lable at ScienceDirect

Biomaterials 35 (2014) 1507e1518

Contents lists avai

Biomaterials

journal homepage: www.elsevier .com/locate/biomater ia ls

Osteogenic differentiation of bone marrow MSCs by b-tricalciumphosphate stimulating macrophages via BMP2 signalling pathway

Zetao Chen a, Chengtie Wu b,*, Wenyi Gu a,c, Travis Klein a, Ross Crawford a, Yin Xiao a,**

a Institute of Health and Biomedical Innovation, Queensland University of Technology, Brisbane, 60 Musk Ave, Kelvin Grove, Brisbane,Queensland 4059, Australiab State Key Laboratory of High Performance Ceramics and Superfine Microstructure, Shanghai Institute of Ceramics,Chinese Academy of Sciences, 1295 Dingxi Road, Shanghai 200050, People’s Republic of ChinacAustralian Institute for Bioengineering and Nanotechnology, The University of Queensland, Corner College and Cooper Rds, Brisbane,Queensland 4072, Australia

a r t i c l e i n f o

Article history:Received 23 September 2013Accepted 3 November 2013Available online 20 November 2013

Keywords:MacrophageImmunoreactionBone substituteb-TCPOsteogenesis

* Corresponding author. Tel.: þ86 21 52412249; fax** Corresponding author. Tel.: þ61 7 3138 6240; fax

E-mail addresses: chengtiewu@mail.sic.ac.cn (C(Y. Xiao).

0142-9612/$ e see front matter � 2013 Elsevier Ltd.http://dx.doi.org/10.1016/j.biomaterials.2013.11.014

a b s t r a c t

Immune reactions play important roles in determining the in vivo fate of bone substitute materials, eitherin new bone formation or inflammatory fibrous tissue encapsulation. The paradigm for the developmentof bone substitute materials has been shifted from inert to immunomodulatory materials, emphasizingthe importance of immune cells in the material evaluation. Macrophages, the major effector cells in theimmune reaction to implants, are indispensable for osteogenesis and their heterogeneity and plasticityrender macrophages a primer target for immune system modulation. However, there are very few re-ports about the effects of macrophages on biomaterial-regulated osteogenesis. In this study, we usedb-tricalcium phosphate (b-TCP) as a model biomaterial to investigate the role of macrophages on thematerial stimulated osteogenesis. The macrophage phenotype switched to M2 extreme in response tob-TCP extracts, which was related to the activation of calcium-sensing receptor (CaSR) pathway. Bonemorphogenetic protein 2 (BMP2) was also significantly upregulated by the b-TCP stimulation, indicatingthat macrophage may participate in the b-TCP stimulated osteogenesis. Interestingly, when macrophage-conditioned b-TCP extracts were applied to bone marrow mesenchymal stem cells (BMSCs), the osteo-genic differentiation of BMSCs was significantly enhanced, indicating the important role of macrophagesin biomaterial-induced osteogenesis. These findings provided valuable insights into the mechanism ofmaterial-stimulated osteogenesis, and a strategy to optimize the evaluation system for the in vitroosteogenesis capacity of bone substitute materials.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

As exogenous bodies, implants tend to be rejected or isolated bythe human body, mainly by the foreign body reaction. Aiming toavoid such a reaction, inert materials have been developed forin vivo implantation. However, this approach regrettably minimizesreconstructive healing and even worse sometimes enhances theforeign body reaction, resulting in the encapsulation of thematerial[1]. With the emergence of osteoimmunology, the interaction be-tween bone cells and immune cells has been studied more thor-oughly [2]. The immune and skeletal systems are closely related,

: þ86 21 52413903.: þ61 7 3138 6030.. Wu), yin.xiao@qut.edu.au

All rights reserved.

sharing a number of cytokines, receptors, signalling molecules andtranscription factors [3]. Besides its perpetuation leading to chronicinflammation, the immune response plays a vital role in implantintegration and osteogenesis [4,5]. Inflammatory cytokinesinterleukin-1 (IL-1), tumour necrosis factor a (TNF-a), and IL-6 canenhance osteoclast differentiation and resorbing activity, andinhibit osteoblast activity and bone formation, while anti-inflammatory cytokines IL-4, IL-10, and IL-13 have the oppositeeffect [6]. Instead of osteoblastic cells, B cells have been shown to bethe main source of bone marrow-derived osteoprotegerin, with apercentage of 64% [7,8], which suggests that B cells are the maininhibitors of osteoclastogenesis in normal physiology. This under-standing has led to a shift towards the development of “smart”bone substitute biomaterials capable of immunomodulation [4].These “smart” biomaterials should be able to arouse an effectiveimmune response, building an osteogenesis-enhancing environ-ment for bone cells. In vitro methods previously used to evaluate

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181508

bone substitute materials based on the interactions between bonecells and materials are insufficient for evaluating these types ofbiomaterials, as they do not reflect the in vivo condition, whichinvolves early immunoreaction upon transplantation. It is thereforeof great importance to investigate the effects of immune cells onthe biomaterial-regulated osteogenesis.

Among all immune cells, the macrophage is one of the mostimportant effector cells in the material-induced immune response.It has been reported that the long-term immune response to ma-terials and the outcome of the inflammatory reaction are mainlydetermined by macrophages [9]. In addition, the heterogeneity andplasticity of macrophage make it a prime target for immunomo-dulation [10]. Macrophages have several phenotypes, which aredynamics and plastic [11]. In response to the environmental signalsalteration, macrophages can change their phenotype and physi-ology [12]. They are broadly referred to as either M1 or M2 phe-notypes [13]. The classically activated M1 phenotype with typicalsurface markers CD11c and CCR7, help to enhance Th1 (T helper cell1) type inflammation [14]. The alternatively activated M2 pheno-type with the typical surface markers CD163 and CD206, helps toenhance Th2 (T helper cell 2) type inflammation, downregulateinflammation and improve tissue healing [15,16]. In addition totheir effects on inflammation, macrophages are also known to in-fluence bone physiology and pathology [17,18]. It is well known thatmacrophages are the precursors of osteoclasts, which participate inthe bone remodelling and material degradation. Macrophages alsocontribute to osteogenesis through the expression and secretion ofa wide range of regulatory molecules [19], such as BMP2, trans-forming growth factor b (TGF-b), etc [20,21]. Macrophages arerequired for efficient osteoblast mineralization and the depletion ofmacrophages can decrease osteoblast bone forming capacity [18].

Given the important roles of macrophages in the bone dy-namics, some studies have investigated the interactions betweenbone substitute biomaterials (such as bioceramics, polymers, tita-nium, etc) and macrophages. However, most of these studies arelimited to either the classic inflammatory contributions or thedifferentiation into osteoclasts on various bone substitute bio-materials [22e24]. Very few reports have been made on the effectsof macrophages in regulating materials stimulated osteogenesis.

b-TCP is well recognized as an osteoconductive biomaterial andhas been widely used for clinical bone regeneration application[25e27]. Although there are many studies about the in vitro andin vivo osteogenesis of b-TCP bioceramics, the effect of b-TCP on thebiological response of macrophages is unclear, and the effect ofmacrophages on b-TCP-coordinated osteogenesis of BMSCs has notbeen investigated. Therefore, in the current study, b-TCP is used as amodel biomaterial to investigate the interactions between bonesubstitute materials and macrophages, and to further revealwhether macrophages participate in the material-stimulatedosteogenesis.

2. Materials and methods

2.1. Cell culture

The murine-derived macrophage cell line RAW 264.7 cells and human bonemarrow mesenchymal stem cells (BMSCs) were used in this study. RAW 264.7 cellcultures were maintained in Dulbecco’s Modified Eagle Medium (DMEM, LifeTechnologies, Carlsbad, California, USA) supplementedwith 10% foetal bovine serum(FBS, Thermo Scientific, Waltham, Massachusetts, USA), and 1% (v/v) penicillin/streptomycin (Life Technologies, Carlsbad, California, USA) at 37 �C in a humidifiedCO2 incubator. The cells were passaged at approximately 80% confluence by scrapingand expanded through two passages before being used for the study.

BMSCs were isolated and cultured based on protocols from previous studies[28,29]. Briefly, bone marrow was obtained from patients (50e60 years old) un-dergoing hip or knee replacement surgery with informed consent given by all do-nors and the procedure was approved by the Ethics Committee of QueenslandUniversity of Technology. Lymphoprep was added to isolate the mononuclear cellsfrom the bone marrow by density gradient centrifugation (Axis-Shield PoC AS, Oslo,

Norway). The obtained cells were seeded into the tissue culture flasks containingDMEM supplementedwith 10% FBS and 1% penicillin/streptomycin and incubated at37 �C in a humidified CO2 incubator. The culture mediumwas changed every 3 daysuntil the primary mesenchymal cells reached 80% confluence. The unattached he-matopoietic cells were removed through medium change. The confluent cells wereroutinely subcultured by trypsinization. Only early passages (p3e5) of cells wereused in this study.

2.2. Metabolic activity of RAW 264.7 cells with b-TCP extracts

b-TCP was synthesized by a chemical precipitation method, as previousdescribed [30]. The prepared powders were sieved with 60e80 mesh to obtain theparticle size of 180e250 mm before further experiment. Based on the standard ISO/EN 10993-5, the cells were cultured in b-TCP powder extract medium [30,31]. Thepowder extracts were prepared by soaking powders in serum-free DMEM at a seriesof solid/liquid ratio (200, 100, 50, 25, 12.5, 6.25, 0 mg/mL). After incubation at 37 �Cfor 24 h, the mixture was centrifuged and the supernatant was collected and ster-ilized by filtration through 0.2 mm filter membranes (Pall Corporation, Port Wash-ington, New York, USA).

RAW 264.7 cells were seeded in a 96-well tissue culture plate at a density of2000 cells per well. After 24 h of incubation, the culture medium was removed andreplaced by 200 mL of material extract containing 10% FBS. DMEM without extractsupplemented with 10% FBS was used as a control. At days 1 and 3, a MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide] assay was applied to testthe overall metabolic activity. 20 mL of 5 mg/mL MTT solution was added into eachwell. After incubation for a further 4 h, the DMEM-MTT solution was carefullyremoved and replaced by 100 mL of DMSO (dimethyl sulfoxide) to dissolve the for-mazan crystals. The absorbance was read at the wavelength of 495 nm by amicroplate spectrophotometer (Benchmark Plus, Tacoma, Washington, USA). Eachsample was performed in triplicate.

2.3. Flow cytometry and fluorescence microscopy

Expression of M1 and M2 macrophage cell surface marker CD11c and CD163,respectively, were determined by flow cytometry. After 5 days of culture, the cellswere detached by scraping. Nonspecific protein binding was blocked by 1% BSA/PBS.Samples were incubated with CD11c and CD163 antibody (1:100) (AbD Serotec,Raleigh, North Carolina, USA) for 30 min at 4 �C, followed by incubation with aDylight 488-anti-mouse secondary antibody (DAKO, Multilink, California, USA) for30 min at 4 �C. After washing with 1% BSA/PBS, cells were analyzed on a FC500 flowcytometer (Beckman Coulter, Brea, California, USA). The data were analyzed usingFlowing Software (www.flowingsoftware.com). For fluorescence microscopy, thetreated cells were mounted by mounting medium containing 40 ,6-diamidino-2-phenylindole (DAPI) (SigmaeAldrich, St. Louis, Missouri United States), to stainthe cell nuclei. Figures of the stained slides were then captured using Axion softwareunder a fluorescence microscope (Carl Zeiss Microimaging GmbH, Gottingen,Germany).

2.4. Effects of b-TCP extracts on the inflammatory and osteogenic gene expression ofRAW 264.7 cells

2.4.1. Inflammatory gene expression of RAW 264.7 cells with b-TCP extractsRAW 264.7 cells were seeded in a 6-well plate at a density of 50,000 cells per

well. After 24 h of incubation, the culture medium was removed and replaced bymaterial extract containing 10% FBS. Culture medium with 10% FBS was used as anegative control. After 3 days, media from cells cultured in material extract or cul-ture medium were collected, and centrifuged at 1500 rpm to gain the supernatantsfor further conditioned media experiments and ion concentration detection. TotalRNA of both groups was extracted using TRIzol reagent (Life Technologies, Carlsbad,California, USA) for RT-qPCR detection.

500 ng total RNA was used for the synthesis of complementary DNA usingDyNAmo� cDNA Synthesis Kit (Finnzymes, Thermo Scientific, Waltham, Massa-chusetts, USA) following the manufacturer’s instructions. RT-qPCR primers (Table 1)were designed based on cDNA sequences from the NCBI Sequence database. SYBRGreen qPCR Master Mix (Life Technologies, Carlsbad, California, USA) was used fordetection and the target mRNA expressions were assayed on the ABI Prism 7300Thermal Cycler (Applied Biosystems, Foster City, California, USA). Each sample wasperformed in triplicate. The mean cycle threshold (Ct) value of each target gene wasnormalized against Ct value of a housekeeping gene to gain the relative expression.For the calculation of fold change, DDCt method was applied, comparing mRNAexpressions between treated groups and negative group (cells cultured in thecomplete culture medium only).

2.4.2. Mechanism of the inflammatory gene expression changeTo understand the mechanism of the gene expression changes, the ionic con-

centrations of Ca2þ, and PO3�4 ions in complete culture medium, b-TCP extracts and

RAW 264.7 cell conditioned media were quantified by inductive coupled plasmaatomic emission spectrometry (ICP-AES, PerkinElmer,Waltham,Massachusetts, USA).

Two Ca2þ involved inflammation signalling pathways, Wnt5A/Ca2þ, and CaSRwere evaluated by RT-qPCR and Western blot. RAW 264.7 cells were seeded in a

Table 1Primer pairs used in the qRT-PCR.

Genes Primer sequences

IL10 Forward: 50-GAGAAGCATGGCCCAGAAATC-30

Reverse: 50-GAGAAATCGATGACAGCGCC-30

IL1ra Forward: 50-CTCCAGCTGGAGGAAGTTAAC-30

Reverse: 50-CTGACTCAAAGCTGGTGGTG-30

TNFa Forward: 50-CTGAACTTCGGGGTGATCGG-30

Reverse: 50-GGCTTGTCACTCGAATTTTGAGA-30

IL1b Forward: 50-TGGAGAGTGTGGATCCCAAG-30

Reverse: 50-GGTGCTGATGTACCAGTTGG-30

IL6 Forward: 50-ATAGTCCTTCCTACCCCAATTTCC-30

Reverse: 50-GATGAATTGGATGGTCTTGGTCC-30

BMP2 Forward: 50-GCTCCACAAACGAGAAAAGC-30

Reverse: 50-AGCAAGGGGAAAAGGACACT-30

VEGFa Forward: 50-GTCCCATGAAGTGATCAAGTTC-30

Reverse: 50-TCTGCATGGTGATGTTGCTCTCTG-30

TGFb1 Forward: 50-CAGTACAGCAAGGTCCTTGC-30

Reverse: 50-ACGTAGTAGACGATGGGCAG-30

WNT5A Forward: 50-CAACTGGCAGGACTTTCTCAA-30

Reverse: 50-CCTGATACAAGTGGCAGAGTTTC-30

Fz5 Forward: 50-AAGTCCATTACGGCGCTG-30

Reverse: 50-AGCCTCGTAGTGAGTTCAGG-30

CaSR Forward: 50-GGCGGACTCAGAAATGACC-30

Reverse: 50-TGATGACGAAGCTCCGAGAG-30

Actinb Forward: 50-CATACCCAAGAAGGAAGGCTGG-30

Reverse: 50-GCTATGTTGCTCTAGACTTCGAGC-30

ALP Forward: 50-TCAGAAGCTAACACCAACG-30

Reverse: 50-TTGTACGTCTTGGAGAGGGC-30

OPN Forward: 50-TCACCTGTGCCATACCAGTTAA-30

Reverse: 50-TGAGATGGGTCAGGGTTTAGC-30

OCN Forward: 50-GCAAAGGTGCAGCCTTTGTG-30

Reverse: 50-GGCTCCCAGCCATTGATACAG-30

COL1A1 Forward: 50-AGAACAGCGTGGCCT-30

Reverse: 50-TCCGGTGTGACTCGT-30

SMAD1 Forward: 50-GTATGAGCTTTGTGAAGGGC-30

Reverse: 50-TAAGAACTTTATCCAGCCACTGG-30

SMAD4 Forward: 50-CTCCAGCTATCAGTCTGTCAG-30

Reverse: 50-CCCGGTGTAAGTGAATTTCAAT-30

SMAD5 Forward: 50-TCATCATGGCTTTCATCCCACC-30

Reverse: 50-GCTCCCCAACCCTTGACAAA-30

BMPR2 Forward: 50-GGCAGCAGTATACAGATAGGTG-30

Reverse: 50-CTGCCCTGTTACTGCCATTATT-30

BMP1A Forward: 50-CCTGGGCCTTGCTGTTAAATTCA-30

Reverse: 50-TCCACGATCCCTCCTGTGAT-30

18S Forward: 50-CGGAACTGAGGCCATGATTAAG-30

Reverse: 50-GTATCTGATCGTCTTCGAACCTCC-30

Z. Chen et al. / Biomaterials 35 (2014) 1507e1518 1509

6-well plate at a density of 50,000 cells per well. After 24 h of incubation, the culturemediumwas removed and replaced by material extract containing 10% FBS. Culturemediumwith 10% FBS was used as a negative control. After 3 days, media from cellscultured in material extract or culture medium were collected. Gene expression ofWnt5A, Frizzled family receptor 5 (Fz5), CaSR were detected following the samemethod in Section 2.4.1. Thewhole cell lysateswere collected after 24 h of culture forthe Western Blot detection (calmodulin-dependent protein kinase II (CaMKII), nu-clear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha (IkB-a)). 15 mg proteins from each sample were separated on SDS-PAGE gels and thentransferred onto a nitrocellulose membrane (Pall Corporation, East Hills, New York,USA). After being blocked in Odyssey blocking buffer for 1 h (LI-COR Biosciences,Lincoln, Nebraska, USA), the membranes were incubated with primary antibodiesagainst IkB-a (1:1000, rabbit anti-human/mouse; Cell Signaling Technology, Dan-vers, Massachusetts, USA), CamKII (pan, 1:1000, rabbit anti-human/mouse; CellSignaling Technology, Danvers, Massachusetts, USA), and a-tubulin (1:5000, rabbitanti-human; Abcam, Cambridge, United Kingdom) overnight at 4 �C. The mem-branes were washed three times in TBS-Tween buffer, and then incubated with anti-mouse/rabbit HRP conjugated secondary antibodies at 1: 4000 dilutions for 1 h atroom temperature. The protein bands were visualized using the Odyssey infraredimaging system (LI-COR Biosciences, Lincoln, Nebraska, USA). The relative intensityof protein bands was quantified using Image J software (National Institutes ofHealth, Bethesda, Maryland, USA).

2.4.3. Osteogenic gene expression of RAW 264.7 cells stimulated with b-TCP extractsTo investigate the effect of b-TCP extracts on osteogenic gene expression by

RAW 264.7 cells, BMP2, Transforming growth factor beta 1 (TGFb1) and Vascularendothelial growth factor (VEGF) were analyzed by RT-qPCR as described in Section2.4.1.

2.5. Effects of RAW 264.7 cells-conditioned b-TCP extracts on the osteogenicdifferentiation of BMSCs

2.5.1. Bone-related gene and protein expression of BMSCsTo investigate whether RAW 264.7 cells could regulate osteogenesis of BMSCs in

response to material extracts, the BMSC culture medium was supplemented withthe collected supernatants detailed in Section 2.4.1 at a ratio of 1:2 to obtain theconditioned media [32]. The pure material extract and culture medium withoutbeing conditioned by RAW 264.7 cells were used as controls.

For the detection of bone-related gene [alkaline phosphatase (ALP), osteopontin(OPN), osteocalcin (OCN), and Collagen, type I (COL1)] and protein expression (ALP,OPN) of BMSCs, BMSCs were plated at a density of 20,000 cells per well in separate6-well plates. After 24 h of incubation, the culture medium was removed andreplaced by conditioned medium or control medium. Total RNA of all four groupswas extracted using TRIzol reagent after 1 and 3 days of culture, and analyzed by RT-qPCR, as described in Section 2.4.1. The whole cell lysates were collected after 14days of culture, and analyzed by Western blot, as described in Section 2.4.2, withprimary antibodies against ALP (1:10,000, rabbit anti-human; Abcam, Cambridge,United Kingdom), OPN (1:1000, rabbit anti-human; Abcam, Cambridge, UnitedKingdom), and a-tubulin (1:5000, rabbit anti-human; Abcam, Cambridge, UnitedKingdom).

2.5.2. The mineralization of BMSCsIn order to identify mineralization nodules, Alizarin Red S staining was

measured on day 14 after BMSCs grown in conditioned or control media in a 96-wellplate with or without osteogenic supplements. The medium was removed and thecells werewashed with ddH2O and fixed in 4% paraformaldehyde for 10min at roomtemperature. After gently rinsing with ddH2O, the cells were stained in a solution of2% Alizarin Red S at pH 4.1 for 20 min and were then washed with ddH2O. Thesamples were air-dried and figures were taken under a light microscope.

2.5.3. BMP2-related signalling pathway of BMSCsTo investigate the activation of the BMP2 signalling pathway in BMSCs, BMP2

signalling pathway related genes [Mothers against decapentaplegic homologue 1/4/5(SMAD1, SMAD4, SMAD5), bone morphogenetic protein receptor, type IA (BMPR1A)and Bonemorphogenetic protein receptor type II (BMPR2)] were analyzed by RT-qPCRas described in Section 2.5.1.

2.6. Statistical analysis

All the analyses were performed using SPSS software (IBM SPSS, Armonk, NewYork, USA). All the data were shown as means � standard deviation (SD) andanalyzed using one-way ANOVA followed by least significant difference post hoc test.A level of significance was set at P < 0.05.

3. Results

3.1. Metabolic activity of RAW 264.7 cells

RAW 264.7 cells grew well in both culture medium and mediumwith series material extracts. MTT analysis showed that the overallmetabolic activity of most groups with various ratios of materialextracts increased in a time-dependent manner (Fig. 1). The mediumwith highest solid/liquid ratios of material extracts (200mg/mL) wasused for the following in vitro studies.

3.2. Flow cytometry and fluorescence microscopy

Forward scatter decreased after stimulation of b-TCP extract,whereas side scatter increased with same treatment (Fig. 2 A, D),indicating that RAW 264.7 cells experienced morphology changesin response to the stimulation. Besides the morphology change, themean fluorescence intensity of CD163 increased after the materialstimulation (Fig. 2 B, E). On the contrary, the mean fluorescenceintensity of CD11c decreased under the same treatment (Fig. 2H, K).Fluorescence microscopic analysis showed CD11c or CD163 posi-tively stained cells in both b-TCP and control groups (Fig. 2C, F, I,and L).

3.3. Ionic concentration

The concentrations of Ca2þ and PO3�4 ions in all collected sam-

ples are shown in Table 2. After soaking with b-TCP powders

Fig. 1. The overall metabolic activity of RAW 264.7 cells in medium with differentconcentration of material extracts, detected by MTT. *Significant difference (P < 0.05)compared day1 groups with day3 groups.

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181510

(200 mg/mL), both Ca2þ and PO3�4 ionic concentrations in the cul-

ture medium were decreased. Although both Ca2þ and PO3�4 ionic

concentrations in the material extract were increased after beingconditioned by RAW 264.7 cells, they were still lower than that ofmacrophage conditioned complete culture medium.

3.4. Inflammatory gene expression of RAW 264.7 cells

Anti-inflammatory genes IL-10 and IL-1ra expression weresignificantly upregulated by the stimulation of b-TCP extract(P < 0.05) (Fig. 3 A, B). On the contrary, inflammatory genes IL-1band IL-6 expression were significantly downregulated with thesame treatment (P < 0.05) (Fig. 3 D, E). TNFa gene expressionshowed no significant change after stimulation (Fig. 3C).

To explore the mechanism of inflammation related geneexpression changes, we examined two calcium involved inflam-mation signalling pathways (Wnt5A/Ca2þ, Ca2þ/CaSR). AlthoughWnt5A gene expression was significantly upregulated (P < 0.05),the downstream molecules Fz5 and CamKII showed no significantchange (Fig. 4 A, B, D). This indicated that the inflammatory effectmay not be caused though this pathway. We then examined theCaSR gene expression, which showed an activation of this pathwaywith significantly upregulated CaSR genes and increased IkB-aprotein (P < 0.05, Fig. 4C & D).

3.5. Osteogenic gene expression of RAW 264.7 cells

BMP2 gene expression was upregulated by the stimulation of b-TCP extract (P < 0.05) (Fig. 5 A). However, VEGF and TGF-b1 geneexpression showed no significant change with the same treatment(Fig. 5 B, C).

3.6. Osteogenic differentiation of BMSCs

Expression of three measured mineralization-related genes(OPN, OCN, COL1) by BMSCs stimulated by RAW 264.7 cell-conditioned media containing b-TCP extracts were significantlyupregulated compared to b-TCP extract alone (P < 0.05) (Fig. 6 A, B,

C, and D). However, RAW 264.7 cell conditioned medium alonewithout b-TCP extract showed no significant changes in expressionof ALP and COL1 in comparison with the normal culture medium(P > 0.05) (Fig. 6 A, D).

Western blot of ALP and OPN showed a similar trend to theRT-qPCR results. Both measured proteins were significantly upre-gulated by BMSCs stimulated by RAW 264.7 cell conditioned b-TCPextract relative to b-TCP extract alone (P < 0.05) (Fig. 7).

Alizarin Red S staining showed evidence of calcium depositionand nodule formation. Mineralized nodules were observed in allgroups with osteogenic medium (Fig. 8). More distinct noduleswere observed in BMSCs stimulated by RAW 264.7 cell conditionedb-TCP extract compared to that by b-TCP extract alone (Fig. 8).

3.7. BMP2 signalling pathway-related gene expression

To further explore the mechanisms of enhanced osteogenesis,we examined the gene expressions of BMP2 signalling pathways.RAW 264.7 cell conditioned medium alone without b-TCP extractcould not upregulate the BMP2 signalling pathways related genes,compared with the normal culture medium. BMPR2 and SMAD4were even downregulated (Fig. 9 A, D). However, The expressions ofBMPR2, SMAD5 and SMAD4 in BMSCs culturedwith RAW264.7 cellconditioned b-TCP extracts showed significantly upregulationcompared to those cultured with b-TCP extract alone (P < 0.05)(Fig. 9 A, C, E).

4. Discussion

Macrophages play important roles in the interaction betweenthe host body and bone substitute materials. Such effects arecommonly attributed to an inflammatory response and materialdegradation. While these effects are undoubtedly important, ourstudy further revealed that in response to the b-TCP extract, mac-rophages could switch their phenotype to an M2 extreme, leadingto the release of osteoinductive molecules and anti-inflammatorycytokines, which enhanced the osteogenesis of BMSCs throughthe BMP2 pathway. These findings demonstrate that macrophageparticipated and made important contributions to the b-TCP stim-ulated osteogenesis.

4.1. Effects of b-TCP extracts on macrophages

The macrophage phenotype occurs along a continuum betweenM1 and M2 extremes [33], and can switch its phenotype inresponse to environmental change. Whether macrophages couldrespond to b-TCP extract leading to a phenotype switchwas still notclear, therefore, the effects of b-TCP extract on macrophages wereinvestigated first. Three criteria (functional properties, surfacemarkers, and inducers) were applied to define a specific macro-phage phenotype [11]. With the induction of b-TCP extract, moremacrophages expressed M2 surface marker CD163 and anti-inflammatory cytokines (IL-10 and IL-1ra). These results indicatedthat b-TCP extract elicited some effects on macrophage polariza-tion, which switched it towards the M2 extremity.

Calcium has been known to involve in some inflammation sig-nalling pathways [34,35]. Wnt/Ca2þ signalling pathway is amongone of them [34]. Wnt5A can bind to Fz5 activating the Wnt/Ca2þ

signalling pathway via CaMKII and Protein kinase C, which culmi-nates the expression of downstream inflammatory cytokine genesvia transcription factor NFkB [34]. When b-TCP was immersed inculture medium, dissolved Ca and P in the medium were removedfrom the solution, forming a deposition on surface of the ceramics,which led to the decrease of Ca2þ concentration in the culturemedia [36]. It is a logical extension to hypothesize that the decrease

Table 2The ionic concentrations of b-TCP powder extract (200 mg/mL) and macrophageconditioned media.

Material immersed e e þ þMacrophageconditioned

e þ e þ

Ionicconcentrations(mg/L)

Ca2þ 68.70 � 1.81 81.60 � 0.40 50.82 � 0.09 74.81 � 2.14PO3�

4 35.41 � 0.25 37.20 � 0.15 18.42 � 0.37 21.40 � 0.36

Fig. 2. FACS and fluorescence staining results: CD163 (AeF), CD11c (GeL); RAW 264.7 cells cultured in complete medium (AeC, GeI), macrophages stimulated by material extract(DeF, JeL).

Z. Chen et al. / Biomaterials 35 (2014) 1507e1518 1511

of Ca2þ concentration may inhibit the Wnt/Ca2þ signallingpathway, leading to the anti-inflammatory effects. However, Fz5and CaMKII showed no significant change by the stimulation ofb-TCP extract, while Wnt5Awas even found to be upregulated. Thissuggests that the observed anti-inflammatory effects were notproduced via this pathway.

Extracellular Ca2þ has also been found to be able to activate theCaSR signalling cascade leading to the production of Wnt5A, whichcan reduce the expression of TNFa via inhibition of NFkB anddownregulates the TNFR1 via the Wnt5a/Ror2 signalling pathway,

Fig. 3. Relative mRNA expressions of inflammation related genes IL-10 (A), IL-1ra (B), TNFa (C), IL-1b (D) and IL-6 (E), relative to housekeeping gene b-actin, by RAW 264.7 cells withor without stimulation of material extract. *Significant difference (P < 0.05) for b-TCP extract group compared to the negative group.

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181512

thereby reducing the inflammation [35]. In this study, both CaSRand Wnt5A gene expressions were upregulated by the materialstimulation, and the NFkB inhibitor IkB was also upregulated,which indicated that the anti-inflammatory effects might be causedthrough CaSR pathway. It should be noted that activation of CaSRsignalling pathway was always related to the high concentration ofCa2þ. However, the Ca2þ concentration in this study was decreasedby the material immersion. How the low concentration of Ca2þ

activates the CaSR signalling pathway requires further studies.

4.2. b-TCP stimulated macrophages enhanced the osteogenicdifferentiation of BMSCs

Although macrophages have been proved to be involved in theosteogenesis, there is still no consensus on which phenotype ismore beneficial for the osteogenesis. Guihard et al. reported that itwas the classically activated inflammatory M1, not M2 macro-phages, which induced osteogenesis in MSCs via Oncostatin M [37].However, it was also reported that culture media from lipopoly-saccharide activated macrophages induced pre-osteoblast cells todifferentiate towards fibroblasts even with osteogenic medium,which might be related to the upregulation of TNFa, IL-1b and IL-6[38,39].

After implantation, bone substitute materials are eitherdegraded and replaced by new bone, or encapsulated and replacedby inflammatory fibrous tissue. M2 macrophages participate inboth ends, via secreting relating cytokines. They may contributeosteoinductive and osteogenic cytokines (BMP2, VEGF etc) toenhance osteogenesis, but also elaborate inflammatory and fibrousagents (TNFa, VEGF, TGF-b1, TGF-b3 etc) to promote pathologicalfibrosis [39,40]. In addition, it was indicated that excess switch ofM2 phenotype may result in scar tissue or a delay inwound healing[41]. Therefore, the conclusion still could not be made that M2macrophages played a major role in regulating osteogenesis. Aneffective and adequate switch in macrophage phenotype might beof great importance for enhancing osteogenesis.

BMP2 is a founding member of the BMP family, acting as anosteoinductive agent [42e44]. On receptor activation, BMP2transmit signals through SMAD-dependent and SMAD-independent pathways, leading to osteoblast differentiation ofmesenchymal progenitor cells [45]. Macrophages produce BMP2,which contributes to osteogenesis during bone healing [39]. In thisstudy, we found that b-TCP extract significantly upregulated theBMP2 expression by macrophages. The released BMP2 could bindto the BMPR2, leading to the activation of SMAD5 [45]. ActivatedSMAD5 could then form a complex with SMAD4, which trans-locates into the nucleus, facilitating the expression of distallesshomeobox5 (Dlx5) [45]. Dlx5 has been shown to induce theexpression of Runt-related transcription factor 2 (Runx2), resultingin osteogenic differentiation of BMSCs [45]. The mechanism for theupregulation of BMP2 after b-TCP stimulation was still not clear. Inaddition to the anti-inflammatory effects, CaSRwas also reported tocause BMP2 synthesis and secretion via phosphatidylinositol3-kinase activation [46]. The deletion of CaSR completely inhibitedBMP2 secretion [46]. The upregulation of BMP2 might be related tothe upregulation of CaSR.

TNFa contributes to fibrosis via the mediator TNFR2, which thenacts through Extracellular signal-regulated kinases 1/2 (ERK1/2) tostimulate proliferation and inhibit collagen degradation [47]. TNFawas also found to activate the ERK-specific mitogen-activatedprotein kinase pathway, leading to increased TGF-b1 production infibroblasts [48]. TGF-b1 is a key cytokine in the fibrosis process,where it induces fibroblast proliferation via the Fibroblast growthfactor 2/ERK pathway [49]. VEGF is another important cytokine forfibrosis, playing fundamental role in the angiogenesis and vascu-logenesis [50]. All three tested fibrosis-promoting genes TNFa,VEGF and TGF-b1, showed no significant changes. This indicatesthat the b-TCP-induced macrophage phenotype switch did notenhance the inflammatory fibrosis, but was effective and adequatefor improving osteogenesis. These findings align with previousin vivo and clinical studies, demonstrating enhanced new boneformation by b-TCP [25,27,51,52].

Fig. 4. Relative mRNA expressions of WNT5A (A), Fz5 (B) and CaSR (C) relative to housekeeping gene b-actin by RAW 264.7 cells with or without stimulation of material extract.Western blotting analysis of CaMKII and IkB-a expression of RAW 264.7 cells with or without stimulation of material extract (D). *Significant difference (P < 0.05) for b-TCP extractgroup compared to the negative group.

Z. Chen et al. / Biomaterials 35 (2014) 1507e1518 1513

Possible effects of macrophages in the process of b-TCP materialinducing osteogenesis are summarized in Fig. 10. Degradationproducts released from the materials elicit some effects on themacrophages. The stimulated macrophages switch their phenotype

Fig. 5. Relative mRNA expressions of BMP2 (A), TGF b1 (B), and VEGF (C) relative to houseke*Significant difference (P < 0.05) for b-TCP extract group compared to the negative group.

into M2 via CaSR signalling pathway, resulting in the down-regulation inflammatory cytokines (IL-1b, IL-6), and the upregula-tion of anti-inflammatory cytokines (IL-10, IL-1ra). The stimulatedmacrophages also uprelulate BMP2 gene expression, which

eping gene b-actin by RAW 264.7 cells with or without stimulation of material extract.

Fig. 6. Fold changes of ALP (A), OPN (B), OCN (C) and COL1 (D), by comparing BMSCs cultured in complete culture medium with material immersed medium, or RAW 264.7 cellconditioned media. *Significant difference (P < 0.05).

Fig. 7. Western blot analysis of ALP and OPN expression by BMSCs cultured in complete culture medium, material immersed medium, or RAW 264.7 cell conditioned media.*Significant difference (P < 0.05).

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181514

Fig. 8. Alizarin Red S staining results: A. BMSCs cultured in complete culture medium only, B. BMSCs cultured in material extract, C. BMSCs cultured in complete culture mediumwith osteogenic supplements (OS), D. BMSCs cultured in material extract with OS, E. BMSCs cultured in RAW 264.7 cell conditioned complete culture medium with osteogenicsupplements (OS), F. BMSCs cultured in RAW 264.7 cell conditioned material extract with osteogenic supplements (OS).

Z. Chen et al. / Biomaterials 35 (2014) 1507e1518 1515

activates the BMP2 signalling pathway in BMSCs, leading to oste-ogenic differentiation.

For evaluating the in vitro osteogenesis capacity of bone sub-stitute materials, osteoblastic cells (e.g. preosteoblasts, osteoblasts)are typically used [53e55]. The potentially osteogenesis-promotingmaterials will then be further evaluated in vivo. However, thein vitro and in vivo biological responses to biomaterials are not al-ways consistent, due to the lack of immune cells in vitro [56,57].b-TCP itself was not sufficient to improve the osteogenic effect of

BMSCs, however, macrophages conditioned b-TCP could releaseosteogenic cytokines, thereby enhancing the osteogenic effect ofBMSCs, which was more consistent with in vivo results. Theconfirmation that macrophages behave actively during the processof osteogenesis caused by b-TCP is meaningful. Inadequately acti-vated macrophages might cause failure of bone regeneration,leading to undesired inflammatory fibrosis. Immune cells, espe-cially macrophages should be taken into consideration for evalu-ating the in vitro osteogenesis capacity of bone substitute materials.

Fig. 9. Relative mRNA expressions of BMP2 signalling pathway genes BMPR2 (A), BMPR1A (B), SMAD5 (C), SMAD1 (D) and SMAD4 (E) relative to housekeeping gene 18S rRNA.BMSCs cultured in complete culture medium, material immersed medium, or RAW 264.7 cell conditioned media. *Significant difference (P < 0.05).

Fig. 10. Summary of possible effects of macrophages in the process of b-TCP material inducing osteogenesis.

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181516

Z. Chen et al. / Biomaterials 35 (2014) 1507e1518 1517

It should be noted that only RAW 264.7 macrophage cell line andb-TCP bioceramics were used in this study. Future research will aimto test primary macrophages and other types of bone substitutematerials to draw more general conclusions.

5. Conclusions

Macrophages played important roles in b-TCP-induced osteo-genesis. b-TCP extracts elicited significant effects on macrophages,inducing an effective and adequate M2 macrophage phenotypeswitch, which inhibited the inflammation and enhance the osteo-genic differentiation of BMSCs. Adequately activated macrophagesare vital for the success of material-stimulated bone regeneration.The interactionwith immune cells, especially macrophages, shouldbe elucidated when evaluating the in vitro osteogenesis capacity ofbone substitute biomaterials.

Acknowledgements

The study was supported by an Australian Research CouncilDiscovery Project (DP120103697), Recruitment Program of GlobalYoung Talent, China (Dr. Wu) and Natural Science Foundation ofChina (Grant 81201202 and 31370963).

References

[1] Bryers JD, Giachelli CM, Ratner BD. Engineering biomaterials to integrate andheal: the biocompatibility paradigm shifts. Biotechnol Bioeng 2012;109:1898e911.

[2] Takayanagi H. Osteoimmunology: shared mechanisms and crosstalk betweenthe immune and bone systems. Nat Rev Immunol 2007;7:292e304.

[3] Walsh MC, Kim N, Kadono Y, Rho J, Lee SY, Lorenzo J, et al. Osteoimmunology:interplay between the immune system and bone metabolism. Annu RevImmunol 2006;24:33e63.

[4] Franz S, Rammelt S, Scharnweber D, Simon JC. Immune responses to implants -a review of the implications for the design of immunomodulatory biomaterials.Biomaterials 2011;32:6692e709.

[5] Takayanagi H. Inflammatory bone destruction and osteoimmunology.J Periodontal Res 2005;40:287e93.

[6] Lorenzo J. Interactions between immune and bone cells: new insights withmany remaining questions. J Clin Invest 2000;106:749e52.

[7] Li Y, Toraldo G, Li A, Yang X, Zhang H, Qian WP, et al. B cells and T cells arecritical for the preservation of bone homeostasis and attainment of peak bonemass in vivo. Blood 2007;109:3839e48.

[8] Pacifici R. The immune system and bone. Arch Biochem Biophys 2010;503:41e53.

[9] Bartneck M, Heffels KH, Pan Y, Bovi M, Zwadlo-Klarwasser G, Groll J. Inducinghealing-like human primary macrophage phenotypes by 3D hydrogel coatednanofibres. Biomaterials 2012;33:4136e46.

[10] Brown BN, Ratner BD, Goodman SB, Amar S, Badylak SF. Macrophage polari-zation: an opportunity for improved outcomes in biomaterials and regener-ative medicine. Biomaterials 2012;33:3792e802.

[11] Martinez FO, Sica A, Mantovani A, Locati M. Macrophage activation and po-larization. Front Biosci 2008;13:453e61.

[12] Mantovani A, Biswas SK, Galdiero MR, Sica A, Locati M. Macrophage plasticityand polarization in tissue repair and remodelling. J Pathol 2013;229:176e85.

[13] Mills CD, Kincaid K, Alt JM, Heilman MJ, Hill AM. M-1/M-2 macrophages andthe Th1/Th2 paradigm. J Immunol 2000;164:6166e73.

[14] Brown BN, Valentin JE, Stewart-Akers AM, McCabe GP, Badylak SF. Macro-phage phenotype and remodeling outcomes in response to biologic scaffoldswith and without a cellular component. Biomaterials 2009;30:1482e91.

[15] Brown BN, Londono R, Tottey S, Zhang L, Kukla KA, Wolf MT, et al. Macro-phage phenotype as a predictor of constructive remodeling following theimplantation of biologically derived surgical mesh materials. Acta Biomater2012;8:978e87.

[16] Lau SK, Chu PG, Weiss LM. CD163-A specific marker of macrophages inparaffin-embedded tissue samples. Am J Clin Pathol 2004;122:794e801.

[17] Alexander KA, Chang MK, Maylin ER, Kohler T, Muller R, Wu AC, et al. Ostealmacrophages promote in vivo intramembranous bone healing in a mousetibial injury model. J Bone Miner Res 2011;26:1517e32.

[18] Chang MK, Raggatt LJ, Alexander KA, Kuliwaba JS, Fazzalari NL, Schroder K,et al. Osteal tissue macrophages are intercalated throughout human andmouse bone lining tissues and regulate osteoblast function in vitro andin vivo. J Immunol 2008;181:1232e44.

[19] Pettit AR, Chang MK, Hume DA, Raggatt LJ. Osteal macrophages: a new twiston coupling during bone dynamics. Bone 2008;43:976e82.

[20] Honda Y, Anada T, Kamakura S, Nakamura M, Sugawara S, Suzuki O. Elevatedextracellular calcium stimulates secretion of bone morphogenetic protein 2 bya macrophage cell line. Biochem Biophys Res Commun 2006;345:1155e60.

[21] Wahl SM, McCartney-Francis N, Allen JB, Dougherty EB, Dougherty SF.Macrophage production of TGF-beta and regulation by TGF-beta. Ann N YAcad Sci 1990;593:188e96.

[22] Velard F, Braux J, Amedee J, Laquerriere P. Inflammatory cell response tocalcium phosphate biomaterial particles: an overview. Acta Biomater 2013;9:4956e63.

[23] Oh SA, Lee GS, Park JH, Kim HW. Osteoclastic cell behaviors affected by thealpha-tricalcium phosphate based bone cements. J Mater Sci Mater Med2010;21:3019e27.

[24] Xia Z, Huang Y, Adamopoulos IE, Walpole A, Triffitt JT, Cui Z. Macrophage-mediated biodegradation of poly(DL-lactide-co-glycolide) in vitro. J BiomedMater Res A 2006;79:582e90.

[25] Chappard D, Guillaume B, Mallet R, Pascaretti-Grizon F, Basle MF, Libouban H.Sinus lift augmentation and beta-TCP: a microCT and histologic analysis onhuman bone biopsies. Micron 2010;41:321e6.

[26] Rokn A, Moslemi N, Eslami B, Abadi HK, Paknejad M. Histologic evaluation ofbone healing following application of anorganic bovine bone and beta-tricalcium phosphate in rabbit Calvaria. J Dent (Tehran) 2012;9:35e40.

[27] Gaasbeek RD, Toonen HG, van Heerwaarden RJ, Buma P. Mechanism of boneincorporation of beta-TCP bone substitute in open wedge tibial osteotomy inpatients. Biomaterials 2005;26:6713e9.

[28] Mareddy S, Crawford R, Brooke G, Xiao Y. Clonal isolation and characterizationof bone marrow stromal cells from patients with osteoarthritis. Tissue Eng2007;13:819e29.

[29] Singh S, Jones BJ, Crawford R, Xiao Y. Characterization of a mesenchymal-likestem cell population from osteophyte tissue. Stem Cells Dev 2008;17:245e54.

[30] Zhang M, Wu C, Li H, Yuen J, Chang J, Xiao Y. Preparation, characterization andin vitro angiogenic capacity of cobalt substituted [small beta]-tricalciumphosphate ceramics. J Mater Chem 2012;22:21686e94.

[31] Petit A, Mwale F, Tkaczyk C, Antoniou J, Zukor DJ, Huk OL. Induction of proteinoxidation by cobalt and chromium ions in human U937 macrophages. Bio-materials 2005;26:4416e22.

[32] Birmingham E, Niebur GL, McHugh PE, Shaw G, Barry FP, McNamara LM.Osteogenic differentiation of mesenchymal stem cells is regulated by osteocyteand osteoblast cells in a simplified bone niche. Eur Cell Mater 2012;23:13e27.

[33] Mosser DM, Edwards JP. Exploring the full spectrum of macrophage activa-tion. Nat Rev Immunol 2008;8:958e69.

[34] De A. Wnt/Ca2þ signaling pathway: a brief overview. Acta Bioch Bioph Sin2011;43:745e56.

[35] MacLeod RJ, Hayes M, Pacheco I. Wnt5a secretion stimulated by the extra-cellular calcium-sensing receptor inhibits defective Wnt signaling in coloncancer cells. Am J Physiol Gastrointest Liver Physiol 2007;293:G403e11.

[36] Chen Z, Wu C, Yuen J, Klein T, Crawford R, Xiao Y. Influence of osteocytes inthe in vitro and in vivo beta-tricalcium phosphate-stimulated osteogenesis.J Biomed Mater Res A 2013. In press.

[37] Guihard P, Danger Y, Brounais B, David E, Brion R, Delecrin J, et al. Induction ofosteogenesis in mesenchymal stem cells by activated monocytes/macro-phages depends on oncostatin M signaling. Stem Cells 2012;30:762e72.

[38] Zheng F. The effect of PMMA stimulated complement-macrophage cascade onosteogenesis of preosteoblast-like MC3T3-E1 cells on PMMA surface. Ph.DDissertation. Ohio State University; 2010.

[39] Champagne CM, Takebe J, Offenbacher S, Cooper LF. Macrophage cell linesproduce osteoinductive signals that include bone morphogenetic protein-2.Bone 2002;30:26e31.

[40] Freytes DO, Kang JW, Marcos-Campos I, Vunjak-Novakovic G. Macrophagesmodulate the viability and growth of human mesenchymal stem cells. J CellBiochem 2013;114:220e9.

[41] BrownBN,BadylakSF.Expandedapplications, shiftingparadigmsandan improvedunderstanding of host-biomaterial interactions. Acta Biomater 2013;9:4948e55.

[42] Bais MV, Wigner N, Young M, Toholka R, Graves DT, Morgan EF, et al. BMP2 isessential for post natal osteogenesis but not for recruitment of osteogenicstem cells. Bone 2009;45:254e66.

[43] Fromigué O, Marie PJ, Lomri A. Bone morphogenetic protein-2 and trans-forming growth factor-b2 interact to modulate human bone marrow stromalcell proliferation and differentiation. J Cell Biochem 1998;68:411e26.

[44] Sitasuwan P, Lee LA, Bo P, Davis EN, Lin Y, Wang Q. A plant virus substrateinduces early upregulation of BMP2 for rapid bone formation. Integr Biol(Camb) 2012;4:651e60.

[45] Soltanoff CS, Yang S, Chen W, Li YP. Signaling networks that control thelineage commitment and differentiation of bone cells. Crit Rev Eukaryot GeneExpr 2009;19:1e46.

[46] Peiris D, Pacheco I, Spencer C, MacLeod RJ. The extracellular calcium-sensingreceptor reciprocally regulates the secretion of BMP-2 and the BMP antagonistNoggin in colonic myofibroblasts. Am J Physiol Gastrointest Liver Physiol2007;292:G753e66.

[47] Theiss AL, Simmons JG, Jobin C, Lund PK. Tumor necrosis factor (TNF) alphaincreases collagen accumulation and proliferation in intestinal myofibroblastsvia TNF receptor 2. J Biol Chem 2005;280:36099e109.

[48] Sullivan DE, Ferris M, Pociask D, Brody AR. Tumor necrosis factor-alpha in-duces transforming growth factor-beta1 expression in lung fibroblaststhrough the extracellular signal-regulated kinase pathway. Am J Respir CellMol Biol 2005;32:342e9.

Z. Chen et al. / Biomaterials 35 (2014) 1507e15181518

[49] Xiao L, Du Y, Shen Y, He Y, Zhao H, Li Z. TGF-beta 1 induced fibroblast prolif-eration is mediated by the FGF-2/ERK pathway. Front Biosci 2012;17:2667e74.

[50] Ahluwalia A, Tarnawski AS. Critical role of hypoxia sensoreHIF-1alpha inVEGF gene activation. Implications for angiogenesis and tissue injury healing.Curr Med Chem 2012;19:90e7.

[51] Jang BJ, Byeon YE, Lim JH, Ryu HH, Kim WH, Koyama Y, et al. Implantation ofcanine umbilical cord blood-derived mesenchymal stem cells mixed withbeta-tricalcium phosphate enhances osteogenesis in bone defect model dogs.J Vet Sci 2008;9:387e93.

[52] Zhang M, Wang K, Shi Z, Yang H, Dang X, Wang W. Osteogenesis of theconstruct combined BMSCs with beta-TCP in rat. J Plast Reconstr Aesthet Surg2010;63:227e32.

[53] Wu CT, Fan W, Zhou YH, Luo YX, Gelinsky M, Chang J, et al. 3D-printing ofhighly uniform CaSiO3 ceramic scaffolds: preparation, characterization andin vivo osteogenesis. J Mater Chem 2012;22:12288e95.

[54] Zhang ML, Wu CT, Lin KL, Fan W, Chen L, Xiao Y, et al. Biological responses ofhuman bone marrowmesenchymal stem cells to Sr-M-Si (M ¼ Zn, Mg) silicatebioceramics. J Biomed Mater Res A 2012;100A:2979e90.

[55] Wu C, Fan W, Gelinsky M, Xiao Y, Simon P, Schulze R, et al. Bioactive SrO-SiO2glass with well-ordered mesopores: characterization, physiochemistry andbiological properties. Acta Biomater 2011;7:1797e806.

[56] Chakravorty N, Ivanovski S, Prasadam I, Crawford R, Oloyede A, Xiao Y. ThemicroRNA expression signature on modified titanium implant surfaces in-fluences genetic mechanisms leading to osteogenic differentiation. Acta Bio-mater 2012;8:3516e23.

[57] Lang NP, Salvi GE, Huynh-Ba G, Ivanovski S, Donos N, Bosshardt DD. Earlyosseointegration to hydrophilic and hydrophobic implant surfaces in humans.Clin Oral Implants Res 2011;22:349e56.