Journal of Materials Science: Materials in Medicine

, Volume 23, Issue 7, pp 1569–1581

Development of strong and bioactive calcium phosphate cement as a light-cure organic–inorganic hybrid

Authors

  • M. Barounian
    • Materials and Energy Research Center
    • Materials and Energy Research Center
  • A. Kazemzadeh
    • Materials and Energy Research Center
Article

DOI: 10.1007/s10856-012-4637-z

Cite this article as:
Barounian, M., Hesaraki, S. & Kazemzadeh, A. J Mater Sci: Mater Med (2012) 23: 1569. doi:10.1007/s10856-012-4637-z

Abstract

In this research, light cured calcium phosphate cements (LCCPCs) were developed by mixing a powder phase (P) consisting of tetracalcium phosphate and dicalcium phosphate and a photo-curable resin phase (L), mixture of hydroxyethylmethacrylate (HEMA)/poly acrylic-maleic acid at various P/L ratios of 2.0, 2.4 and 2.8 g/mL. Mechanical strength, phase composition, chemical groups and microstructure of the cured cements were evaluated at pre-set times, i.e. before and after soaking in simulated body fluid (SBF). The proliferation of Rat-derived osteoblastic cells onto the LCCPCs as well as cytotoxicity of cement extracts were determined by cell counting and 3-{4,5-dimethylthiazol-2yl}-2,5-diphenyl-2H-tetrazolium bromide assay after different culture times. It was estimated from Fourier transforming infrared spectra of cured cements that the setting process is ruled by polymerization of HEMA monomers as well as formation of calcium poly-carboxylate salts. Microstructure of the cured cements consisted of calcium phosphate particles surrounded by polymerized resin phase. Formation of nano-sized needlelike calcium phosphate phase on surfaces of cements with P/L ratios of 2.4 and 2.8 g/mL was confirmed by scanning electron microscope images and X-ray diffractometry (XRD) of the cured specimen soaked in SBF for 21 days. Also, XRD patterns revealed that the formed calcium phosphate layer was apatite phase in a poor crystalline form. Biodegradation of the cements was confirmed by weight loss, change in molecular weight of polymer and morphology of the samples after different soaking periods. The maximum compressive strength of LCCPCs governed by resin polymerization and calcium polycarboxylate salts formation was about 80 MPa for cement with P/L ratio of 2.8 g/mL, after incubation for 24 h. The strength of all cements decreased by decreasing P/L ratio as well as increasing soaking time. The preliminary cell studies revealed that LCCPCs could support proliferation of osteoblasts cultured on their surfaces and no cytotoxic effect was observed for the extracts of them.

1 Introduction

Since Charnley [1] introduced polymethylmethacrylate (PMMA) bone cement for prosthetic fixation in 1960, it has been widely used in orthopedic surgery. While it is clinically successful, there are several problems associated with the use of this cement. The most serious one is its nonadhesiveness to bone, namely, the formation of a fibrous layer between the bone surface and the cement [2], which is one of the major contributions to the loosening of cemented femoral components. PMMA cement relies on mechanical interlocking with bone rather than adhesive chemical bonding to form a stable cement–bone union [3]. A solution to overcome this problem at the interface is to use bioactive bone cement. Various types of bioactive fillers, based on calcium phosphates, hydroxyapatite (HA) powders and bioactive glass or glass–ceramics powder, may be used [46]. However the resulted bioactive bone cements exhibit decreased mechanical properties in comparison to pure PMMA cement [7].

Hydraulic calcium phosphate cements (CPCs) with unique characteristics such as osteoconductivity, bioactivity and low temperature setting reaction are also appropriate alternatives for bone defect treatments. CPCs generally consist of a powder phase (mixture of various calcium phosphates such as tetracalcium phosphate (TTCP) and dicalcium phosphate) and an aqueous liquid phase (such as aqueous solution of Na2HPO4), which are mixed to form a paste [8, 9]. The paste is placed into a defect as a substitute for the damaged part of bone [10, 11]. CPC forms HA in an aqueous environment at body temperature, hence it is more similar to biological apatite than sintered HA formed at high temperatures.

However the main disadvantage of CPC is its poor mechanical strength once after setting which makes it unsuitable for load-bearing applications [11]. Another drawback which limits clinical use of self-setting CPC is its hardening behavior during operation. It persuades the surgeons to place CPC paste into a defect within a prescribed time [12]. Mixing CPC paste that undergoes setting phenomenon (precipitation of apatite crystals) may result in destruction of the configured precipitated crystals and thus lead to formation of inhomogeneous cement with undesirable consistency and low mechanical strength.

To solve such problems, it is suitable to develop strong and bioactive bone cement in which a paste is prepared in advance under well-controlled conditions and is not hardened during preparation or usage. The goals were achieved in our study where strong and bioactive bone cement with photo-curable setting behavior was introduced. The cement composition consisted of powder phase including mixture of TTCP and dicalcium phosphate and liquid phase comprising hydroxyethylmethacrylate (HEMA)/poly(acrylic-maleic acid, PAMA) resin. The powder has been reported to be used as solid phase of hydraulic calcium phosphate bone cements which are used for treatment of osteoporotic fractures and other related usages [13]. The resin phase is a biocompatible material widely used in preparation of light-cure glass-ionomer dental cements for lining and canal sealing [14]. Light cured cements with various powder to liquid ratios were prepared and studied in relation to strength, biodegradation and biocompatibility. The structural properties and in vitro nanoapatite formation ability of the best specimen in term of mechanical strength was also evaluated and compared with those of conventional hydraulic CPC (c-CPC) reported in literatures. In this report, the vocabularies c-CPC and hydraulic CPC are synonyms.

2 Materials and methods

2.1 Starting materials

TTCP, dicalcium phosphate anhydrous (DCPA, Merck, Germany) and commercially available {HEMA/PAMA} resin were employed for cement preparation. The resin commonly used as liquid part of light-cure glass-ionomer dental cements was purchased from Fuji Company (Fuji II LC, Japan). It consists of an aqueous solution of HEMA and PAMA containing camphorquinone, a photo-initiator that absorb ultra-violet radiation at 250 nm and visible light at 470 nm [15]. TTCP was synthesized by solid state reaction as describe elsewhere [16]. Briefly, equimolar amount of DCPA, and calcium carbonate (Merck, 2069, Germany) was heated in an electric furnace (CARBOLITE BLF, 16/3, England) at 1,500 °C for 6 h, quenched to room temperature, crushed in an agate mortar and then grounded in a planetary mill. Other reagents such as NaHCO3, K2HPO4·3H2O, KCl, NaCl, Na2SO4, CaCl2·2H2O, MgCl2·6H2O, HCl and tris(hydroxyl-methyl) aminomethane used for preparation of simulated body fluid (SBF) solution were purchased from Merck Company (Germany).

2.2 Cement formulations

A homogenous powder consisting of 1 mol of TTCP with an average particle size of 10 μm (data obtained by Laser Particle Size Analyser, Fritsch Analysate 22) and 1 mol of DCPA with an average particle size of 6 μm was used as solid phase of light-cure CPCs (LCCPCs) and photo-curable resin of HEMA/PAMA was used as liquid phase. Several cement pastes were prepared by mixing the powder phase (P) and the liquid phase (L) at different P/L ratios according to Table 1.
Table 1

Chemical formulations of LCCPCs with different P/L ratios

Code

Powder phase

Liquid phase

P/L ratio (g/mL)

LC-2.0

TTCP + DCPA

HEMA–PAMA resin

2.0

LC-2.4

TTCP + DCPA

HEMA–PAMA resin

2.4

LC-2.8

TTCP + DCPA

HEMA–PAMA resin

2.8

To select maximum P/L ratio to have a workable paste, cement pastes with various P/L ratios (from 2.0 to 3.0 g/mL) were tested in terms of consistency based on the method described previously [17]. Briefly, the cement paste was pressed between two glassy slabs under compressive pressure of 20 N. The sample was considered workable when no cracks were appeared at the edges of pressed paste.

2.3 Preparation of LCCPCs

LCCPCs were prepared as follows: Powder phase was mixed with the liquid to achieve a homogenous paste. The paste was transferred into glass-tubing template and cylindrical-shaped specimen with 12 mm in height and 6 mm in diameter was light-polymerized by irradiating sequentially each face of specimen assembly for 80 s (according to manufacturer’s suggestion) with UV light source (Farazmehr Isfahan) at room temperature.

The accurate ratio of polymer/calcium phosphate powder was measured by using thermo-gravimetry analysis (TGA) at a heating rate of 5 °C/min. Powdered alumina was used as reference material.

2.4 In vitro behavior of LCCPCs

The SBF solution that resembles composition of inorganic constituent of blood plasma was prepared by dissolving NaCl 7.995 g, NaHCO3 0.353 g, KCl 0.224 g, K2HPO4·3H2O 0.228 g, MgCl2·6H2O 0.305 g, CaCl2 0.227 g, and Na2SO4 0.071 g into 1 L distilled water and the solution was buffered by tris (hydroxymethylaminomethane) and pH 7.4 was adjusted at 7.4 using HCl [18]. The in vitro degradation studies were performed by immersing the cylindrical samples in SBF (solid to liquid loading of 1 g per 100 mL) at 37 °C. After soaking for different preset time intervals (1, 7, 14 and 21 days), the samples were taken out from the SBF solution, rinsed gently with de-ionized pure water and dried in air condition for further characterization. It should be noted that during evaluation, the SBF solution was renewed every 24 h.

2.4.1 Compressive strength

The compressive strength of samples soaked in SBF solution for various times was measured according to ASTM standard F451-08 using universal testing machine (Zwick/Roell-HRC 25/400) with a crosshead speed of 1 mm min−1. In this part of study, the compressive strength of cured cements after incubation at 37 °C for 24 h was also determined for comparison with hydraulic CPCs.

2.4.2 Morphology

Surface microstructure of the soaked samples was analyzed by using a scanning electron microscope (SEM, Stereoscan S 360 Cambridge) that operated at an accelerating voltage of 20 kV and coupled with energy dispersive X-ray analysis (EDXA, Oxford, US). Due to the poor electrical conductivity of the samples, their surfaces were coated with a thin layer of gold before testing. The surface morphologies of samples before storing in SBF were also provided for comparison. The X-ray maps of Ca and C elements were also taken to determine distribution of calcium phosphate and resin in the composite.

2.4.3 Weight loss

The weight loss (WL) of samples was determined by the following equation:
$$ {\text{W}}_{\text{L}} {\text{\%}}=100 \times ({\text{W}}_{0} - {\text{W}}_{\text{d}}) / {\text{W}}_{0}, $$
(1)
where W0 was original weight, and Wd was dried weight of samples.

2.4.4 Molecular weight

The molecular weight of polymer component of cured cements before and after immersion was measured at 37 °C by gel permeation chromatography (Agilent1100) on a Waters 2410 instrument. The sample was dissolved in dimethylformamide-pyridine mixture solvent and filtered with a 0.22 μm filter to remove the calcium phosphate particles, thus producing a dimethylformamide-pyridine solution of polymer. Polystyrene standard (Polyscience Co.) was chosen as calibration and Waters Millennium 32 as the data-processing software. The weight-averaged molecular weights (Mw) of the scaffolds were determined.

2.4.5 Phase composition and chemical groups

Phase composition and structural groups of the cement with maximum compressive strength was characterized before and after soaking using X-ray diffractometry, XRD (Philips PW3710) with Cu Kα radiation and Fourier transforming infrared (FTIR) spectroscopy, respectively. For FTIR spectroscopy, two milligrams of the ground specimen was mixed with 800 mg of ground spectroscopic grade KBr and pressed to make a transparent KBr pellets. The Infrared spectra between 400 and 4,000 cm−1 were measured at a resolution of 2 cm−1 using BRUKER VECTOR 33 device. All specimens were ground to powder and then analyzed.

2.5 Cell studies

2.5.1 Proliferation and morphology of cells on LCCPCs

To evaluate cell responses of the LCCPC, osteoblastic cells were derived from newborn rat calvaria and isolated by sequential collagenase digestion from calvaria of newborn (2–5 days) Wistar and cultured in Dulbecco modified Eagle medium (Gibco-BRL, Life Technologies, Grand Island, NY) supplemented with 15 % fetal bovine serum (Dainippon Pharmaceutical, Osaka, Japan) and 100 g/mL penicillin–streptomycin (Gibco-BRL, Life Technologies) in 5 % CO2 and 95 % air atmosphere at 37 °C for 1 week. The medium was changed every 2 days. The confluent cells were dissociated with trypsin and subcultured to 3 passages which were used for tests.

The disc-shaped LCCPC specimens (5 mm in diameter and 2 mm in height) were sterilized using 70 % ethanol and the osteoblastic cells were seeded on tops of the cement discs at 1 × 104 cells/disc. Polystyrene discs with the same surface area were also seeded as control group. The specimen/cell constructs were placed into 24-wells culture plates and left undisturbed in an incubator for 4 h to allow the cells to attach to them and then an additional 3 mL of culture medium was added into each well. The cell/specimen constructs were cultured in a humidified incubator at 37 °C with 95 % air and 5 % CO2 for 1,7, and 14 days. The medium was changed every 3 days.

The viability of the osteoblastic cells on cement specimens was determined using the MTT (3-{4,5-dimethylthiazol-2yl}-2,5-diphenyl-2H-tetrazolium bromide) assay. For this purpose, at the end of each evaluating period, the medium was removed and 2 mL of MTT solution was added to each well. Following incubation at 37 °C for 4 h in a fully humidified atmosphere at 5 % CO2 in air, MTT was taken up by active cells and reduced in the mitochondria to insoluble purple formazan granules. Subsequently, the medium was discarded and the precipitated formazan was dissolved in dimethylsulfoxide (150 mL/well), and optical density of the solution was read using a microplate spectrophotometer (BIO-TEK Elx 800, Highland park, USA) at a wavelength of 570 nm.

Morphological evaluation of the cells on the surfaces of the cement specimens was carried out as follows: The cells were cultured onto the discs as mentioned above. After 14 days, the medium was removed, the cell-cultured specimens were rinsed with phosphate buffered saline (PBS) twice and then the cells were fixed with 500 mL/well of 3 % glutaraldehyde solution (diluted from 50 % glutaraldehyde solution (Electron Microscopy Science, USA) with PBS). After 30 min, they were rinsed again and kept in PBS at 4 °C. Specimens were then fixed with 1 % Osmium tetroxide (Polyscience, Warmington, PA, USA). After cell fixation, the specimens were dehydrated in ethanol solutions of varying concentration (30, 50, 70, 90, and 100 %) for about 20 min at each concentration. The specimens were then dried in air, coated with gold and analyzed by SEM (Streoscan S 360, Cambridge).

2.5.2 MTT assay on LCCPCs extracts

In this part of study, the extracts of LCCPCs were prepared in full culture medium according to International Standard Organization (ISO/EN 10993-5) [17] and used for cytotoxicity evaluations. The dissolution extracts of various cements were prepared by adding each LCCPC in powdered form to serum free α-MEM culture medium at a ratio of 200 mg/mL. After incubation at 37 °C for different times (1, 7, and 14 days), the mixture was centrifuged and filtered through a membrane. The extracts were sterilized using 0.22 μm filter and used in the cell culture experiments. The osteoblastic cells were seeded at a density of 1 × 104 cells/well into 96-well plate and incubated for 24 h. After cell adhesion, the culture medium was removed and each extract supplemented with 15 % FCS was added to the plate inoculating the cells. The culture medium supplemented with 10 % FCS without addition of extracts was used as a blank control (control group). Then, the cells were incubated at 37 °C in 5 % CO2 for 1, 7 and 14 days. The medium was removed every 3 days and fed with fresh one (including extracts or control medium). The viability of the bone cells after espousing to various extracts was determined as described in Sect. 2.5.1.

2.6 Statistical analysis

Data were processed using Microsoft Excel 2003 software. The results were produced as mean ± standard deviation of at least four experiments. Statistical significance between mean values was determined by a one-way analysis of variance, and significance differences of means were evaluated by Tukey’s post hoc test (SPSS v10.0, Chicago, IL, USA). The P ≤ 0.05 was considered significant.

3 Results

3.1 Optimum P/L ratio of pastes and polymer/calcium phosphate contents

Figure 1 shows TGA graphs of different cements. The weight loss in the range of 280–400 °C relates to removal of polymeric phase from the composite which is correlated with the content of resin phase in the LCCPCs composition. Total weight loss of samples in the range of 50–500 °C is 23, 28 and 33 wt% for LC-2.8, LC-2.4 and LC-2.0, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig1_HTML.gif
Fig. 1

The TGA graphs of LCPCS with different P/L ratios

The results of the consistency test showed that the maximum P/L ratio was 2.8 g/mL and as shown in Fig. 2, cracked pastes with undesirable configurations were obtained using P/L ratios higher than this value (e.g. P/L = 3.0 g/mL).
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig2_HTML.jpg
Fig. 2

Visual situation of different LCCPCs after pressing between two glassy slabs under 20 N pressure

3.2 Compressive strength

Figure 3 shows the compressive strength of LCCPCs before and after soaking in SBF solution for various periods. The LC-2.8 specimen exhibits an ultimate compressive strength value of about 80 MPa after incubation at 37 °C for 24 (before soaking). The value of compressive strength decreases with decreasing the P/L ration. Furthermore, LCCPCs tends to weaken by soaking them in SBF solution in a time-dependent manner. For example a 45 % decrease in compressive strength of LC-2.8 specimen is observed after immersing for 7 days which reaches to about 185 % after 21 days. Compressive strength values of 3 and 17 MPa have been reported for c-CPC after 24 h of incubation and 7 days of soaking, respectively [17]. The hydraulic cements had been tested under conditions similar to this study. These data claim that the LCCPCs is more mechanically robust than c-CPC.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig3_HTML.gif
Fig. 3

The compressive strength of different LCCPCs 24 h after incubation at 37 °C and 1–21 days soaking in SBF solution

3.3 Morphology and image analysis

Figure 4 represents micrograph of LCCPC prepared at P/L ratio of 2.4 g/mL (LC-2.4 specimen) along with distribution map of Ca, P and C elements in its microstructure. The figures shows uniform distribution of Ca, P and C elements in the composite, reflecting a nearly homogenous distribution of calcium phosphate (from the maps of Ca and P elements) and polymer (from the map of C) phases.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig4_HTML.jpg
Fig. 4

Elemental image maps of Ca, P and C in cured LC-2.4 specimen determining distribution of calcium phosphate powder in polymer phase

Figure 5 shows SEM micrographs taken from the surfaces of different LCCPCs before and after soaking in SBF solution along with the EDXA analysis taken from the region marked by circles in the SEM images of 21-days-soaked samples. In all samples, before soaking, fine particles of reactants (calcium phosphate) have been surrounded by a monolithic polymer phase to yield a compacted microstructure. The micropores observed in the composite microstructures are due to the air bubble trapping during mixing powder and liquid phases. After 21 days of soaking, large pores (5–20 μm) with good interconnectivity are observed in the microstructures of all samples. The size, number and interconnectivity of the pores increase with decreasing P/L ratio (or increasing in content of resin phase) of cements. SEM pictures with higher magnifications reveal deep important difference in morphologies of different samples after soaking for 21 days. Confluent nanosized needlelike crystals with thickness of 10–20 and length of 50–100 nm are observed in the microstructures of LC-2.4 and especially LC-2.8 cements. These crystals have been produced on samples during soaking them in SBF solution and are quite similar to the morphology of apatite layer formed on bioactive glasses and glass–ceramics [19]. No sign of these crystals is observed in the microstructure of 21 days-soaked LC-2.0 specimen.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig5_HTML.gif
Fig. 5

The SEM micrographs taken from the surface of different LCCPCs before and after soaking in SBF (21 days) (accelerating voltage is 20 kV) along with the EDXA patterns of the soaked samples prepared from the regions marked by circle in the related SEM images

The EDXA patterns reveal the presence of Ca, P, C, O and Cl elements in the structure of all soaked cements. Ca, P and O comes from calcium phosphate salts used as starting materials and/or apatite crystals precipitated during soaking (in the cases of LC-2.8 and LC-2.4 specimens) Cl appears to present in very low amounts in EDXA patterns probably due to the trace amounts of chlorinated solvents used in processing the resins or even residue (CaCl2) for the SBF despite washing. It should be noticed that EDXA is not a quantitative experiment and since penetration depth of electron beam is larger than just the formed layer thickness on specimens, the elemental composition of lower layers are included in EDXA data.

3.4 Weight loss and molecular weight

Figure 6 indicates that all LCCPCs, after in vitro soaking, tend to decrease their masses with increasing time. The mass of the cements decreases slightly at the beginning and then by the increased constant rate thereafter. Additionally, the amount of mass loss of the cements increases with decreasing P/L ratio and thus the LC-2.8 specimen exhibits the least loss of its initial weight in compared to other samples. The results are consistent with results of SEM and mechanical tests where large pores are produced in cement structure during soaking led to decrease in compressive strength.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig6_HTML.gif
Fig. 6

Changes in weight of LCCPCs as a function of immersion time

Change in the molecular weight of the polymer component of LCCPCs in at the 1st day to day 21th of soaking is illustrated in Fig. 7. It can be seen from the results that the average-weight molecular weight of polymer in the composites with higher resin content reduced faster than that with lower resin phase. This result is in agreement with the trend of mass loss, which indicates that degradation of the cement increase with decreasing P/L ratio. The decrease in molecular weight may be resulted from hydrolysis and macromolecular scission of Poly-HEMA or PAMA.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig7_HTML.gif
Fig. 7

Changes in weight average molecular weight of polymer phase of the LCCPCs as a function of immersion time

3.5 Chemical groups and phase composition

Figure 8 shows the FTIR spectra of both cured and 21-days-soaked LC-2.8 specimen. The spectrum of cement liquid phase is also shown for comparison. These FTIR patterns help us to understand the mechanism of cement hardening and chemical characteristics of the crystals formed on the surface of sample in SBF. Table 2 lists important chemical groups with their related absorption wavenumber in these spectra.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig8_HTML.gif
Fig. 8

The FTIR transmission spectra of (a) resin phase of cement, (b) cured cement (LC-2.8) after setting and (c) after soaking in SBF solution for 21 days

Table 2

Important chemical groups with their related absorption wavenumber appeared in the FTIR spectra of resin phase, cured cement and LCCPC soaked in SBF solution

Material

Group

Wavenumber (cm−1)

Resin

C=CH2

813

C=C

1650

C=O

1710

CH

2950

Cured cement

PO4

563, 603, 950, 1060

COO salt

1330–1450, 1550–1650

C=O

1710

CH

2950

Soaked cement

PO4

563, 603, 950, 1060

CO3

865, 1400, 1450

H2O

1590, 3400

OH (in apatite lattice)

630

HPO4

985

CH

2950

From the position and assignment of the peaks, the following features can be pointed out: For resin, the bands appeared at 813 and 1,650 cm−1 are respectively assigned to C=CH2 and C=C groups in HEMA monomer (Fig. 9), which are disappeared in the pattern of hardened (cured) cement. It confirms curing process of the cement through polymerization reaction. The bands around 1,330–1,450 and 1,550–1,650 cm−1 are assigned to the symmetric and asymmetric stretches of the carboxyl group of calcium carboxylate salt [20].
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig9_HTML.gif
Fig. 9

Chemical structure of HEMA monomer

For the soaked LC-2.8, the FTIR spectrum reveals the formation of phosphate and carbonate groups that are individually found in apatite lattice. The bands observed at 562 and 603 cm−1 and 1,060 relate to stretching mode of PO43− in apatite crystal and the band at 865 cm−1 is assigned to vibrational mode of CO32− substituted for PO43− group [21]. The bands observed at 1,416 and 1,457 cm−1 can also confirm the substitution of CO32− groups for PO43− in apatite lattice [21]. Appearance of the absorption band at around 630 cm−1 indicates the formation of hydroxylated apatite phase during the soaking process, even though the resulted phase may not be stoichiometric in chemistry, i.e. Ca10(PO4)6(OH)2.

Another considerable point in FTIR spectrum of the soaked LC-2.8, compared to unsoaked sample, is reduction of its carbonyl and CH band intensities, which confirms typical degradation of polymeric phase. In other words, omission of these bands indicates that the polymeric phase of the LCCPC is degraded which confirms the results of SEM images. In other words, using the FTIR spectra, it can be stated that the pores observed in the SEM images of 21-soaked samples is obtained due to the leaching out of the polymerized resin and poly salts into the SBF solution during soaking period.

Figure 10 shows the changes in phase composition of LC-2.8 samples before and after soaking in SBF solution for various time schedules. Diffraction pattern of powder phase of the cement is also presented for comparison. The set cement specimen has similar XRD data to the cement powder after setting and before soaking in SBF solution, except that noise-like fluctuations found in LC-2.8 pattern, which are probably due to the amorphous nature of polymerized resin and setting reaction products (see comments on FTIR results). Signs of apatite phase are observed in XRD pattern of the sample after immersion in SBF for 7 days. Increase in amount of apatite phase is observed with increasing in soaking time; however TTCP and dicalcium phosphate phases also exist in cement composition even after 21 days. Regarding to the peak intensity of apatite and TTCP, equal amounts of these phase are suggested in LC-2.8 composition after soaking in SBF for 21 days.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig10_HTML.gif
Fig. 10

The XRD patterns of light cured cement (LC-2.8) samples before and after soaking in SBF solution for various periods

3.6 In vitro cell studies

Figure 11 shows the change in numbers of viable cells attached on LCCPCs as a function of time. After cell attachment, the osteoblasts cultured on all samples began a period of proliferation due to the significant difference in cell numbers between days 1, 7 and 14th (P < 0.05). Osteoblasts exhibit better proliferation on top of all LCCPCs because of significantly higher numbers of viable cells compared to control group. Furthermore, after 14 days, the number of cells proliferated on LC-2.8 specimens is significantly higher than that of LC-2.0.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig11_HTML.gif
Fig. 11

Numbers of viable cells attached on the surfaces of different LCCPCs after different culturing periods

Figure 12 shows the morphology of the osteoblastic cells well attached onto the surface of all LCCPCs after culturing for 14 days. The surfaces have been covered by spindle-like and confluent osteoblastic cells with elongated cytoplasmic membrane.
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig12_HTML.jpg
Fig. 12

SEM pictures of the rat calvarium-derived osteoblastic cells attached on the surface of LCCPCs after culturing for 14 days (magnification ×500)

The results of MTT assay done using extracts of different cements are shown in Fig. 13. In this figure, the amount of produced Formazan of osteoblasts at the presence of sample extract at each evaluating period has been normalized to the value of cells cultured at the presence of polystyrene. The extracts of samples collected at different times does not exhibit cytotoxic effect on osteoblasts, because no significant difference in number of viable cells is observed between control group and other samples even for samples kept in culture medium for 14 days (P ≫ 0.05).
https://static-content.springer.com/image/art%3A10.1007%2Fs10856-012-4637-z/MediaObjects/10856_2012_4637_Fig13_HTML.gif
Fig. 13

The results of MTT tests performed by extracts of different LCCPCs collected by immersing powdered cement in culture medium for 1, 7, and 14 days. Data have been normalized to the numbers of cells cultured on polystyrene plastics (control group)

4 Discussion

CPCs have been widely used as an alternative for reconstruction of bone defects. They are also known as bone tissue engineering scaffold because of their in vivo resorption nature [22]. Biodegradability of CPCs originates from the poor crystallinity, nanostructured nature and deviation from stoichiometry of the apatite crystals formed during setting as well as highly porous structure of the cements. However, hydraulic CPCs (c-CPCs) exhibit low biodegradation rate [23]. From another point of view, while the initial compressive strength of hydraulic CPCs is low (3–10 MPa, depending on particle size and P/L ratio [17, 24]), these materials become stronger after immersion in SBF (e.g. a value of about 17–25 MPa after 7 days storage in SBF [17, 25]) due to the growth and developed microstructure of apatite crystals. Thus, hydraulic CPCs are designed for non-load-bearing applications. However, an appropriate scaffold must have an enough mechanical strength in early stages of implantation until the newly formed tissue regained its lost strength created by degradation process [26]. Additionally, as an initial requirement of tissue engineering scaffold, degradable implant should gradually weaken as the natural tissue progress, because more rapid healing process is occurred when stresses are gradually transferred from the scaffold into the formed tissue [26].

In this study strong calcium phosphate-based cements were developed using a photo-curable resin. The resin phase of cement composed of HEMA and PAMA, two important biocompatible organic materials which are hydrophilic in nature. Polymerized HEMA has been used as a main component of glass-ionomer cements. Poly-HEMA has been successfully developed since 1960 as a biocompatible material and used in different fields of medicine including ophthalmology, cardiovascular, plastic surgery and otolaryngology [2729]. Also, it has a widespread usage as an adhesive and diluents in dental materials because it facilitates self-adhesion to mineral phase of dental tissues [30] and has a tendency to interact with HA particles which results in improvement of matrix/filler coherency [31].

In this study, it is suggested that the hardening behavior of LCCPCs occurs through two different mechanisms: (i) Polymerization of HEMA monomers based on the following reaction:
$$ n{\text{CH}}_{ 2}\,=\,{\text{C(CH}}_{3} ) {\text{COOCH}}_{ 2} {\text{CH}}_{ 2} {\text{OH}} \to [ {\text{CH}}_{ 2}{-} {\text{C(CH}}_{3} ) {\text{COOCH}}_{ 2} {\text{CH}}_{ 2} {\text{OH]}}_{n} $$
(2)
and (ii) formation of poly-carboxylate salts through the reaction of Ca ions with polyacid groups.

From the FTIR spectra of cured LC-2.8 cement (Fig. 8), absence of stretching bands at 813 and 1,650 cm−1 assigned to C=CH2 and C=C groups confirms the suggested polymerization phenomenon. Furthermore, bands around 1,330–1,450 and 1,550–1,650 cm−1 are assigned to the symmetric and asymmetric stretches of the carboxyl group of polycarboxylate salt [19].

Compared to c-CPC, LCCPCs with the same powder phase and photo-curable HEMA based resin exhibited much higher compressive strength after curing, while a continuous decrease in strength was observed after immersion in SBF solution (Fig. 3). Adverse behavior has been reported for c-CPCs with the same powder phase and particle size in which the compressive strength increased from ~3 MPa at 24 h after setting to 17 MPa at 7 days after soaking [17]. Superior mechanical strength of LCCPCs especially at P/L ratio of 2.8 g/mL is also comparable to that of non-resorbable bone cements based on PMMA (~80 MPa) [32] which are suggested for load-bearing applications. The elevated compressive strength of LCCPCs originates from its compacted microstructure (Fig. 5) in which calcium phosphate particles bind to polymeric matrix by specific organic–inorganic interactions such as covalent bond and chelating of Ca ions with some functional groups. The strength was greatly depends on powder content of cement and increased with increasing P/L ratio probably due to the higher concentration of organic–inorganic bonds. Loss in mechanical strength of soaked LCCPCs is due to leaching out of polymeric poly-carboxylate salts and polymerized residual resin phase as confirmed by change in weight of samples and molecular weight of polymer phase. Cement degradation introduces pores in the cement structure (Fig. 5). These pores with the size of 20–50 μm are appropriate spaces for penetration of macrophages, white cells and osteoclasts to induce active resorption as well as osteoblasts and blood vessels for development of new tissue. Poly-HEMA can absorb a significant amount (30 wt%) of water and may be an additional reason for loss in mechanical strength after immersion in SBF [33].

In addition to its superior mechanical strength and biodegradable nature, it is suggested that the light cured cements produced in this study have the ability to bond to living bone by means of biologically formed active apatite layer on its surfaces (Figs. 5, 10) though the rate of apatite formation on LCCPC is slower than that reported for c-CPC [17]. Precipitation of apatite crystals on the surfaces of LCCPCs may be ruled by concentration of TTCP/DCPA mixture in the cement paste because at low P/L ratio (LC-2.0) no precipitation was observed. The inhibited precipitation of apatite crystals on LC-2.0 and the limited precipitation or growth of apatite phase on LCCPC with P/L ratios of 2.4 and 2.8 g/mL can be discussed by the presence of PAMA in cements compositions. The inhibitory effect of poly acrylic acid on apatite precipitation has been reported elsewhere [34].

In vitro ageing is a good initial indicator of the ability of the material to form an interfacial bond with hard tissues such as bone [35, 36]. The mechanisms of apatite formation on surfaces of bioactive materials have been discussed in previous studies [37]. As observed in FTIR spectra of set LC-2.8 (Fig. 8), COOH groups are found in the cement structure (due to the presence of broad band at 1,710 cm−1). It has been reported that this functional group can induce formation of bonelike apatite layer on surfaces of materials because it acts as an active site for apatite nucleation [38]. LCCPCs may bond to bone through another mechanism that is chemical bond formed between free carboxyl groups of the cement and HA of bone mineral. Such bond is reported to be ionic and produced by replacement of PO43− ions of the substrate (apatite phase of bone) by carboxyl groups of material [39].

In vitro formation of nano-sized apatite crystals on cements surfaces shows that they can mimic the naturally occurring processes which occur in a biological setting, however more in vivo experiments is required to prove cements bioactivity.

Cell counting was performed to quantify proliferation of cells on sample which is related to cell viability. The results of MTT tests (Figs. 11, 13) show that the LCCPCs had no cytotoxic effect on osteoblasts and favorably, their composition and topographical features encourage proliferation of cells. The results is in agreement with literatures that demonstrates enhanced proliferation of osteoblasts on surfaces of HA-containing composites [40, 41]. It reveals that the cement component influenced the metabolism of the osteoblasts. However there are some studies which report suppressed proliferation of osteoblasts on HA bodies [42, 43]. Discrepancies may originate from differences in osteoblastic cell type, topographical parameters, materials crystallinity and solubility and cell culture conditions.

This study confirms that surface morphology of LCCPC modulates the adhesion and proliferation of the rat-derived osteoblastic cells because they were easily adhered and spread on the cement surface. Adhesion of osteoblastic cells to surfaces of implants is accepted that playing essential role in the process of osteoconduction [44].

5 Conclusions

Light cured and biodegradable CPCs with improved compressive strength were developed using TTCP/dicalcium phosphate powder and poly-HEMA/PAMA resin. The compressive strength of the cement is comparable to PMMA, but increase 6–16-fold compared to hydraulic CPC, depending on its P/L ratio. The LCCPC is gradually degraded during soaking in SBF solution and its compressive strength decreases due to the formation of macropores on its microstructure. Nano-sized apatite crystals can also precipitate on the surfaces of LCCPCs prepared at elevated P/L ratios when soaking them in SBF solution. It reveals that the cements can mimic natural processes which occur in a biological setting. The results of this study designate LCCPCs as potential alternatives for bone tissue engineering scaffold.

Acknowledgments

Authors wish to acknowledge Pasteur Institute and Medical College of Shahid Beheshti University for the cell experiments.

Copyright information

© Springer Science+Business Media, LLC 2012