface ions, and increased fluid drain.The SEM micrographs corroborate this observation and revealed higher porosity in TCP samples, which probably resulted in a surface texture more propitious for cell adhesion.On in vivo conditions, these micropores can act as fluid drains, supplying nutrients to   the neoformed tissue 37,38 .Hence, solubility of TCP is an important parameter to increase its biocompatibility.
Based in the results of this study we conclude that as a bone substitute material, the TCP ceramic has a faster biological response.Even though TCP does not exhibit bioactive characteristics of HAp ceramic, it is bioresorbable, has the density and roughness that make it have higher cell adhesion and proliferation.This in turn indicates that this material (TCP) is a better bone substitute.Even though biphasic ceramics are considered to be more bioactive, TCP ceramics are capable of inducing faster responses at the initial stages, suggesting that calcium phosphate based ceramics with a Ca/P concentration gradient can be more suitable for use as implant materials.A surface layer of TCP can further accelerate the osteointegration process and improve the fields of application of these ceramics.On general, the experiments reported in the literature have concentration gradient can be used as an implant.The best calcium phosphate ceramic implant seems to be TCP coated BCP for quicker osteointegration.

Figure 8 .
Figure 8. FMM1 cell growth on HAp/TCP ceramic substrates as a function of different cell exposure time.Different letters mean that the values are statistically different, for the same exposure time.

Table 1 .
Characteristics of experimental groups of calcium phosphate ceramics.

Table 2 .
Characteristics of HAp and β-TCP powders as raw materials for calcium phosphate ceramics.