Characterization, physicochemical properties and biocompatibility of La-incorporated apatites

Available online at www.sciencedirect.com

Acta Biomaterialia 5 (2009) 3512–3523 www.elsevier.com/locate/actabiomat

Characterization, physicochemical properties and biocompatibility of La-incorporated apatites D.G. Guo *, A.H. Wang, Y. Han *, K.W. Xu State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Xi’an, Shaanxi 710049, China Received 24 November 2008; received in revised form 12 May 2009; accepted 19 May 2009 Available online 27 May 2009

Abstract In this study, the physicochemical properties and biocompatibilities of La-containing apatites were intensively investigated together with their characterizations in terms of composition, structure, valent state and morphology using X-ray diffraction, Fourier-transform infrared spectra, X-ray photoelectron spectroscopy, scanning electron microscopy and energy dispersive X-ray spectroscopy, respectively. The results indicate that the La3+ ion can be incorporated into the crystal lattice of hydroxyapatite resulting in the production of La-incorporated apatites (LaxCa10x(PO4)6(OH)2+x2yOyhyx (x 6 0.5, y < 1 + x/2) or LaxCa10x(PO4)6Oyhyx (0.5 < x < 2, y = 1 + x/2)) by high-temperature solid phase synthesis. For La content <20%, the product is composed of the major phase, LaxOAP, as well as a small amount of tricalcium phosphate, but for a La content of 20%, the product is pure La-incorporated oxyapatite with the formula La2Ca8(PO4)6O2 (La2-OAP, x = 2, y = 2). It is also found that the La content plays important roles in both the physicochemical properties and biocompatibilities of the La-incorporated apatites. In contrast to La-free apatite, La-incorporated apatites possess a series of attractive properties, including higher thermal stability, higher flexural strength, lower dissolution rate, larger alkaline phosphatase activity, preferable osteoblast morphology and comparable cytotoxicity. In particular, the sintered La-incorporated apatite block achieves a maximal flexure strength of 66.69 ± 0.98 MPa at 5% La content (confidence coefficient 0.95), increased 320% in comparison with the La-free apatite. The present study suggests that the La-incorporated apatite possesses application potential in developing a new type of bioactive coating material for metal implants and also as a promising La carrier for further exploring the beneficial functions of La in the human body. Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. Keywords: La-containing apatite; Flexural strength; Degradability; Thermal stability; Biocompatibility

1. Introduction Hydroxyapatite (Ca10(P04)6(OH)2; HA), owing to its similarity to the inorganic phase of human bone and thus excellent bioactivity and biocompatibility, has become one of the most extensively researched and widely applied biomedical materials. In fact, biological apatite is not pure HA, but a non-stoichiometric HA with a general formula Me10xhx(PO4)6y * Corresponding author. Address: State Key Laboratory for Mechanical Behavior of Materials, Xi’an Jiaotong University, Department of Materials Science and Engineering, Xianning West Road, Xi’an, Shaanxi 710049, China. Tel.: +86 029 82668614; fax: +86 029 82663453. E-mail addresses: [email protected] (D.G. Guo), yonghan@ mail.xjtu.edu.cn (Y. Han)..

(HPO4)y(OH)22xh2x [1], where the Me/P atomic ratio can vary widely over the range 1.50–1.67. Also, in this formula, Me are often associated with a wide variety of trace ions such as Ca2+, Mg2+, Sr2+, Ba2+, Na+, K+, Pb2+, Zn2+, Cu2+, Fe3+, etc.; PO43 can be substituted by CO32, SiO44, P2O74; and OH can be substituted by F and Cl [2–4]. It has been demonstrated that the incorporation of some of these ions instead of equivalent Ca2+ or PO43 or OH into the apatite crystal lattice had significantly improved the physicochemical and biological properties of HA [5–9]. For example, introducing CO32 [5] or SiO44+ ion [6] into the crystal lattice of HA is effective in improving its degradation rate. It is worthy of attention that the incorporation of Sr2+ into HA instead of equivalent Ca2+ in the form of Sr-incorporated HA bone cement exhibits preferable biocompatibility, better bioactivity,

1742-7061/$ - see front matter Ó 2009 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.actbio.2009.05.026

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

higher compressive strength and a faster biodegradable rate than conventional Sr-free HA bone cement [7–9]. Rare earth elements (REE) are one type of important strategic resource widely used in various fields, including industry, agriculture, medicine and daily life, but eventually accumulated in the human body. Naturally, the effects of REE on human health are increasingly the concern of scientists. REE, including all lanthanide elements could be found in bovine whole blood reference material at a wide concentration range of 0.90– 1880 pg g1 [10]. In particular, lanthanum (La) is one of the most important REE widely researched in recent years. He et al. [11] evaluated the neurotoxicity of La by exposing Wistar rats to lanthanum chloride and found that La exposure played a dose-dependent role in brain functions. Huang et al. [12] investigated the effects of La on the femur bone mineral of male Wistar rats, and the result confirmed that La was finally accumulated in bone and would be lost with bone mineral degradation. More interestingly, La3+ promotes the formation of osteoclast-like cells and significantly increases the number and surface area of the resorption pits at the concentration of 1.00  108 mol l1, but inhibits bone-resorption activity at higher concentrations [13]. In addition, La3+ has also shown other beneficial functions, such as restraining the Ca pump of the human red cell [14], treating hyperphosphataemia [15], resisting or preventing cancer cells from defusing [16], depressing the release of insulin-like growth factor binding proteins (IGFBP) from cell types (>80% for GM10 and T98G cells and <65% for MDBK cells) [17] and mediating the effect on platelet function by decreasing the lipid fluidity of the surface membrane of human platelet [18]. Based on this existing research, the introduction of REE (especially for La) at controlled doses into some biomedical material or carriers could become one effective way to improve human health. Naturally, one question would be raised: what will happen to the structure and properties of La-containing apatite after La is incorporated into its crystal lattice instead of equivalent Ca2+? It was reported elsewhere [19] that La element could be incorporated into the crystal lattice of HA, but the characterizations presented for the La-incorporated apatite are poor, and the investigations in physicochemical properties and biological properties are also few. Thus, the present study characterized La-incorporated apatites with various La content by combining multi-techniques, and initially investigated their physicochemical and biological properties, which were considered for two potential purposes (1) to test the possibility of improving clinical serving functions of HA (e.g., serving as a bioactive coating of metal implants) via La incorporation; (2) to develop a potential carrier of REE in the form of REEincorporating apatite, through which the beneficial functions of these elements on human health would be further extended. 2. Materials and experiments 2.1. Starting materials and synthesizing methods The La-incorporated apatites with La content (i.e., the molar ratio of La/(Ca + La)) of 5%, 10% and 20%, respec-

3513

tively, were synthesized from a series of stoichiometric mixtures composed of calcium carbonate (CaCO3), diammonia hydrogen phosphate ((NH4)2HPO4) and lanthanum oxide (La2O3) through a high-temperature solid-state reaction. As a control, the La-free apatite was prepared by a similar method, but its precursor consisted of only CaCO3 and (NH4)2HPO4 in a stoichiometric mixture. The sintering steps and parameters shown in Table 1 are similar to those of Series A reported elsewhere [19]. The heating units of the high-temperature furnace were silicomolybdic rods. For the purpose of accelerating the solid-state reaction dominated by a diffusion mechanism, before sintering the stoichiometric mixtures were first refined by a ball-milling process with a liquid medium of absolute ethyl alcohol. The dosage of liquid medium used here was 30 ml in volume per 50 g raw powder. The materials of the milling ball and vial were alundum (Al2O3) and nylon, respectively, where the milling balls used for each milling include 100 smaller balls with a diameter of 5 mm, and 15 larger balls with a diameter of 10 mm, and the capacity of the vial was 200 ml in volume. The total mass and the mean size of the raw powders were 50 g and 30–50 lm, respectively, and the rotation rate was 400 rpm. The apparatus used was planet ball-milling machine (QM-1SP, Nanjing University Instrument Corp., China). After sintering, the products were crushed and then wetly ball-milled again for 24 h. The mean size of the final ground powder particles was in range 2.71–3.93 lm, determined by a laser particle analyzer (JL9200, Jinan Weina Device Ltd. Corp., China). 2.2. Flexural strength measurements A standard three-point flexural test with a span of 20 mm was used to fracture the specimens at a cross-head speed of 0.5 mm min1 on a computer-controlled high-frequency all-purpose fatigue testing machine (100HFP5100, Zwick/Roel Corp., German). The samples for testing flexural strength (FS) were processed as follows: first, the asprepared powders were pressed into the quadrate blocks of width 5 mm, height 4 mm and length 45 mm, where the pressure used for each pressing was 20 MPa, and polyvinyl alcohol with a suitable dosage was used as the additive for easy pressing; secondly, the pressed samples were sintered at 1500 °C for 2 h, followed by an in situ cooling process in a furnace. During testing, the loading– displacement curve was recorded on a computer. A mean FS value was determined from at least six samples under

Table 1 Sintering steps and parameters of La-free and La-incorporated apatite with different La content. La content (x, %)

1st step

0 5 10 20

950 °C 950 °C 950 °C 950 °C

(2 h) + 1100 °C (2 h) + 1100 °C (2 h) + 1100 °C (2 h) + 1100 °C

2nd step (48 h) (48 h) (48 h) (48 h)

– +1200 °C (24 h) +1200 °C (60 h) +1450 °C (120 h)

3514

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

the same processing conditions. Without other special description, the samples used for the following tests were pressed blocks or pressed discs or free powders which consistently experienced an additional sintering process at 1500 °C for 2 h.

mLi ¼ 100%ðmi  m0 Þ=m0

2.3. Porosity measurement methods

2.5. Characterization methods

The porosities, including open cell porosity, close cell porosity and total porosity, of the ceramic samples after the three-point flexural tests were measured by the Archimedes principle. The steps for measurement of the open cell porosity (Popen) were as follows (1) the tested sample in the form of a block was dried in a vacuum oven at 110 °C for 10 h and then was weighed on an electronic balance with a precision of 0.1 mg (denoted as mdry); (2) the dried sample was immersed in distilled water, and the water was heated at 100 °C for 2 h, after which the wet sample was slightly taken out, and then the residual water layers on the surface of the wet sample was carefully wiped off with fully wet cotton, followed by weighing again (denoted as mwet); (3) the wet sample was again immersed in distilled water and weighed under the water surface (denoted as mwater). The open cell porosity of the sample was calculated as follows:

Phase compositions of the powder samples were determined by X-ray diffraction (XRD) with Cu Ka radiation and an Ni filter, where the electrical voltage and electrical current used were 35 kV and 30 mA, respectively. The Fourier-transform infrared (FTIR) absorption spectra were detected using a Nexus 870 spectrometer. The fractured surfaces of the samples were observed by scanning electron microscopy (SEM; JSM6460, Japan). Energy dispersive X-ray (EDX) spectroscopy coupled with the SEM device was used to characterize the chemical elements in the various samples. The electron binding energies of the core levels of atoms on the sample surface were detected using X-ray photoelectron spectroscopy (XPS; AXZS ULTRA, Kratos Analytical Ltd., UK), where the electrical voltage and electrical current used were 15 kV and 30 mA, respectively. The precision of the binding energy determination was <0.1 eV.

P open ¼ ðmwet  mdry Þ=ðmdry  mwater Þ100%

ð1Þ

Also, the close cell porosity (Pclose) of the tested ceramic sample was calculated as follows: P close ¼ 1  P open  ðmdry d water =½d real ðmdry  mwater Þ100% ð2Þ where dwater is the density of water and dreal is the real density of the tested ceramic sample measured by the Archimedes principle. Here, the testing method of dreal was similar to steps (1) and (3) above, except that the tested sample was changed from block-like to fine powder-like (<50 lm). Then, the total porosity (Ptotal) was calculated as follows: P total ¼ P open þ P close

ð3Þ

2.4. In vitro evaluation of the degradation rate A series of static immersion experiments without stirring were performed in a physiological saline (0.9% NaCl solution) at 37 ± 1 °C for in vitro evaluation of the degradation (or dissolving) rates of various samples. The tested samples were the pressed discs (diameter 13 mm and height 2 mm) with different La content. The immersion time was 2 weeks, 4 weeks, 6 weeks, 8 weeks and 10 weeks, respectively, and the immersion solutions were refreshed every 2 days. After being taken out of the immersion solution after the pre-set immersion time, the immersed samples were first dried for 5 h in a vacuum oven working at 100 °C with a pressure of 1.0–1.2  101 Pa and then weighed on an electronic balance with a precision of 0.1 mg. The mass loss ratio (mLi) for each sample immersed for i weeks was calculated according to the following equation:

ð4Þ

where m0 denotes the mass of the tested sample before immersion, and mi the dried mass of the sample after immersion in physiological saline for i weeks.

2.6. Cell experiments 2.6.1. Cytotoxicity assay Testing for cytotoxicity was performed by the addition of different dilutions of powder sample extract to a L929 cell culture on a 96-multiwell plate. The tested samples were the La-free and the La-containing powders with La content 0%, 5%, 10% and 20%, respectively. The positive and negative controls used were 0.64% phenol solution and free culture, respectively. Five samples were tested for each group at the same time. Each sample in 1.0 g, sterilized by Co-60 rays, was individually poured into 100-ml glass flasks which were loaded with 10 ml Dulbecco’s modified Eagle medium (DMEM)BFS culture medium (pH 7.2) containing 10 vol.% bovine fetal serum. In total, four flasks were simultaneously incubated in a humid atmosphere of 5% CO2 at 37 ± 1 °C for 72 h. Then, the supernatant was filtered through a membrane and serial dilutions were made of 100, 50, 10 and 1 vol.% pure extract. The L929 cells were cultivated in a DMEM-BFS medium placed in an incubator at 37 ± 1 °C with a humid atmosphere of 5% CO2. The culture medium was taken out of the incubator, and the cells were washed with a calcium- and magnesium-free saline phosphate buffer (PBS-CMF). A 2.5 g l1 trypsin solution was then added to detach the cell layer from the bottle, and the cells were washed again with PBS-CMF and then re-suspended in DMEM-BFS. The suspension was adjusted to 1  104 cells ml1 and inoculated into a the 96multiwell culture plate, and then the plate was incubated for cell adhesion with 200 lm suspension for each well.

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

After cell adhesion for 24 h, the culture medium was removed, and each serial dilution and controls were added to the multiwell plate for inoculating L929 cells. Each concentration of tested extracts was made in six duplications to get a mean value. The plate was incubated in a humid incubator with 5% CO2 at 37 ± 1 °C for 72 h, after which 20 ll methyl thiazoly tetrazolium (MTT) solution at 5 g l1 was added into the well of the multiwell plate, and incubation was continued for 4 h in the same conditions. Thereafter, the supernatant of each well was carefully sucked and then added to 150 ll dimethyl sulfoxide (DMSO), followed by 10 min shaking. A DG3022 enzyme linked immunosorbent machine with a wavelength of 490 nm was used to determine the optical absorption data (OD), which were used to calculate the relative growth rate (RGR) of L929 cells with the following equation [20]: 1=2

RGR ¼ E=C  100%  s; s ¼ ½s2E  1=E2 þ ðs2C  E2 =C 4 Þ

ð5Þ where E is the absorbance of each group, C is the average absorbance of control, s is the standard deviation, sE is the standard deviation of E, and sC is the standard deviation of C. 2.6.2. Primary culture of osteoblasts The sequential collagenase digestion method was used to culture primary osteoblasts [21]. Cells were obtained from the calvaria of neonatal SD rats (1–2 days) after removing the soft tissue. The bone was washed with PBS three times and then digested for 2 h with insulin. After the supernatant was eliminated, the bone was digested twice with bone digestion liquid within 1 h. The second supernatant was collected and then centrifuged to gather the cells. Finally, the gathered cells were cultured in DMEM (Gibco) supplemented with 10% NBS at 37 ± 1 °C under a humid atmosphere of 5% CO2. 2.6.3. Cell adhesion The samples tested for cell adhesion were La-free and Lacontaining apatite discs (£ 13  2 mm). After caking at 120 °C for 20 min, the discs were taken out and cooled under aseptic conditions. The sterilized discs were preliminarily immersed in PBS-CMF for 4 h and then placed into the 24multiwell culture plate. Then 1000 ll (2  104 cells) of the cell dispersion was inoculated into the samples (6 wells/ group) and cultured at 37 ± 1 °C under a humid atmosphere of 5% CO2 for 30 min, 60 min and 120 min, respectively. The culture plate without a sample was used as the negative control. The viable cells were determined by a modified MTT assay method. After the original culture medium was removed with a macro-pipette, 1000 ll of the fresh culture medium was added to each well, and 200 ll of the MTT solution (5 mg ml1) was added to each well of the culture plate and incubated at 37 ± 1 °C under a humid atmosphere of 5% CO2 for 4 h. The supernatant of each well was carefully sucked and then added with 600 ll/eye DMSO, followed

3515

by 10 min shaking. Then 200 ll of the solution was transferred into another 96-multiwell culture plate, and finally the OD at a wavelength of 490 nm were determined. 2.6.4. Cell ALP assay The samples tested for alkaline phosphatase (ALP) activity assay were also La-containing or La-free apatite discs (13 mm diameter  2 mm). After being sterilized, these samples were put into the 24-multiwell culture plate. Then 1000 ll (5  104 cells) of the cell dispersion was inoculated into the samples (6 wells/group) and cultured for 5 days. The culture plates without samples served as the negative controls. In each group, osteoblasts were lysed using distilled water, and then three freeze–thaw cycles were performed on the 5th day. The classics method of Lowry et al. [22] was used to assay the ALP in the cell lysates. Aliquots (100 ll) of the distilled water or supernatants of each group were incubated with 500 ll of the reaction solution containing 2-amino-2-methyl-propanol (pH 10.0) and p-nitrophenylphosphate at 37 ± 1 °C for 40 min. The conversion from p-nitrophenol to p-nitrophenylate was stopped by the addition of 50 ll of 1 mol l1 NaOH. The OD of these samples were measured on a spectrophotometer (MR600 Spectrophotometric Microplate Reader; Dynatech) at 410 nm. 2.6.5. Cell morphology Two samples from each group were fixed on the 3rd day in 2.5% glutaraldehyde/1% paraformaldehyde in 0.1 ml sodium cacodylate buffer, and the fixation buffer was then replaced by guanine–HCl–tannic acid for 15 min and washed twice in distilled water. The cells were further fixed with 1% osmium tetroxide in fixation buffer for 30 min, rinsed in sodium cacodylate buffer, dehydrated in a graded series of ethanol, and finally immersed in hexamethyldisilazan for 3 min. The samples were finally dried in air and examined by SEM (JSM-840, Japan) after being goldcoated.

Fig. 1. XRD patterns of as-prepared La-free and La-containing products with different La content.

3516

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

2.7. Statistical analysis Significant differences between the groups were identified by the t-test with SPSS. The family confidence coefficient (cc) was given as 0.95. 3. Results Fig. 1 shows the XRD patterns of the as-prepared products with La content 0%, 5%, 10%, 20%, respectively. The major diffraction peaks and diffraction angles of the La-free products were well consistent with those of pure HA in standard JCPDS cards and, in addition, one weak diffraction peak located at 30.962° corresponding to (0 2 1 0) crystal plane of b-TCP was also detected. With the increase in La content, all the diffraction peaks of the products identifiably shifted towards the direction of lower diffraction angles. Moreover, all the diffraction peaks and the positions of the 20% La-containing products were well consistent with those of pure La-incorporated OAP (La2Ca8(PO4)O2; La2-OAP),

the only La-incorporated OAP phase presently found in standard JCPDS card base. In addition, with the increase in La content, the diffraction peak intensity of the (2 1 1) plane of apatite gradually increased, while those of the (1 1 2) and (3 0 0) planes slightly decreased, showing a slight effect of La made on the preferential orientation of apatite crystal. Fig. 2 shows the EDX spectra of elements on the surface of the as-prepared products. The element characteristic peaks corresponding to Ca, La, O and P elements could be clearly detected and, in particular, the intensities of the La peaks of the products gradually increased as the La content increased. Fig. 3 shows the FTIR spectra of the as-prepared products with various La content. Both the stretching and librating bands of OH– located at 3571 cm1 and 631 cm1, respectively, were clearly observed (see Fig. 3a) in the FTIR spectrum corresponding to the La-free product. However, the two OH bands (see Fig. 3b–d) corresponding to the La-containing products basically diminished. This indicates that the substitution of Ca2+ by La3+ decreases the number

Fig. 2. EDX spectra of the as-prepared La-free and La-containing products with different La content: (a) 0%; (b) 5%; (c) 10%; and (d) 20%.

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

Fig. 3. FTIR spectra of the as-prepared La-free and La-containing products with different La content: (a) 0%; (b) 5%; (c) 10%; and (d) 20%.

of OH ions, contributing to the transformation of OH ion in the apatite structure into O2 ion [19]. In addition, an obvious La–O absorption peak appeared between 510 and 525 cm1 and, more interestingly, its peak intensity gradually grew, and the center of the absorption peak shifted from 510 to 526 cm1 as the La content increased from 5% to 20%. Fig. 4 shows the total XPS electron binding energy spectra of the atoms on the surface of the as-prepared products with various La content. Obviously, the intensities of both the La3d3/2 peak and the La3d5/2 peak gradually increase, while the intensities of both the Ca2s peak and the Ca2p3/2 peak slightly decrease with the increase in La content, which accords well with the above result obtained from EDX spectra. By Gaussian fitting, it was found that the O1s-electron binding energy (530.9–531.4 eV) and P2s-electron binding energy (189.9–190.3 eV) of the as-prepared products containing various La content are well consistent with those data detected in some typical apatite specimens such as Sr104xNa2xLa2x(PO4)6(OH)2, Sr10(PO4)6F2 (Sr-FAP), Sr10(PO4)6Cl2 (Sr-ClAP) and Sr10(PO4)6(OH)2(Sr-HAP) reported elsewhere [23]. This implies that the present electron binding energy data of all atoms on the surface of the as-prepared products with various La content could be comparable with those obtained in the literature [23]. In particular, the La3d5/2-binding energy of La atom corresponding to 5% La-containing products was 835.9 eV, which is significantly different from that of La2O3 (834.9 eV) but very close to that of Sr8NaLa(PO4)6OH (835.6 eV) and Sr6Na2La2(PO4)6OH (836.1 eV) reported in the same literature. In other words, the La3d5/2-electron states of La atoms in the present La-containing products are similar to those of La atoms obtained in other apatite-like specimens [23], and different from those of La atoms in the reactant La2O3. Therefore, combined with the above XRD and FTIR results, it could be deduced that La had been incorporated into the apatite crystal lattice instead of Ca, and the as-prepared La-containing products

3517

Fig. 4. The total XPS spectra of the atoms on the surface of as-prepared products with various La content. The electron binding energies of all atoms were calibrated by C1s peak (284.8 eV).

with La content <10% were composed mainly of La-incorporated apatite, except for mixing a little b-TCP, while the 20% La-containing product was pure La-incorporated OAP with a formula of La2Ca8(PO4)6O2 (La2-OAP). Fig. 5 shows the XRD patterns of the La-incorporated and La-free apatite blocks after further sintering at 1500 °C for 2 h. In comparison with the XRD patterns of the as-prepared products shown in Fig. 1, some diffraction peaks of a-TCP with considerable intensities obviously appeared on the patterns of the sintered La-free apatite blocks and, simultaneously, only one a-TCP peak located at 30.74° could be clearly identified on the patterns of all the sintered La-incorporated apatite blocks. This indicates that the degree of decomposition of both the La-incorpo-

Fig. 5. XRD patterns of pressed La-free and La-incorporated products with different La content after further sintering at 1500 °C.

3518

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

Fig. 6. The flexure strengths of pressed La-free and La-incorporated products with different La content after further sintering at 1500 °C.

rated and La-free apatite enlarges slightly at 1500 °C, and also the b-TCP is basically transferred into a-TCP. Fig. 6 shows the FS of the La-incorporated and La-free apatite blocks after further sintering at 1500 °C for 2 h. The La-incorporated apatite blocks achieved the highest FS value (66.69 ± 0.98 MPa) at La content of 5%, and no obvious statistical difference was observed on the FS values of the La-incorporated apatite blocks for La content between 10% and 20% (cc = 0.95). By contrast with that

of the La-free HA block, the FS values of the La-incorporated apatite blocks increased 2.44–3.20-fold. Fig. 7 shows SEM photographs of the flexural fractured surfaces of the sintered La-incorporated and La-free apatite blocks. Some obvious pores and cracks on a micrometer scale were observed in the fractured surfaces of the Lafree apatite blocks. No obvious cracks but fewer pores were observed in the fractured surfaces of the La-incorporated apatite blocks. From the porosity data shown in Table 2, both the close cell porosity and the total porosity of the La-incorporated apatite decreased slightly with the increase in La content, consistent with the above SEM observations. Also, there were identifiable differences in both the crystal size and crystal shape between the La-incorporated and La-free apatite blocks. The La-incorporated apatite crystals exhibited a polyhedron-like shape, while the Lafree apatite crystals took on a strip-like shape, the latter probably caused by the decomposition of HA. In addition, the crystal size of the 5% La-incorporated apatite was much larger than that of the other La-incorporated apatites and La-free apatite blocks, indicating that La plays a slight role in the growth of La-incorporated apatite during sintering. Fig. 8 shows the mass loss ratios of the La-incorporated and La-free apatite discs after immersion in 0.9% physiological solution for different times. The mass loss ratios of all La-incorporated and La-free apatite blocks gradually increased as the immersion time was prolonged (cc = 0.95). Moreover, at each immersion stage, the mass loss ratio of

Fig. 7. The SEM photographs of the fractured surfaces of pressed La-free and La-incorporated apatite discs with different La content after further sintering at 1500 °C: (a) 0%; (b) 5%; (c) 10%; and (d) 20%.

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

3519

Table 2 Open pore percentages (Popen), closed pore percentages (Pclose) and total pore percentages (Ptotal) of La-free and La-incorporated apatite with different La content. [La/(La + Ca)]100%

0%

5%

10%

20%

Popen Pclose Ptotal

5.41 ± 2.26 8.63 ± 1.32 14.04 ± 0.96

1.57 ± 0.56 6.70 ± 0.76 8.27 ± 0.35

1.08 ± 0.07 4.80 ± 0.10 5.88 ± 0.04

2.55 ± 0.18 3.32 ± 0.02 5.87 ± 0.20

the apatite block decreased remarkably with the increase in La content (cc = 0.95), indicating that the incorporation of La instead of equivalent Ca into the apatite crystal structure significantly decreases its degradation rate. Fig. 9 shows the RGR percentages of L929 cells incubated in various extracts of the La-incorporated and La-free apatite powders. All the extracts with various concentrations exhibited no or low cytotoxicity with a score of 0 or 1 grade according to the six-grade criteria of the Chinese National Standard for Medical Devices and Implants-GB/T 161751996 [24]. With careful comparison, the RGR percentages of the extracts at lower concentrations were slightly larger than those of the extracts at higher concentrations, but no significant difference in RGR percentage was observed among the extracts of the La-free and La-incorporated apatite in the same concentration. Fig. 10 shows the OD values of ALP activity after incubating L929 cells in the extracts of the La-incorporated and La-free apatite discs for 5 days. Based on the statistical analysis, almost no significant variation was found in the mean OD values of the ALP activity corresponding to the extracts of La-free and La-incorporated apatite powders at La content <20%, but a slight increase was observed in the mean OD values of the ALP activity corresponding to the extracts of 20% La-containing apatite powders (cc = 0.95). In addition, the mean OD values of the ALP activity corresponding to the extracts of all the tested powders were slightly lower than that of the control. The result

In past decades, a considerable amount of research work involved the incorporation of various ions by replacing Ca2+ or PO43 or OH into the HA crystal lattice, which has become one of the most important ways of improving the physicochemical and biological properties of HA [5–9]. It has also been reported that the REE, including the La element in particular, exhibits many beneficial effects on

Fig. 8. The mass loss ratios of the pressed La-free and La-incorporated apatite discs (sintered at 1500 °C for 2 h) after immersion in 0.9% physiological solution for different times.

Fig. 9. The RGR percentages of L929 cells incubated in various extracts of the La-free and La-incorporated apatite powders (sintered at 1500 °C for 2 h) with different La content.

indicates that incorporation of La instead of equivalent Ca into the apatite has a slightly positive effect on the function of osteoblasts. Fig. 11 shows the morphologies of L929 cells incubated on the surface of the pressed La-incorporated and La-free apatite discs for 3 days. Many living osteoblasts were found on the surfaces of these tested samples. Some of them exhibited an approximately triangle-like shape on the surfaces of the La-free (see Fig. 11a) and 5% La-containing apatite samples (see Fig. 11b), and the others took on a spindle-like shape, with long mini-filopodias spreading (see Fig. 11c and d) on the surfaces of the 10% La- and 20% La-containing apatite samples. More cells with more rich cellular plasma were observed on the apatite samples with greater La content. This indicates that the incorporation of La element promotes the proliferation and adhesion of osteoblast, well consistent with the above results of the ALP activity test. 4. Discussion

3520

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

Fig. 10. OD values of ALP activity after L929 cells incubated for 5 days in extracts of the pressed La-free and La-incorporated apatite discs (sintered at 1500 °C for 2 h) with different La content.

human health [10–18]. So, the question is whether the incorporation of REE into HA in the form of REE-incorporated apatites plays an important role in the physicochemical and biological properties of HA. Based on this original idea, the authors synthesized and characterized the La-incorporated apatites with different La content via a high-temperature solid-state reaction process, and initially investigated their physicochemical and biological properties in the present study. Up to now, only a few contributions have mentioned the preparation and characterization of La-incorporated apatite [19,25]. Usually, the

wet method is one of the simplest ways of synthesizing conventional HA powder. Therefore, some researchers naturally considered that the La-incorporated apatites could be obtained using the conventional wet method [25], but there was a lack of evidence to demonstrate whether La had really been incorporated into the crystal structure of HA. Actually, the present authors had also made many unsuccessful attempts to synthesize La-incorporated apatite using the conventional wet method before starting the present study. It might be speculated that a higher external energy was needed to modify the crystal structure of HA by 0 ˚ ) with larger La replacing smaller Ca0 ions (radii = 0.99 A ˚ ), which is probably difficult to ions (radii = 1.016 A achieve just in the hydrous solution. However, successful synthesis process of La-incorporated apatite was realized through a long-term solid-state reaction at high temperatures [19], which to some degree provides experimental proof of the above speculation. In the present study, a similar high-temperature solidstate reaction was also performed to synthesize the Laincorporated apatite for the purpose of investigating its physicochemical and biological properties for different La content. However, a small amount of TCP was always observed in the La-incorporated apatites at lower La content or in La-free apatite, which is different from the result reported elsewhere [19], except that only pure La2-OAP was gained at a high La content of 20%. The result that the incorporation of La instead of Ca significantly enhanced the thermal stabilities of the apatite crystal structure is well in agreement with the result reported elsewhere

Fig. 11. The SEM photographs of L929 cells incubated for 3 days in extracts of pressed La-free and La-incorporated apatite discs (sintered at 1500 °C for 2 h) with different La content: (a) 0%; (b) 5%; (c) 10%; and (d) 20%.

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

[19]. Interestingly, it differs greatly from the influence of Sr content on the thermal stabilities of the apatite crystal structure found in a previous study [26], where the incorporation of Sr instead of Ca into apatite crystal structure contrarily decreased its thermal stabilities. In addition, by combining the present result of XRD, XPS, EDX and FTIR and also the valuable analyses made in the literature [19], the formulas of the La-free apatite and La-incorporated apatite were Ca10(PO4)6(OH)22yOyhy (y 6 1) and LaxCa10x(PO4)6(OH)2+x2yOyhyx (x 6 0.5, y < 1 + x/2) or LaxCa10x(PO4)6Oyhyx (0.5 < x 6 2, y = 1 + x/2), respectively. With the increase in La content, some important variations were identifiably detected in the La-incorporated apatite, including the transformation of OH ion into O2 ion, a major diffraction peak shift, preferential orientation of the crystal plane and other aspects reported elsewhere [19]. One major reason for the diffraction peak shift is attributed mainly to the increase in the crystal plane space of apatite as a result of the substitution of larger La ˚ ) for smaller Ca ions (0.99 A ˚ ). ions (1.016 A In the present study, the FS values of the La-incorporated apatite blocks with different La content were determined (see Fig. 6), and it was first found that the FS values of the Laincorporated apatite blocks were remarkably higher than those of the La-free apatite blocks. In particular, the FS values of the La-incorporated apatite blocks with La content 5% increased 3.20-fold in comparison with that of the Lafree HA block. In general, the composition and the microstructure, including porosity, crystal size, crack, etc., are important factors which affect the mechanical properties of ceramic materials. By comparing the porosity data shown in Table 2, one conclusion could be drawn that the incorporation of La into an apatite structure promotes densification of its microstructure during sintering, which is quite favorable for its strength. However, observed from SEM photographs shown in Fig. 7, the crystal size of the apatite first increases and then decreases with the increase in La content. It is worthy of attention that the crystal size of the 5% Laincorporated apatite was much larger than those of the other La-incorporated apatites and La-free apatite blocks. Although one a-TCP peak located at 30.74° Could be clearly detected on the patterns of all the sintered La-incorporated apatite blocks (see Fig. 5), no obvious difference in the content of a-TCP could be identified. In summary, by comparing the 5% La-incorporated apatite with the other La-incorporated apatites, although both the total porosity and crystal size of the former are identifiably larger than the latter, the former achieves a higher FS value. As a matter of fact, this result confirms that La content is the key factor affecting the mechanical strength of La-incorporated apatite. In addition, the La-free HA block had the lowest FS values in comparison with all the La-incorporated apatite blocks, probably attributed to its higher degree of decomposition, larger porosity and no La incorporation. Also, it has been considered that the decomposition of sintered calcium phosphates at high temperature leads to a decrease in its densification [27].

3521

Interestingly, the tendency for influence of La content on the FS of apatite is quite similar to that of Sr content on the compressive strength of apatite bone cement reported in a previous study [7], where apatite bone cement also achieves a higher compressive strength at an Sr content of 5%. It has even been speculated that the increase in the strength of apatite by the incorporation of ions might be caused mainly by crystal lattice strengthening or crystal lattice distortion [7,28]. However, the speculation was still short of experimental proof, and the mechanism of La incorporation affecting the strength of apatite needs to be further elucidated in the future work. The degradation rate of a would be implant or coating material is always of great concern. In this study, the Laincorporated apatite exhibited a much lower dissolution rate than the La-free apatite and, with the increase in La content, the degradation rate of La-incorporated apatite gradually decreased. This tendency for variation is very different from the influence of Sr on the degradation rate of Sr-containing HA [9]. But from the viewpoint of the in vivo serving life, the La-incorporated apatite with a much lower dissolution rate would become a better potential coating material for metal implants such as titanium and its alloys in comparison with the conventional HA coating [29]. In fact, apatite belongs to the ionic crystals, and its energy of crystal lattice has a tight relationship with the ion radius and its electric charge [30]. With increasing ion charge and decreasing ion radius, its crystal lattice energy is thus enhanced. In comparison with bivalent ˚ , the ion charge of trivaCa2+ with an ion radius of 0.99 A lent La3+ apparently increases by 50%, though there is a ˚ ). Therefore, from slight increase in La ion radius (1.016 A the viewpoint of the crystal lattice energy [31], it could be understood that La-incorporated apatite possesses a lower dissolution rate and a higher thermal stability in comparison with La-free HA. The concentration of La3+ in vivo released from a certain La-containing carrier or La-containing compound is a key factor affecting human health. For example, at a concentration <1.00  108 mol l1, La3+ promoted the proliferation, differentiation and function expression of osteoblasts but, at a higher concentration (>1.00  104 mol l1), it produced obvious damage in microvillus, membrane, endoplasmic reticulum and caryon of the osteoblasts [32]. So far, it is still not quite clear in which precise concentration range La is beneficial to human health. However, the present in vitro results show that La-incorporated apatite samples exhibit a comparable RGR, higher ALP activity, excellent cell morphology and cell proliferation, and cell adhesions in comparison with La-free HA. In other words, La-incorporated apatite sample with La content <20% has a qualified cytotoxicity, and excellent biocompatibility and bioactivity. Certainly, the present results, to some degree, are related to the lower degradation rate of La-incorporated apatite, which leads to a smaller amount of La released and is thus safe for the tested cells. Therefore, important work is worth performing in a further study to investigate the La release

3522

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523

behavior of La-incorporated apatite and the influence of La concentration on osteoblasts. An ideal bioactive coating material is always expected to possess a series of excellent serving functions: higher thermal stability (especially for coating prepared from the plasma-spray technique or other high-temperature process), lower dissolution rate, higher mechanical strength, excellent biocompatibility and bioactivity, and other preferable cytology properties. Therefore, the present study first provides valuable experimental proof of the potential application of La-incorporated apatite in biomedical engineering fields, e.g., as a potential carrier for exploiting the beneficial functions of the element La in the human body and a prospective bioactive coating material for metal implants such as titanium and its alloys instead of conventional pure HA coating. 5. Conclusions La-incorporated apatites with different contents were synthesized via a high-temperature solid-state reaction process, and their composition, microstructure and element electron state were characterized. In particular, their physicochemical properties and biocompatibilities were first evaluated. It was found that La-incorporated apatite exhibits a higher FS, lower dissolution rate, higher thermal property, qualified cytotoxicity, and excellent biocompatibility and bioactivity in comparison with conventional La-free HA. These excellent properties suggest that Laincorporated apatite possesses at least two prospective clinical applications: as a bioactivation coating material in metal implants; and as the La carrier exerting a beneficial effect in the human body at a controlled release rate. Further work is needed to study the releasing behavior of La loaded in La-incorporated apatite, the influence of La concentration on osteoblasts, and the bonding characters of this new biomaterial with bone tissue. Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 50702043), the Ph.D. Program Foundation of the Ministry of Education of China (20070698092) and the High Technology Research and Development Program of China (2006AA03Z447). The authors would like to thank Ms. Wei Yanping and Mr. Zhao Yantao, doctorial candidates at the Fourth Military Medical University, who are acknowledged for the results of the cell tests in this paper. References [1] Lerous L, Lacout JL. Preparation of calcium strontium hydroxyapatites by a new route involving calcium phosphate cements. J Mater Res 2001;16:171–6.

[2] LeGeros RZ. Calcium phosphates in oral biology and medicine. In: Myers HM, editor. Monographs in oral science. Karger: Basel; 1991. p. 31. [3] Blake GM, Zivanovic MA, Mcewan AJ. Sr-89 therapy strontium kinetics in metastatic bone disease. J Nucl Med 1986;27:1030–5. [4] Blake GM, Zivanovic MA, Mcewan AJ. Sr-89 strontium kinetics in disseminated carcinoma of the prostate. Eur J Nucl Med 1986;12:447–51. [5] Tang RK, Henneman ZJ, Nancollas GH. Constant composition kinetics study of carbonated apatite dissolution. J Cryst Growth 2003;249:614–24. [6] Alexandra EP, Claudia MB, Maria AL, Jose DS, Serena MB, William B. Ultrastructural comparison of dissolution and apatite precipitation on hydroxyapatite and silicon-substituted hydroxyapatite in vitro and in vivo. J Biomed Mater Res 2004;69A:670–9. [7] Guo DG, Xu KW, Zhao XY, Han Y. Development of a strontiumcontaining hydroxyapatite bone cement. Biomaterials 2005;26(19): 4073–83. [8] Guo DG, Xu KW, Yue J, Han Y, Zhao XY. PH value evolution of a novel Sr-containing calcium phosphate cement pastes and cytotoxicities of their hardened bodies. J Inorg Mater 2005;20(5):1159–66. [9] Guo DG, Xu KW, Han Y, Zhao XY. The influence of the Sr doses on the in vitro biocompatibility and in vivo degradability of the single-phase Srincorporated HAP cement. J Biomed Mater Res 2008;86A(4):947–58. [10] Fujimori E, Hayashi T, Inagaki K, Haraguchi H. Determination of lanthanum and rare earth elements in bovine whole blood reference material by ICP-MS after coprecipitation preconcentration with heme-iron as coprecipitant. Fresen J Anal Chem 1999;363:277. [11] He X, Zhang ZY, Zhang HF, Zhao YL, Chai ZF. Neurotoxicological evaluation of long-term lanthanum chloride exposure in rats. Toxicol Sci 2008;103(2):354–61. [12] Huang J, Zhang TL, Xu SJ, Li RC, Wang K, Zhang J, et al. Effects of lanthanum on composition, crystal size, and lattice structure femur bone mineral of Wistar rats. Calcif Tissue Int 2006;78:241–7. [13] Zhang JC, Zhang TL, Xu SJ, Wang K, Yu SF, Yang MS. Effects of lanthanum on formation and bone-resorbing activity of osteoclastlike cells. J Rare Earths 2004;22(6):891–5. [14] Luterbacher S, Schatzmann HJ. The site of action of La3+ in the reaction cycle of the human red cell membrane Ca2+-pump ATPase. Experientia 1983;39:311–2. [15] Alastair JH, Mary S, Fouad AB. Reducing high phosphate levels in patients with chronic renal failure undergoing dialysis: a 4-week, dose-finding, open-label study with lanthanum carbonate. Nephrol Dial Transplant 2004;19:1902–6. [16] Jiang WH, Chen D, Hao LM, Meng XT. Influence of light REE compound (LaCl3) on the growth of liver cancer cells. Stud Trace Elem Health 2005;22(3):1–3. [17] Robert HMC, David RC. Use of lanthanum to accurately quantify insulin-like growth factor binding to proteins on cell surfaces. J Cell Biochem 1997;66:256–67. [18] Sauerheber RD, Zimmermann TS, Esgate JA, VanderLaan WP, Gordon LM. Effects of calcium, lanthanum, and temperature on the fluidity of spin-labeled human platelets. J Membr Biol 1980;52:201–19. [19] Serret A, Cabanas MV, Vallet-Regi M. Stabilization of calcium oxyapatite with Lanthanum (III)-created anionic vacancies. Chem Mater 2000;12:3836–41. [20] Dion I, Rouais F, Baquey C, Lahaye M, Salmon R, Trut L, et al. Physico-chemistry and cytotoxicity of ceramics. Part I. Characterization of ceramic powders. J Mater Sci Mater Med 1997;8(5):325–32. [21] Zhao Y, Zhang Y, Ning F, Guo D, Xu Z. Synthesis and cellular biocompatibility of two kinds of HAP with different nanocrystal morphology. J Biomed Mater Res 2007;82B(2):533–44. [22] Lowry OH, Roberts NR, Wu M, Hixton WS, Crawford EJ. The quantitative histochemistry of brain II. Enzyme measurements. J Biol Chem 1954;207:19–37. [23] Shpak AP, Karbovskii VL, Vakhney AG. Electronic structure of isomorphically substituted strontium apatite. J Electron Spectrosc Relat Phenom 2004;137–140:585–9.

D.G. Guo et al. / Acta Biomaterialia 5 (2009) 3512–3523 [24] Chinese National Standard for Medical Devices and Implants-GB/T 16175-1996; 2003. p. 6–8. [25] Zhang YM, Fu T, Xu KW, An H. Wet synthesis and characterization of lanthanum-substituted hydroxyapatite. Chin J Conser Dent 2000;10(3): 149–51. [26] Guo DG, Xu KW, Han Y. Thermal stability of stoichiometric strontium-incorporated hydroxyapatite. Rare Metal Mater Eng 2006;35(3):379–82. [27] LeGeros RZ, LeGeros JP, Daculsi G, Kijkowska R. Calcium phosphate biomaterials: preparation, properties, and biodegradation. In: Wise DL, Trantolo DJ, Altobelli DE, Yaszemski MJ, Gresser JD, Schwartz ER, editors. Encyclopaedia handbook of biomaterials and bioengineering. Part A, vol. 2, No. 43. New York: Marcel Dekker; 1995. p. 1429–63.

3523

[28] Chen DM, Fu YF. Evaluation on the mechanic properties of the solid solution of strontium substituted hydroxyapatite. Chin J Dent Mater Dev 2001;10(4):178–9. [29] Lee Eung-Je, Lee Su-Hee, Kim Hae-Won, Kong Young-Min, Kim Hyoun-Ee. Fluoridated apatite coatings on titanium obtained by electron-beam deposition. Biomaterials 2005;2(18):3843–51. [30] Priest ND, Van de Vyver FL. Trace metals and fluoride in bones and teeth. Boca Raton, FL: CRC Press; 1990, p. 136. [31] Makishima A, Tamura Y, Sakaino T. Elastic moduli and refractive indices of aluminosilicate glasses containing Y2O3, La2O3 and TiO2. J Am Ceram Soc 1978;61:247–9. [32] Zhang JC, Li XX, Xu SJ, Wang K, Yu SF, Lin Qin. Effect of rear earth ion on the proliferation, differentiation and function expression of osteoblast in vitro cultured. Prog Nat Sci (Chin) 2004;14(4):404–9.