rsif.royalsocietypublishing.org

Research Cite this article: Ding Y, Tang S, Yu B, Yan Y, Li H, Wei J, Su J. 2015 In vitro degradability, bioactivity and primary cell responses to bone cements containing mesoporous magnesium – calcium silicate and calcium sulfate for bone regeneration. J. R. Soc. Interface 12: 20150779. http://dx.doi.org/10.1098/rsif.2015.0779

Received: 30 August 2015 Accepted: 9 September 2015

Subject Areas: bioengineering, biomaterials, biomedical engineering Keywords: mesoporous magnesium– calcium silicate, calcium sulfate, bone cement, degradability, cytocompatibility

Authors for correspondence: Jie Wei e-mail: [email protected] Jiacan Su e-mail: [email protected]

In vitro degradability, bioactivity and primary cell responses to bone cements containing mesoporous magnesium – calcium silicate and calcium sulfate for bone regeneration Yueting Ding1, Songchao Tang1, Baoqing Yu2, Yonggang Yan3, Hong Li3, Jie Wei1 and Jiacan Su2 1 Key Laboratory for Ultrafine Materials of Ministry of Education, East China University of Science and Technology, Shanghai 200237, People’s Republic of China 2 Department of Orthopaedics, Changhai Hospital, Second Military Medical University, Shanghai 200433, People’s Republic of China 3 College of Physical Science and Technology, Sichuan University, Chengdu 610041, People’s Republic of China

Mesoporous calcium sulfate-based bone cements (m-CSBC) were prepared by introducing mesoporous magnesium–calcium silicate (m-MCS) with specific surface area (410.9 m2 g21) and pore volume (0.8 cm3 g21) into calcium sulfate hemihydrate (CSH). The setting time of the m-CSBC was longer with the increase of m-MCS content while compressive strength decreased. The degradation ratio of m-CSBC increased from 48.6 w% to 63.5 w% with an increase of m-MCS content after soaking in Tris –HCl solution for 84 days. Moreover, the m-CSBC containing m-MCS showed the ability to neutralize the acidic degradation products of calcium sulfate and prevent the pH from dropping. The apatite could be induced on m-CSBC surfaces after soaking in SBF for 7 days, indicating good bioactivity. The effects of the m-CSBC on vitamin D3 sustained release behaviours were investigated. It was found that the cumulative release ratio of vitamin D3 from the m-CSBC significantly increased with the increase of m-MCS content after soaking in PBS (pH ¼ 7.4) for 25 days. The m-CSBC markedly improved the cell-positive responses, including the attachment, proliferation and differentiation of MC3T3-E1 cells, suggesting good cytocompatibility. Briefly, m-CSBC with good bioactivity, degradability and cytocompatibility might be an excellent biocement for bone regeneration.

1. Introduction Calcium sulfate-based bone cement (CSBC) has been widely applied in bone reconstruction owing to its excellent biocompatibility and resorption in vivo. However, CSBC cannot form chemical bonds between the calcium sulfate grafts and bone tissue because of its poor bioactivity [1,2]. In addition, previous studies pointed out that the degradation products (such as sulfate radical ions) of implanted CSBC would decrease the pH and produce an acidic microenvironment, which might lead to an inflammatory reaction with host tissues in vivo [3]. Mesoporous magnesium–calcium silicate (m-MCS)-based biomaterials with excellent bioactivity have emerged in recent years, which have stimulated the growth of osteoblasts and promoted osteointegration [4]. These characteristics made them one of the most intriguing materials in the biomedical field. The bredigite (Ca7MgSi4O16) with apatite layer induced in simulated body fluid (SBF) could support cell attachment and stimulate cell proliferation and differentiation [5]. It was reported that some ions (such as calcium, magnesium and silicon) released from these bioceramics could change the biological environment in vivo and promote bone reconstruction [6].

& 2015 The Author(s) Published by the Royal Society. All rights reserved.

2.1. Synthesis and characterization of m-MCS powders m-MCS powders were synthesized by the template method, using triblock copolymer of poly(ethylene glycol)-block-poly(propylene glycol)-block-poly(ethylene glycol) (P123, Sigma Aldrich Chemistry) as porogen; concentrated hydrochloric acid (HCl, 36% v/ v) as catalyst; ethyl silicate (TEOS, (C2H5)4SiO4), calcium nitrate tetrahydrate (Ca(NO3)2 . 4H2O) and magnesium nitrate hexahydrate (Mg(NO3)2 . 6H2O) as raw materials. All the chemicals were obtained from Shanghai Lingfeng Chemical Regent co., Ltd. In brief, P123 (6 g) was dissolved in deionized water (217 ml) until the solution was clarified [15], then HCl (17 ml) was dropped into solution with stirring for 30 min. After that, TEOS (12.9 g), Ca(NO3)2 . 4H2O (7.3 g) and Mg(NO3)2 . 6H2O (7.9 g) were added into the mixed aqueous solution with stirring for 4 h; the temperature was kept at 388C to form well-dispersed collosol, which was aged at 958C for 3 days and then dried at 808C. The m-MCS powders were obtained by sintering the

2.2. Preparation and characterization of m-CSBC CSH powders were obtained by sintering calcium sulfate dehydrate (CSD) at 1208C for 8 h, then raised to 1608C at a heating rate of 18C min21 and kept for 4 h. The m-CSBC pastes were prepared by mixing CSH and m-MCS (0 w%, 20 w% and 40 w%) powders with a ratio of cement powder to liquid (water) of 1 : 0.85 g ml21. The setting time of the cements was tested with a Vicat needle as mentioned in a previous study [16]. The cement pastes were injected into the Teflon moulds (Ø 12  2 mm). Then the cements were cured at 378C, 100% relative humidity for 5 days. The compressive strength of solidified specimens (Ø 6  6 mm) was measured using a universal testing machine (AG-2000A, Shimadzu Autograph, Shimadzu Co., Ltd, Japan) at a loading rate of 1 mm min21 [16]. The surface morphology and phase composition of the cement specimens were examined using scanning electron microscopy (SEMJSM-6360LV, JEOL, Japan) and XRD.

2.3. In vitro degradability of m-CSBC in Tris –HCl solution The degradation of m-CSBC was investigated by monitoring the weight-loss ratio and pH change of the solution after the samples were immersed in Tris – HCl solution ( pH ¼ 7.4) for different times (with a ratio of liquid to solid of 20 ml g21); the Tris – HCl solution was changed once a week. The initial weight (W0) of each sample was recorded. The samples were taken out at different time points (7, 14, 21, 42, 56 and 84 days) from the solution and dried at 808C for 48 h, and the dried cements were weighed again (Wt). The weight-loss ratio was calculated as weight loss ð%Þ ¼

W0  Wt  100: W0

The pH change of the solution after the cements were soaked for different times was measured by a pH monitor (PHS-3C, INESA, Shanghai Instrument).

2.4. Mineralization of m-CSBC in SBF The assessment of in vitro bioactivity of the cements was carried out in SBFs ( pH ¼ 7.4) [17]. The samples (Ø 12  2 mm) were immersed in SBF (liquid/cements ¼ 20 ml g21, 378C) by shaking. The specimens were collected after soaking in SBF for 7 days, then washed gently with distilled water and dried at 608C for 24 h. The surface morphology and composition were determined by SEM, EDS and XRD. The change of ion concentration (Ca, Mg, P and Si) in SBF solution after soaking 40 m-CSBC for different times was determined by inductively coupled plasma-atomic emission spectroscopy (ICP-AES, IRIS 1000, Thermo Elemental, USA).

2.5. Sustained release of vitamin D3 The vitamin D3 (0.05 g) was dissolved in 100 ml of ethanol, then the m-MCS powders were added to vitamin D3 solution (mMCS/vitamin D3 solution: 5 g 50 ml21) with strong stirring for about 2 h, centrifuged so as to remove the unabsorbed vitamin D3 in the supernatant, and then dried at 378C for 12 h. The

2

J. R. Soc. Interface 12: 20150779

2. Material and methods

dried powders at 2008C for 4 h and then raised to 6008C at a heating rate of 18C min21 and kept for 6 h. The morphology and composition of m-MCS were characterized by transmission electron microscopy (TEM, JEM-2010, JEOL Ltd, Japan), X-ray diffraction (XRD, D/max 2550 VB/PC, Rigaku Co., Japan) and energy dispersive spectrometry (EDS, JEOL6360LV, Japan). In addition, the pore diameter distribution, surface areas/pore volumes were determined by Brunauer – Emmett – Teller (BET, Tristar 3000, Micromeritics, USA).

rsif.royalsocietypublishing.org

The magnesium (Mg) ion is a small but important component of bone tissue: deficiency of Mg ions seems to be a risk factor for osteoporosis in humans [7], and Mg ions were reported to be associated with the mineralization of calcified tissues and could influence mineral metabolism through the activation of alkaline phosphatase (ALP) [8,9]. The degradation products of magnesium-containing materials were not only non-toxic but were also involved in various biological processes such as cellular processes, metabolism of vitamin D and enhanced osteogenic gene expression [10]. In addition, Feng et al. pointed out that silicates would neutralize the acidic degradation by-products and stabilize the pH of the surrounding environment [11]. Zhu et al. [12] mentioned that calcium and silicon ions could significantly induce the formation of a bone-like apatite layer. Mesoporous materials with high specific surface areas/ volumes could adsorb a large amount of drugs into their mesoporous structures and diffuse over time in a controlled manner [13]. Lu et al. [4] revealed that the ordered mesoporous magnesium–calcium silicate (om-MCS) would promote new bone formation by inducing apatite deposited on its surfaces. Zhang et al. [14] also explored that ceria (CeO2)-incorporated mesoporous calcium silicate materials had apatite-formation ability and sustained drug-delivery behaviours. Zhu et al. [12] figured out that mesoporous strontium-substituted calcium silicate (Sr-CaSiO3) materials had stimulated the proliferation and enhanced ALP activity of MC3T3-E1 cells. Considering the few reports of m-MCS-based biomaterials used in bone cement and drug delivery for bone generation, in this study, the novel mesoporous calcium sulfate-based cement (m-CSBC) was developed by incorporating m-MCS into calcium sulfate hemihydrate (CSH), and the effects of m-MCS content on the physical – chemical and biological properties of m-CSBC were investigated. It is reasonable that the m-CSBC might have an appropriate setting time, compressive strength, degradability, good bioactivity and cytocompatibility. Furthermore, the mesoporous channels of the m-CSBC allowed it to be used as a drug delivery agent during drug controllable release. Therefore, vitamin D3 was loaded into m-CSBC and its sustainable release behaviours were studied.

MC3T3-E1 cells were used to investigate the cytocompatibility of the cements (Ø 12  2 mm), which were sterilized with 100% ethanol for 6 h and UV radiated for 1 h; 300 ml of cell suspension was added to the samples at a density of 5  103 cells well21 in 24-well plates, and then incubated at 100% humidity with 5% CO2. At 6, 24 and 72 h, the medium was removed, and the specimens were gently rinsed with PBS three times and fixed with 4% glutaraldehyde for 2 h [18]. The fixative was removed by washing with PBS followed by sequential dehydration in graded ethanol (30, 50, 70, 90 and 100%). The morphology of the cell on the cements was observed by SEM.

2.7. Cell proliferation MC3T3-E1 proliferation was inspected using the methylthiazol tetrazolium (MTT) test. Briefly, the cells (5  103) were seeded on the cements (Ø 12  2 mm) and incubated in a humidified atmosphere of 95% air and 5% CO2 for different times. The culture mediums were replaced every 2 days. At 1, 3 and 7 days, 40 ml of a 0.5 mg ml21 MTT solution (Amresco, Solon, USA) was added to the well plate and incubated at 378C for 4 h. After 4 h of culture, the medium was removed and 200 ml of dimethyl sulfoxide (DMSO, Sinopharm, Shanghai, China) was added to each well to stop the reaction and completely dissolve the purple formazan product. An aliquot of 100 ml was taken from each well and transferred to a fresh 96-well plate. The absorbance was measured at l ¼ 590 nm on an ELISA plate reader (ELx800, BIO-TEK). A cell culture plate without materials was used as the blank control. All the results were presented as optical density values minus the absorbance of the blank wells.

2.8. Alkaline phosphatase activity The ALP activity assay evaluated the effect of the m-CSBC (Ø 12  2 mm) on osteogenic differentiation of MC3T3-E1 cells. A total of 2.5  104 of cells were seeded on each specimen in 24-well plates, then cultured at 378C in an atmosphere of 100% humidity with 5% CO2. At each culture point, the medium was removed and samples were washed with PBS three times. Then the cell lysate was obtained by adding 1 ml of 0.2% Nonidet P40 (NP-40) solution at room temperature. Aliquots of cell lysates were incubated with 50 ml of 1 mg ml21 p-nitrophenyl phosphate solution (Sigma, USA) at 378C for 15 min. The conversion of p-nitrophenyl phosphate to p-nitrophenol was stopped by adding 100 ml of 0.1 M sodium hydroxide solution. The ALP activity was normalized by the total intracellular protein contents and absorbance at 405 nm was measured with a spectrophotometer (SPECTR Amax 384, Molecular Devices). The ALP activity of MC3T3-E1 cells cultured on blank culture plates served as control.

A minimum of five samples was tested per group for physical and chemical properties, and six parallel samples per group for the cell experiments. p , 0.05 was considered statistically significant. All quantitative data were analysed with Origin pro 8.0 and expressed as the mean + standard deviation (M + s.d.).

3. Results 3.1. Characterization of m-MCS and m-CSBC From the TEM image (figure 1a), it can be seen that the mMCS showed ordered mesoporous channels and uniform pore size. From figure 1b, the m-MCS displayed a narrow pore size distribution, and the average pore size was approximately 9 nm. The BET results revealed that the specific surface area of m-MCS was 410.9 m2 g21 and pore volume was 0.8 cm3 g21. Figure 1c shows the EDS of m-MCS; clearly, the m-MCS consisted of Ca, Mg and Si elements. Figure 1d shows the XRD of the m-MCS, which was amorphous with a broad peak at 2u ¼ 248. The phase compositions of the m-CSBC after being cured for 5 days were characterized by XRD as shown in figure 2. It can be seen from figure 2a that the diffraction peak of CSD appeared at 2u ¼ 11.68, 20.78, 23.38, 29.18, 31.18 and 33.38, indicating that the CSH (CaSO4 . 1/2H2O) had transformed into CSD (CaSO4 . 2H2O) [2]. Correspondingly, the peak of CSD was found in 20 m-CSBC (figure 2b) and 40 m-CSBC (figure 2c), but the peak intensity of CSD significantly decreased with the increase of m-MCS content. Figure 3 shows the surface morphology of the cements. According to 1 k magnification images, the surfaces of the cements were dense and rough (figure 3a,c,e), and some m-MCS clusters were observed on the m-CSBC surfaces (figure 3c,e). From higher magnification images, CSBC presented as rod-like microcrystalline structures (figure 3b) while a large amount of amorphous m-MCS was evenly distributed onto the 20 m-CSBC and 40 m-CSBC (figure 3d,f ).

3.2. Setting time and compressive strength The setting time of the cements was prolonged with the addition of m-MCS (CSBC ¼ 5.3, 20 m-CSBC ¼ 8.3 and 40 m-CSBC ¼ 11.5 min) as shown in table 1. In contrast with the setting time, the compressive strength of the cements slightly decreased with the increase of m-MCS content; the compressive strength of CSBC, 20 m-CSBC and 40 m-CSBC was approximately 6.4, 5.8 and 5.5 MPa as shown in table 1.

3.3. Degradability of cements Figure 4a exhibits weight loss of the cements in Tris –HCl solution. It is found that the weight loss of the cements increased with time. The weight loss of CSBC (48.6 w%) was lower than 20 m-CSBC (59.3 w%) and 40 m-CSBC (63.5 w%) after soaking in Tris– HCl solution for 84 days. In figure 4b, the pH value of the solution for the CSBC decreased sharply within 14 days (from 7.4 to 6.96), and further decreased to 6.89 after soaking for 84 days, which presented an acidic solution. However, the pH of 20 m-CSBC and 40 m-CSBC showed mild changes (from 7.4 to 7.22, from 7.4 to 7.26) within 14 days, and gradually increased to 7.28 and 7.34 after soaking for 84 days.

3

J. R. Soc. Interface 12: 20150779

2.6. Cell attachment

2.9. Statistical analysis

rsif.royalsocietypublishing.org

cements loaded with vitamin D3 were prepared by mixing CSH and m-MCS powders containing vitamin D3 in different ratios. Then, the cements were put into a Teflon mould (F 12  2 mm) and cured at 378C, 100% relative humidity for 5 days. Using the same method, the CSBC, 20 m-CSBC and 40 m-CSBC containing the same vitamin D3content were prepared. For sustained release of vitamin D3 from the cements, the standard curve with certain concentrations of vitamin D3 was measured in order to make a regression equation. Then the specimens (CSBC, 20 m-CSBC, 40 m-CSBC) were immersed in PBS ( pH ¼ 7.4). The PBS solution (2 ml) was drawn out at each time point (1, 3, 5, 7, 10, 15, 20 and 25 days) and the OD value was tested by enzyme-linked immunosorbent assay (ELISA, SPECTRAmax 384, Molecular Devices, USA). The concentration and release ratio was calculated by regression equation and Origin pro 8.0 software.

(a)

(b)

4

0.30 0.25 0.20 0.15 0.10 0.05 0

100 nm

2

4

6

8 10 12 14 pore diameter (nm)

16

18

20

(d) 1400

(c) Si

intensity (CPS)

1200

Ca Mg

1000 800 600 400 200 0

1

4

2 3 energy (KeV)

10

20

30

40 50 2q (°)

60

70

80

Figure 1. (a) TEM image, (b) pore diameter distribution, (c) EDS and (d ) XRD of mesoporous magnesium –calcium silicate (m-MCS).

* (a) * intensity (CPS)

*

*

* (b)

10

(c)

*

*

*

*

*

* *

*

*

* *

* * 20

30

40

2q (°)

Figure 2. (a) XRD of CSBC, (b) 20 m-CSBC and (c) 40 m-CSBC after setting for 5 days; asterisks represent the peaks of calcium sulfate (CaSO4).

3.4. Apatite-formation on cements Figure 5 shows SEM images of surface morphology for the cements after soaking in SBF for 7 days. The lower magnification shows the ball-like apatite scattered on CSBC surface (figure 5a), while a lot of microspheres aggregates on the m-CSBC surfaces (figure 5c,e). Obviously, the apatite aggregates almost covered the 40 m-CSBC surfaces, which were denser than 20 m-CSBC and CSBC (dependent on the m-MCS content). In the higher magnification, the aggregates were found to be made up of regular microsphere crystalline

structures, which were the typical morphology of apatite. The apatite became more and more dense on the m-CSBC surfaces, which were m-MCS content dependent (figure 5b,d,f ). To further determine apatite formation on cement surfaces, XRD scans of the samples after soaking into SBF for 7 days are shown in figure 6a. It was found that the apatite peaks (at approx. 2u ¼ 25.58, 32.18, 42.48 and 49.48) appeared on all the samples [5], and other peaks appeared at approximately 2u ¼ 20.88, 23.58, 29.28 and 47.98, which belong to the cements. Clearly, the intensity of the peaks for 40 m-CSBC (figure 6a) and 20 m-CSBC (figure 6b) were significantly higher than CSBC (figure 6c), which was in accord with the SEM images. The EDS of the cements are shown in figure 6b. The results revealed the mole ratio of calcium to phosphorus was approximately 1.66, confirming the formation of the apatite layer on the 40 m-CSBC surface. Figure 7 reveals changes of Ca, Mg, P and Si ion concentrations in SBF after soaking the 40 m-CSBC for different times (12, 24, 72, 120 and 168 h). Ca ions increased sharply in the first 24 h, then declined slightly and gradually went up after 72 h. P ions kept declining, while Mg and Si ions sustained increases during the whole mineralization process.

3.5. Vitamin D3 sustained released from cements Figure 8 shows the curve for vitamin D3 sustained release with time. On the first day, the cumulative release ratio of vitamin D3 from CSBC was approximately 7.5 w%, while for 20 m-CSBC

J. R. Soc. Interface 12: 20150779

dV/dD pore volume (cm3 g–1)

rsif.royalsocietypublishing.org

0.35

(a)

5

(b)

rsif.royalsocietypublishing.org

20 m

J. R. Soc. Interface 12: 20150779

(c)

8m

(d )

8m

20 m

(f)

(e)

20 m

8m

Figure 3. SEM images of surface morphology of CSBC (a,b), 20 m-CSBC (c,d ) and 40 m-CSBC (e,f ). Scale bars represent 20 mm (a,c,e) and 8 mm (b,d,f ). Table 1. Setting time and compressive strength of CSBC, 20 m-CSBC and 40 m-CSBC (n ¼ 5).

samples CSBC 20 m-CSBC 40 m-CSBC

setting time (min)

compressive strength (MPa)

5.3 + 0.5

6.4 + 0.2

8.3 + 0.3 11.5 + 0.5

5.8 + 0.3 5.5 + 0.3

it was 6.6 w%, which was slightly higher than 40 m-CSBC of 3.6 w%. After that, however, the cumulative release ratio of vitamin D3 from CSBC was slow, and the m-CSBC displayed a higher cumulative release ratio of vitamin D3 with the increase of m-MCS content. At 25 days, the cumulative release ratio of vitamin D3 from 40 m-CSBC, 20 m-CSBC and CSBC was 41.3 w%, 30.6 w% and 17.5 w%, respectively.

3.6. Cell morphology and proliferation The cell morphologies on the cement surfaces for different times are illustrated in figure 9. It was found that the

osteoblasts stretched out and spread well on the coarse surface of the samples at 6 h (figure 9a–c). At 1 day, the osteoblasts were completely spread and tightly attached to the samples, and exhibited normal osteocyte morphology (figure 9d–f ). At 3 days, the osteoblasts formed a confluent layer with close attachment to the surfaces of the samples, as shown in figure 9g–i. The results indicated that all the samples had no negative effects on cell morphology and viability. The MTT assay was used to investigate cell proliferation; the results from figure 10 show that the OD value for 40 m-CSBC was higher than for 20 m-CSBC and CSBC at day 1, with the culture time extension, the OD value for 40 m-CSBC and 20 m-CSBC significantly higher than CSBC on day 3; meanwhile, the 40 m-CSBC was also higher than 20 m-CSBC. However, on day 7, the cell proliferation on m-CSBC became slower but the OD value was still higher than CSBC, with no obvious differences for m-CSBC between day 3 and day 7. The results suggested that the cements containing m-MCS could promote cell proliferation as compared with CSBC, showing good cytocompatibility.

(a) 70

weight loss (%)

50 40 30 20 10 0

7 14 21 28 35 42 49 56 63 70 77 84

(b) 7.4 7.3

pH

7.2 7.1 7.0 6.9 6.8 0

7 14 21 28 35 42 49 56 63 70 77 84 time (d)

Figure 4. (a) Weight loss and (b) pH change of the solution for CSBC, 20 mCSBC and 40 m-CSBC after the samples had been immersed in Tris – HCl solution for different times (7, 14, 21, 42, 56 and 84 days).

3.7. Cell differentiation The ALP activity of MC3T3-E1 cells increased with culture time for all the samples, as shown in figure 11. In general, the cells on m-CSBC presented a higher ALP levels than CSBC ( p , 0.05). No obvious differences were found for all the samples on day 7. However, a continual increase could be noted on subsequent days; the ALP of cells cultured on 40 m-CSBC was significantly higher than CSBC on day 10 and day 14; the 20 m-CSBC also presented a higher ALP activity than that of the CSBC on day 14.

4. Discussion The emergence of newly advanced bioceramics of ordered mesoporous materials displaying good bioactivity and biocompatibility has potential for bone tissue engineering [19]. Previous studies revealed that the too short setting time of CSBC (5 min) was not conducive for injecting cement into bone defects for clinical operations [16,20]. In this study, m-CSBC showed a longer setting time (11 min) compared with CSBC (5 min) so that the cements could be easily handled and injected into irregular bone defect cavities. Some studies have reported that the acidic degradation products, released by the CSBC, would decrease the pH value of the surrounding environment ( produced an acidic microenvironment), which might result in inflammatory responses in vivo [1,2,18,20]. In this study, it was found that

J. R. Soc. Interface 12: 20150779

CSBC 20 m-CSBC 40 m-CSBC

6

rsif.royalsocietypublishing.org

60

there was a gradual decrease of pH in the solution (from 7.40 to 6.89), indicating that acidic products were produced during CSBC degradation. However, the pH of the solution for 20 m-CSBC (from 7.40 to 7.28) and 40 m-CSBC (from 7.40 to 7.34) showed little decrease. The results suggested that the m-CSBC containing m-MCS could prevent the pH from further decreasing in the process of materials degradation. It can be explained that the degradation of the m-MCS in m-CSBC might neutralize the acidic degradation products of CSBC and thus compensated for the decrease in pH. The degradable material would be gradually replaced by newly formed bone tissue in vivo, so the degradability of biomaterials should be in accord with the new bone growth rate [21,22]. In this study, the degradation ratio of CSBC was 48.6 w% after immersing in the Tris –HCl solution for 84 days, while for 20 m-CSBC and 40 m-CSBC it was 59.3 w% and 63.5 w%, respectively. The fast degradation rate of m-CSBC might be attributed to the large surface area and high pore volume of m-MCS. In addition, the degradation of m-CSBC leads to macroporous structures in the cements, which was favourable for new bone tissue growth in materials and promoted bone regeneration [23]. The apatite layer formation on biomaterials in the biological environment plays an important role in the formation of chemical bonds between material surfaces and bone tissues, as well as bone-relating progress [24,25]. In our research, the m-CSBC could induce apatite deposition within one week, while little apatite was observed on the CSBC surface. The Ca ions from CSBC were not sufficient to promote the formation of new apatite on the material surface, and so it is improbable to generate bone connections with the surrounding tissues because the CSBC lacks bioactivity [18,26–28]. However, the excellent bioactivity of m-CSBC was attributed to the synergistic effects of Ca with Si ion release from m-CSBC during apatite formation (silicon ions are favourable for the mineralization), which could be explained as follows: as m-MCS gradually degrades, the Ca-O bond was broken, more Ca ions released and exchanged with H3Oþ in SBF, which facilitates the formation of HSiO34 – ; the concentration of silanol groups provided more nucleation sites for apatite. Then, the Ca2þ, HPO24 – and OH2 in SBF interacted and finally deposited on the nuclei through a dissolution-deposition process, forming the apatite layer on m-CSBC surfaces [29]. A study revealed that the ordered mesoporous bioglass had a superior ability to induce apatite formation compared with a non-porous structure, resulting from the special mesoporous structure of larger surface area and high pore volume [4]. Therefore, with the significantly improved bioactivity, the m-CSBC was expected to form stronger bone-bonding with the surrounding bone tissue by the apatite layer compared with the traditional calcium sulfate cement. Vitamin D3 can turn into 1,25-(OH)2-VitD3 (active form of vitamin D3) after transformation in kidney and liver in vivo; vitamin D3 is capable of conditioning the metabolism of Ca and P ions; inducing osteoblast differentiation; and regulating mRNA and protein expression [30]. Animal experiments have proved that 1,25-(OH)2-VitD3 had the ability to treat osteoporosis and repair articular cartilage defects by stimulating osteoblast growth in implants, which is a promising drug to help bone regeneration and accelerate the progress of bone construction [31].

(a)

7

(b)

rsif.royalsocietypublishing.org

5m

2m

J. R. Soc. Interface 12: 20150779

(c)

(d)

2m

5m

(f)

(e)

5m

2m

Figure 5. SEM images of apatite formation on CSBC (a, b), 20 m-CSBC (c,d ) and 40 m-CSBC (e,f ) after immersing in SBF for 7 days. Scale bars represent 5 mm (a,c,e) and 2 mm (b,d,f ). (a)

intensity (CPS)

(b) *

*

*

*

*

*

*

*

*

P

(i) Ca (ii) (iii)

20

*

30

40 2q (°)

*

*

50

1

2

3 energy (KeV)

4

5

Figure 6. XRD (a) of the samples immersed in SBF for 7 days: (i) 40 m-CSBC, (ii) 20 m-CSBC and (iii) CSBC. Asterisks represent the peaks of apatite. Circles represent the peaks of the cements; EDS (b) of 40 m-CSBC after immersion in SBF for 7 days. (Online version in colour.) In this study, we loaded vitamin D3 into the cements and investigated the sustained release behaviour of vitamin D3 from the cements in 25 days. On the first day, vitamin D3 exhibited a burst release from CSBC because of a large

amount of vitamin D3 exposed on the CSBC surface (easy release). In contrast, m-CSBC presented sustained released of vitamin D3 due to the mesoporous structure. After 5 days, the release of vitamin D3 from CSBC became slow,

ion concentration (mg l–1)

1000 800 600

0

40

20

60

80 100 120 140 160 180 time (h)

Figure 7. Change of ion concentrations of Ca, Mg, P and Si in SBF after soaking the 40 m-CSBC for different times (12, 24, 72, 120 and 168 h).

8 CSBC 20 m-CSBC 40 m-CSBC

40 30 20 10

5

0

10 m

25

10 m

(e)

(f)

10 m

(g)

20

(c)

10 m

(d )

15 time (d)

Figure 8. Cumulative release ratio of vitamin D3 from the cements after immersion in PBS for 25 days.

(b)

(a)

10

10 m

10 m

(i)

(h)

10 m

10 m

10 m

Figure 9. SEM images of cell morphology on the samples surfaces for 6 h (a,b,c), 1 day (d,e,f ) and 3 days (g,h,i). Cell morphology on CSBC (a,d,g), 20 m-CSBC (b,e,h) and 40 m-CSBC (c,f,i). CSBC 20 m-CSBC 40 m-CSBC

*

1.5 * 1.0

0.5

0.6

*

*

* *

ALP activity (OD)/total protein (ng)

optical density (OD)

2.0

0.5

CSBC 20 m-CSBC 40 m-CSBC

*

*

*

10 time (d)

14

0.4 *

0.3 0.2 0.1

0

1

3 time (d)

7

Figure 10. Cell proliferation of MC3T3-E1 cells cultivated on samples for 1, 3 and 7 days, *p , 0.05.

0

7

Figure 11. Alkaline phosphatase activity of MC3T3-E1 cells cultivated on samples for 7, 10 and 14 days, *p , 0.05.

J. R. Soc. Interface 12: 20150779

400 120 100 80 60 40 20 0

50

rsif.royalsocietypublishing.org

Ca Mg Si Pi

cumulative release ratio of vitamin D3 (%)

1200

which might be responsible for stimulating cell proliferation and differentiation, showing good cytocompatibility.

5. Conclusion

Ethics. All participants gave informed consent prior to data collection. Cell experiments were supported by the Ninth People’s Hospital affiliated to Shanghai Jiao Tong University School of Medicine.

Authors’ contribution. Y.D. carried out the experiments and performed data statistical analysis. J.W. conceived and coordinated the study. S.T., B.Y, Y.Y. and H.L. participated in drafting the manuscript. Y.D and J.W. wrote and revised the paper. J.S. performed the cell experiments and collected related data. All authors gave final approval for the publication.

Competing interests. We declare we have no competing interests. Funding. This study was supported by grants from the National Natural Science Foundation of China (51173041, 31271031), the National Science and Technology Support Plan Projects (no. 2012BAD32B01), the International Cooperation Project of the Ministry of Science and Technology of China (2013DFB50280), the Major International Joint Research Project between China and Korea (81461148033, 31311140253) and the Key Medical Program of Science and Technology Development of Shanghai (no. 12441902802).

Acknowledgements. The authors thank Zhaoying Wu, Jie Chen and Liang Cai for comments on earlier versions of the manuscript.

References 1.

2.

3.

Yang GY, Liu JL, Li F, Pan ZY, Ni X, Shen Y, Xu HZ, Huang Q. 2014 Bioactive calcium sulfate/ magnesium phosphate cement for bone substitute applications. Mater. Sci. Eng. C 35, 70 –76. (doi:10. 1016/j.msec.2013.10.016) Chen ZG, Liu HY, Liu X, Lian XJ, Guo ZW, Jiang HJ, Cui FZ. 2013 Improved workability of injectable calcium sulfate bone cement by regulation of selfsetting properties. Mater. Sci. Eng. C 33, 1048– 1053. (doi:10.1016/j.msec.2012.11.019) Chen WL, Chen CK, Lee JW, Lee YL, Ju CP, Chern LJH. 2014 Structure, properties and animal study of a calcium phosphate/calcium sulfate composite cement. Mater. Sci. Eng. C 37, 60 –67. (doi:10. 1016/j.msec.2013.12.034)

4.

5.

6.

Lu JX et al. 2012 Preparation, bioactivity, degradability and primary cell responses to an ordered mesoporous magnesium–calcium silicate. Microporous Mesoporous Mater. 163, 221–228. (doi:10.1016/j.micromeso.2012.06.037) Yi DL, Wu CT, Ma B, Ji H, Zheng XB, Chang J. 2014 Bioactive bredigite coating with improved bonding strength, rapid apatite mineralization and excellent cytocompatibility. J. Biomater. Appl. 28, 1343– 1353. (doi:10.1177/0885328213508165) Wu CT, Ramaswamy Y, Zreiqat H. 2010 Porous diopside (CaMgSi2O6) scaffold: a promising bioactive material for bone tissue engineering. Acta Biomater. 6, 2237–2245. (doi:10.1016/j.actbio. 2009.12.022)

7.

8.

9.

Tamimi F et al. 2011 Biocompatibility of magnesium phosphate minerals and their stability under physiological conditions. Acta Biomater. 7, 2678– 2685. (doi:10.1016/j.actbio. 2011.02.007) Yoshizawa S, Brown A, Barchowsky A, Sfeir C. 2014 Magnesium ion stimulation of bone marrow stromal cells enhances osteogenic activity, simulating the effect of magnesium alloy degradation. Acta Biomater. 10, 2834–2842. (doi:10.1016/j.actbio. 2014.02.002) Diba M, Goudouri OM, Tapia F, Boccaccini AR. 2014 Magnesium-containing bioactive polycrystalline silicate-based ceramics and glass-ceramics for biomedical applications. Curr. Opin. Solid State

J. R. Soc. Interface 12: 20150779

In this study, m-CSBC was prepared by adding m-MCS to CSH. Our findings revealed that the incorporation of m-MCS into cements prolonged the setting time of the m-CSBC, which was beneficial to improving the injectability of the cements. Meanwhile, the compressive strength slightly decreased. In addition, the addition of m-MCS into the cements improved the degradability and apatite-formation ability of the m-CSBC, which was m-MCS content dependent. Moreover, the m-CSBC containing m-MCS showed the ability to neutralize the acidic degradation by-products released from the cements and prevented the pH from deceasing during degradation. It was found that the cumulative release ratio of vitamin D3 from m-CSBC was significantly higher with the increase of m-MCS content. In the cell culture experiments, our findings showed that the m-CSBC containing m-MCS enhanced the attachment, proliferation and ALP activity of MC3T3-E1 cells. In summary, m-CSBC possessed excellent bioactivity, degradability and cytocompatibility, and might be a new biomaterial for bone tissue regeneration.

9

rsif.royalsocietypublishing.org

as most vitamin D3 was wrapped in it, which blocked continued release of vitamin D3 unless from CSBC degradation. In comparison to the increase of m-MCS content, more m-MCS was exposed on the m-CSBC surfaces, so that the cements displayed a fast cumulative release rate of vitamin D3. Furthermore, the degradation rate of m-CSBC was faster than CSBC, thus the accumulated amount of vitamin D3 from m-CSBC was more than CSBC. After 25 days, the cumulative release ratio of 40 m-CSBC, 20 m-CSBC and CSBC was 41.3%, 30.6% and 17.5%, respectively. The results suggested that the mesoporous structure provided a large area for drugs to be loaded, which might be favourable for drug sustained release compared with the CSBC. The cell responses on the interface of biomaterials surfaces and cells including the morphology, proliferation and differentiation were closely associated with cytocompatibility [32,33]. The results revealed that the surfaces of m-CSBC were beneficial to cell adhesion and growth: at 6 h, the morphology of MC3T3-E1 presented normal phenotype with the synapse extend and spread, indicating that the samples have no negative effect on cell attachment. The cell number increased and closely attached to the substrate at 24 h of culture. On the third day, the cells formed a confluent layer on the m-CSBC surface, showing good cytocompatibility. Moreover, through the MTT assay, it was found that the m-CSBC obviously stimulated the proliferation of MC3T3-E1 cells on days 3 and 7. ALP is a specific indicator to evaluate osteogenic activity and differentiation ability; its activity was positively correlated with osteoblast differentiation [34]. The result revealed that ALP activity of the cells cultured on the m-CSBC markedly increased compared with CSBC on days 10 and 14. Previous studies have confirmed that the ion dissolution products containing Ca, Mg and Si ions from bioactive glasses/ ceramics could stimulate osteoblast proliferation and differentiation [35,36]. Moreover, several studies have shown that the silicon ions could improve the ability of osteoblasts and fibroblasts to produce collagen type I, so that cell differentiation can be promoted [37,38]. In this study, the ions (including Ca, Mg and Si) released from m-CSBC might be crucial for cell responses. Therefore, it can be suggested that the m-CSBC provided a Ca-, Mg- and Si-rich environment,

11.

13.

14.

15.

16.

17.

18.

30.

31.

32.

33.

34.

35.

36.

37.

38.

containing bioceramic coatings with high bonding strength on inflammation, osteoclastogenesis, and osteogenesis. ACS Appl. Mater. Inter. 6, 4264–4276. (doi:10.1021/am4060035) Mao L, Tamura Y, Kawao N, Okada K, Yano M, Okumoto K, Kaji H. 2014 Influence of diabetic state and vitamin D deficiency on bone repair in female mice. Bone 61, 102–108. (doi:10.1016/j. bone.2013.12.024) Laine OL, Zheng Y, Zhou H, Trivedi T, Conigrave AD, Seibel MJ, Dunstan CR. 2010 Vitamin D deficiency promotes growth of MCF-7 human breast cancer in a rodent model of osteosclerotic bone metastasis. Bone 47, 795 –803. (doi:10.1016/ j.bone.2010.07.012) Duan B, Wang M. 2010 Customized Ca-P/PHBV nanocomposite scaffolds for bone tissue engineering: design, fabrication, surface modification and sustained release of growth factor. J. R. Soc. Interface 7, S615 –S629. (doi:10.1098/rsif. 2010.0127.focus) Wang C, Lin K, Chang J, Sun J. 2014 The stimulation of osteogenic differentiation of mesenchymal stem cells and vascular endothelial growth factor secretion of endothelial cells by b-CaSiO3/ b-Ca3(PO4)2 scaffolds. J. Biomed. Mater. Res. A 102A, 2096– 2104. (doi:10.1002/jbm.a.34880) Zhou YH, Wu CT, Xiao Y. 2014 Silicate-based bioceramics for periodontal regeneration. J. Mater. Chem. B 2, 3907–3910. (doi:10.1039/c4tb00377b) Ohtsuki C, Kamitakahara M, Miyazaki T. 2009 Bioactive ceramic-based materials with designed reactivity for bone tissue regeneration. J. R. Soc. Interface 6, S349–S360. (doi:10.1098/rsif.2008.0419.focus) Li HY, Xue K, Kong N, Liu K, Chang J. 2014 Silicate bioceramics enhanced vascularization and osteogenesis through stimulating interactions between endothelia cells and bone marrow stromal cells. Biomaterials 35, 3803–3818. (doi:10.1016/j. biomaterials.2014.01.039) Mourino V, Boccaccin AR. 2010 Bone tissue engineering therapeutics: controlled drug delivery in threedimensional scaffolds. J. R. Soc. Interface 7, 209–227. (doi:10.1098/rsif.2009.0379) Chen Z, Mao XL, Tan LL, Friis T, Wu CT, Crawford R, Xiao Y. 2014 Osteoimmunomodulatory properties of magnesium scaffolds coated with b-tricalcium phosphate. Biomaterials 35, 8553–8565. (doi:10. 1016/j.biomaterials.2014.06.038)

10

J. R. Soc. Interface 12: 20150779

12.

19. He QJ, Shi JL, Zhu M, Chen Y, Chen F. 2010 The three-stage in vitro degradation behavior of mesoporous silica in simulated body fluid. Microporous Mesoporous Mater. 131, 314–320. (doi:10.1016/j.micromeso.2010.01.009) 20. Huan ZG, Chang J. 2007 Self-setting properties and in vitro bioactivity of calcium sulfate hemihydratetricalcium silicate composite bone cements. Acta Biomater. 3, 952 –960. (doi:10.1016/j.actbio.2007. 05.003) 21. Fredholm YC, Karpukhina N, Brauer DS, Jones JR, Law RV, Hill RG. 2012 Influence of strontium for calcium substitution in bioactive glasses on degradation, ion release and apatite formation. J. R. Soc. Interface 9, 880–889. (doi:10.1098/rsif.2011.0387) 22. Kanter B, Geffers M, Ignatius A, Gbureck U. 2014 Control of in vivo mineral bone cement degradation. Acta Biomater. 10, 3279–3287. (doi:10.1016/j. actbio.2014.04.020) 23. Hudson SP, Padera RF, Langer R, Kohane DS. 2008 The biocompatibility of mesoporous silicates. Biomaterials 29, 4045–4055. (doi:10.1016/j. biomaterials.2008.07.007) 24. Kalia P, Vizcay-Barrena G, Fan JP, Warley A, Silviol LD, Huang J. 2014 Nanohydroxyapatite shape and its potential role in bone formation: an analytical study. J. R. Soc. Interface 4, 1–11. (doi:10.1098/rsif. 2014.0004) 25. Azama JC, Alkhraisat MH, Rueda C, Torres J, Blanco L, Lo´pez-Cabarcos E. 2014 Magnesium substitution in brushite cements for enhanced bone tissue regeneration. Mater. Sci. Eng. 43, 403– 410. (doi:10. 1016/j.msec.2014.06.036) 26. Vorndran E, Geffers M, Ewald A, Lemmb M, Nies B, Bureck U. 2013 Ready-to-use injectable calcium phosphate bone cement paste as drug carrier. Acta Biomater. 9, 9558–9567. (doi:10.1016/j.actbio. 2013.08.009) 27. Jia JF, Zhou HJ, Wei J, Jiang X, Hong H, Chen FP, Wei SC, Shin JW, Liu CS. 2010 Development of magnesium calcium phosphate biocement for bone regeneration. J. R. Soc. Interface 7, 1171–1180. (doi:10.1098/rsif.2009.0559) 28. Zhang YF, Li S, Wu CT. 2014 The in vitro and in vivo cementogenesis of CaMgSi2O6 bioceramic scaffolds. J. Biomed. Mater. Res. 102A, 105–116. (doi:10. 1002/jbm.a.34679) 29. Wu CT, Chen Z, Yi DL, Chang J, Xiao Y. 2014 Multidirectional effects of Sr-, Mg-, and Si-

rsif.royalsocietypublishing.org

10.

Mater. Sci. 18, 147–167. (doi:10.1016/j.cossms. 2014.02.004) Schumacher S, Roth I, Stahl J, Ba¨umer W, Kietzmann M. 2014 Biodegradation of metallic magnesium elicits an inflammatory response in primary nasal epithelial cells. Acta Biomater. 10, 996–1004. (doi:10.1016/j.actbio.2013.10.030) Feng P, Wei PP, Li PJ, Gao CD, Shuai CJ, Peng SP. 2014 Calcium silicate ceramic scaffolds toughened with hydroxyapatite whiskers for bone tissue engineering. Mater. Charact. 97, 47– 56. (doi:10. 1016/j.matchar.2014.08.017) Zhu YF, Zhu M, He X, Zhang JH, Tao CL. 2013 Substitutions of strontium in mesoporous calcium silicate and their physicochemical and biological properties. Acta Biomater. 9, 6723 –6731. (doi:10. 1016/j.actbio.2013.01.021) Wu ZY et al. 2014 In vitro degradability, bioactivity and cell responses to mesoporous magnesium silicate for the induction of bone regeneration. Colloids Surf. B 120, 38– 46. (doi:10.1016/j.colsurfb. 2014.04.010) Zhang JH, Zhy YF. 2014 Synthesis and characterization of CeO2-incorporated mesoporous calcium-silicate materials. Microporous Mesoporous Mater. 197, 244–251. (doi:10.1016/j.micromeso. 2014.06.018) Gu JL, Huang K, Zhu XY, Li YS, Wei J, Zhao WR, Liu CS, Shi JL. 2013 Sub-150 nm mesoporous silica nanoparticles with tunable pore sizes and wellordered mesostructure for protein encapsulation. J. Colloid Interface Sci. 407, 236–242. (doi:10.1016/ j.jcis.2013.06.028) Liu WJ, Wu CT, Liu WN, Zhai WY, Chang J. 2013 The effect of plaster (CaSO†4 1/2H2O) on the compressive strength, self-setting property, and in vitro bioactivity of silicate-based bone cement. J. Biomed. Mater. Res. B 101B, 279–286. (doi:10.1002/jbm.b.32837) Lu JX, Wei J, Yan YG, Li H, Jia JF, Wei SC, Guo H, Xiao TQ, Liu CS. 2011 Preparation and preliminary cytocompatibility of magnesium doped apatite cement with degradability for bone regeneration. J. Mater. Sci. Mater. Med. 22, 607–615. (doi:10. 1007/s10856-011-4228-4) Chen CC, Wang CW, Hsueh NS, Ding SJ. 2014 Improvement of in vitro physicochemical properties and osteogenic activity of calcium sulfate cement for bone repair by dicalcium silicate. J. Alloys Compound 585, 25–31. (doi:10.1016/j.jallcom.2013.09.138)