at SciVerse ScienceDirect

Biomaterials 33 (2012) 7993e8002

Contents lists available

Biomaterials

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

The role of the Wnt/b-catenin pathway in the effect of implant topographyon MG63 differentiation

Wei Wang a,1, Lingzhou Zhao b,1, Qianli Ma a,1, Qintao Wang b, Paul K. Chu c,*, Yumei Zhang a,*

aDepartment of Prosthetic Dentistry, School of Stomatology, The Fourth Military Medical University, No. 145 West Changle Road, Xi’an 710032, ChinabDepartment of Periodontology and Oral Medicine, School of Stomatology, The Fourth Military Medical University, No. 145 West Changle Road, Xi’an 710032, ChinacDepartment of Physics and Materials Science, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong, China

a r t i c l e i n f o

Article history:Received 11 July 2012Accepted 29 July 2012Available online 11 August 2012

Keywords:Wntb�cateninMG63DifferentiationMicro/nano-textured topography

* Corresponding authors.E-mail addresses: paul.chu@cityu.edu.hk (P.K. C

(Y. Zhang).1 Co-first authors.

0142-9612/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.biomaterials.2012.07.064

a b s t r a c t

Wnt/b-catenin signaling plays a key role in bone formation. To assess the role of this signaling cascade inthe response of osteoblasts to the implant topography, human MG63 osteoblasts are cultured onmicropitted/nanotubular surface topographies (MNTs) and the transcriptional expressions of Wnt/b-catenin pathway receptors, activators, and inhibitors are measured. b-catenin signaling and cell differ-entiation are studied in the absence and presence of exogenous Dickkopf 1 (Dkk1) on the MNTs andexogenous Wnt3a on a smooth surface. The expressions of the Wnt/b-catenin pathway receptor low-density lipoprotein receptor-related protein 6 and pathway ligand Wnt3a are up-regulated by theMNTs whereas those of the pathway inhibitors including Dkk1/2 and secreted frizzled-related protein 1/2 are down-regulated by the MNTs, indicating regulation of the Wnt/b-catenin pathway modulators toactivate the pathway. Consequently, the b-catenin signaling activity is enhanced by the MNTs as well ascell differentiation in terms of osteogenesis-related gene expressions and alkaline phosphatase andcollagen products. On the smooth surface, exogenous Wnt3a stimulates b-catenin signaling and celldifferentiation while exogenous Dkk1 attenuates the enhancement by the MNTs. The results explicitlydemonstrate that the implant topography regulates the product of the Wnt/b-catenin pathway modu-lators from the cells and in turn activates the cell Wnt/b-catenin pathway promoting osteoblastdifferentiation.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The surface topography of a biomedical implant plays animportant role in regulating protein adsorption and cell focaladhesion assembly, which change the intracellular signaling path-ways and consequently influence the cell phenotype and overallbiological response to the implant [1e3]. Various types of topog-raphies on the micro- and nanoscale have been developed to targetbetter osseointegration [4]. Since the natural bone extracellularmatrix (ECM) is composed of nano- to microscale functional blocks,a hierarchical micro/nano-textured topography (MNT) is expectedto yield better biological effects. The MNTs combining nanotubesand micropitted topography exhibit more pronounced effects onosteoblast maturation as well as mesenchymal stem cell osteogenicdifferentiation [5,6]. Nonetheless, the molecular mechanism by

hu), wqtzym@fmmu.edu.cn

All rights reserved.

which the topographical cue affects the functions of cells andtissues is still not well understood and this has hampered optimi-zation of biomaterials topography.

The Wnt/b-catenin pathway which plays an essential role inbone mass and bone cell functions [7e9] is involved in theresponses of cells to various stimulants including bone morpho-genetic protein (BMP) [10], strain [11], oxygen-related stress [12],and implant surface properties [13]. It has also been shown that theWnt/b-catenin pathway mediates the biological effects of theimplant surface topography [14e16], although how the topo-graphical cues affect theWnt/b-catenin pathway is not well known.In addition to the direct influence on cell functions through cells/biomaterials interactions, biomaterials also modulate the cellsecretion profiles to indirectly affect cell behaviors via autocrine/paracrine modes [17]. b-catenin cytosol accumulation and nucleustranslocation, the key event of the canonical Wnt pathway activa-tion, are comprehensively modulated by Wnt proteins and a largenumber of antagonists secreted by cells. The canonical Wntpathway is initiated by Wnt proteins [18]. Furthermore, there isa large number of antagonists in the Wnt/b-catenin pathway,including the Dickkopf (Dkk) family and secreted frizzled-related

W. Wang et al. / Biomaterials 33 (2012) 7993e80027994

protein (sFRP) [19e21]. Hence, the surface topography may influ-ence the osteoblast functionalities by regulating the Wnt/b-cateninpathway modulators secreted from the cells that in turn modulatethe cell Wnt/b-catenin pathway. To test the hypothesis, humanMG63 osteoblasts are cultured on the MNTs combining the nano-tube and micropitted topography and the transcriptional expres-sions of the Wnt/b-catenin pathway receptors, activators, andinhibitors are measured in this work. The b-catenin signaling andcell differentiation are studied in the presence and absence ofexogenous Dkk1 for cells on the MNTs and exogenous Wnt3a forcells on a smooth surface. This study aims at advancing ourunderstanding of the biological effects of implant topographies andproviding insight into how implant osseointegration can besystematically enhanced.

2. Materials and methods

2.1. Specimen preparation

Pure titanium (99.9%, 10 � 10 � 1 mm3, Northwest Institute for NonferrousMetal Research, China) was used as the substrate. After polishingwith SiC sandpaperfrom 400 to 1500 grits and ultrasonic cleaning, the samples were treated with 0.5 wt% hydrofluoric acid for 30 min, rinsed with distilled water, and dried. The sampleswere anodized in an electrolyte containing 0.5 wt % hydrofluoric acid and 1 M

phosphoric acid for 1 h with a DC power supply and a platinum cathode at 5 and20 V to fabricate the MNTs (R-5 and R-20). The polished smooth surface (S) was usedas the control. The morphology of the samples was inspected by field-emissionscanning electron microscopy (FE-SEM, S-4800, Hitachi, Japan). The samples weresterilized by cobalt 60 before cell plating.

2.2. Cell culture

Human MG63 osteoblasts obtained from ATCC company were cultured in Dul-becco’s modified Eagle’s medium (DMEM, Gibco) supplemented with 10% fetalbovine serum (FBS, Gibco) and 1% penicillin/streptomycin and incubated ina humidified atmosphere of 5% CO2 at 37 �C. Only early passage cells were used inthe experiments.

2.3. Quantitative real time PCR

The MG63 cells were seeded on the samples at a density of 2 � 104/well andcultured for 3 and 7 days to evaluate the gene expressions of the Wnt3a, Wnt5a,Axin2, low-density lipoprotein receptor-related protein 5 (LRP5), LRP6, sFRP1/2,Dkk1 and Dkk2. The MG63 cells cultured on the MNTs at a density of 2 � 104 cell/well were treated with 100 ng/mL of human recombinant (rh)Wnt3a (R&D System),and those on the smooth surfacewere treatedwith theWnt inhibitor human rhDkk1(R&D System). After total incubation for 7 days, the expressions of runt-relatedtranscription factor 2 (Runx2), alkaline phosphatase (ALP), BMP, and collagen typeI (ColI) were determined. The total RNA was isolated using the Trizol reagent(Invitrogen). 1 mg of total RNAwas converted to cDNA using the the PrimeScript� RTreagent kit (TaKaRa). The real-time PCR reactions were performed using SYBRPremix Ex� Taq II (TaKaRa) on the CFX96� PCR System (Bio-rad). b-actin was usedas a housekeeping gene and the primers are listed in Table 1.

Table 1Primers used in real time PCR.

Gene Forward primer sequence (50e30) Reverse primer sequence (50e30)

Axin2 GGAGAAATGCGTGGATACC GCTGCTTGGAGACAATGCWnt3a GTCCCGTCCCTCCCTTTC ACCTCTCTTCCTACCTTTCCCWnt5a TCTCAGCCCAAGCAACAAGG GCCAGCATCACATCACAACACLRP5 TGGATTTGAACTCGGACTC GGGAAGAGATGGAAGTAGCLRP6 GCAGAGGAGAACTATGAAAGC GTTGGAGGCAGTCAGAGGsFRP1 ATCAGCCAGTCTCAGATGCC AAATCGCCGTCTCTCTCAGGsFRP2 AAGGAAAAGCCCACCCGAATC ACAACAACCAACCAGACCCAAGDkk1 CCAGACCATTGACAACTACC CAGGCGAGACAGATTTGCDkk2 TGACTTGGGATGGCAGAATC CAGAAATGACGAGCACAGCRunx2 CACTGGCGCTGCAACAAGA CATTCCGGAGCTCAGCAGAATAAALP CCTTGTAGCCAGGCCCATTG GGACCATTCCCACGTCTTCACBMP CAACACCGTGCTCAGCTTCC TTCCCACTCATTTCTGAAAGTTCCColI TCCACATACCTTTATTCCAGGAATC CCCGGGTTTAGAGACAACTTCb-actin TGGCACCCAGCACAATGAA CTAAGTCATAGTCCGCCTAGAAGCA

2.4. Western blot assay

The MG63 cells cultured on the MNTs at a density of 2 � 104 cells/well weretreated with 100 ng/mL of human rhWnt3a, and those on the smooth surface weretreated with the Wnt inhibitor human rhDkk1. The culture medium containingeither Wnt3a or Dkk1 was changed every 48 h for a total period of 7 days. For totalcellular proteins, the cells were lysed in the RIPA buffer (150 mM NaCl, 1% deoxy-cholate, 50 mM Tris (PH 7.4), 5 mM EDTA, 1% TritonX-100, 1 mM NaF, and 1 mM

Na3VO4). Alternatively, the cytosolic and nuclear fractions were prepared using theNuclear and Cytoplasmatic Extraction Kit (Millipore, Billerica, MA, USA). Equalamounts of extracts were separated by 10% SDS-PAGE and transferred to the poly-vinylidene fluoride membrane (Bio-Rad). Blots were blocked for 1 h in 5% bovineserum albumin (BSA, Gibco), followed by incubation with the primary antibodiesovernight at 4 �C and then the horseradish peroxidase-conjugated anti-rabbit oranti-mouse antibody for 1 h at room temperature. Blots were analyzed usingWestern-Light Chemiluminescent Detection System (Peiqing, China). The mono-clonal antibody against b-cateninwas purchased from Cell Signaling Technology andmonoclonal antibody against aetubulin was acquired from Abcam.

2.5. ALP staining

The Wnt3a and Dkk1 treatment processes were the same as above. The cellswere seeded on the substrates at a density of 2 � 104 cells/well and cultured in theosteogenic medium. The osteogenic medium was supplemented with 10 mM b-glycerophosphate (Sigma), 50 mg/mL ascorbic acid (Sigma), and 10�7 M dexameth-asone (Sigma). After culturing for 7 days, the cells were washed with phosphatebuffered saline (PBS) and fixed, and ALP staining was performed with the BCIP/NBTalkaline phosphatase color development kit (Beyotime) for 15 min. The stain waswashed with PBS thrice and then images were acquired.

2.6. Collagen secretion

The cell culture and Wnt3a and Dkk1 treatment processes were the same asthose in the ALP staining assay. After culturing for 14 days, the cells were washedwith PBS, fixed in 4% paraformaldehyde, and stained for collagen secretion in 0.1 wt% sirius red (Sigma) in saturated picric acid for 18 h. The unbound stain was washedwith 0.1 M acetic acid before images were taken. In the quantitative analysis, thestain on the specimens was eluted in 500 mL of destain solution (0.2 M NaOH/methanol 1:1) and the optical density at 540 nm was measured on a spectropho-tometer (Bio-tek, Germany).

2.7. Cell viability assay

The cell culture and Wnt3a and Dkk1 treatment processes were the same asthose in the Western blot assay. After culturing for 3 and 7 days, the cell vitality wasassessed by the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT,Sigma) assay. At the prescribed time points, the samples were rinsed thrice by PBSand transferred to new 24 well culture plates. The MTT solution was added and thesamples were incubated at 37 �C to allow formazen formation, which was dissolvedwith dimethyl sulfoxide. The optical density was measured at 490 nm on thespectrophotometer.

2.8. Cell apoptosis analysis

The cell culture and Wnt3a and Dkk1 treatment procedures were the same asthose in the Western blot assay. To determine cell apoptosis, an apoptosis detectionkit (BD Pharmingen) was used. After culturing for 3 days, the cells were trypsinized,washed with PBS, and resuspended in binding buffer at 1 � 106 cells/mL. 500 mL ofthe cell suspensionwas added to a flow tube and then 5 mL annexin V-FITC and 10 mLpropidium iodide were added to each tube. After incubation in dark at roomtemperature for 10 min, fluorescence was measured immediately on a flowcytometer (FACSVantage SE, BD Biosciences).

2.9. Statistical analysis

All data were expressed as means � standard deviations from at least threeindependent experiments. The data were analyzed by one way ANOVA combinedwith Student-Newman-Keuls post hoc test or Student’s t-test using SPSS 17.0 soft-ware (SPSS, USA). A p value of < 0.05 was considered to be significant.

3. Results

3.1. SEM characterization of the MNTs

The morphology of the fabricated samples is examined by SEM(Fig. 1). At a low magnification, the smooth surface is relatively flathaving parallel grooves, and R-5 and R-20 display a rougher

Fig. 1. SEM pictures showing the morphology of the samples.

W. Wang et al. / Biomaterials 33 (2012) 7993e8002 7995

micropitted morphology. The high-magnification pictures revealthat nanotubes of about 30 and 100 nm are distributed evenly on R-5 and R-20, while there is no obvious nanoscale cue on the smoothsurface.

3.2. Expressions of Wnt/b-catenin pathway modulators on theMNTs

The expressions of Wnt/b-catenin pathway modulators areassessed by real time PCR (Fig. 2). After culturing for 7 days, theWnt3a expression is significantly increased by theMNTs, while thatof Wnt5a is not. The Axin2 expression shows no discernibledifference among the samples. With regard to the Wnt receptors,the expression of LRP5 displays no significant difference among thesurfaces, but that of LRP6 is enhanced by the MNTs at day 3. Theexpressions of Wnt/b-catenin pathway inhibitors including sFRP1,sFRP2, Dkk1, and Dkk2 are down-regulated by the MNTs.

3.3. b-Catenin signaling activation on the MNTs

The nuclear amount of b-catenin which is the marker for the b-catenin signaling activation is examined by Western blot afterincubation for 7 days (Fig. 3). The nuclear b-catenin levels on the

MNTs are 2 folds higher than those on the smooth surface, butthose on R-5 and R-20 show no obvious difference.

3.4. Effect of exogenous Dkk1 or Wnt3a on b-catenin signalingactivity

In the presence and absence of exogenous Dkk1 for cells on theMNTs and exogenous Wnt3a for cells on the smooth surface for 7days, the nuclear b-catenin levels are assessed by Western blot todetermine the activation of b-catenin signaling (Fig. 4). The exog-enous Wnt3a induces one-fold increase in the nuclear b-cateninamount on the smooth surface. In comparison, the exogenous Dkk1dramatically decreases the nuclear b-catenin amounts on the MNTsto a level similar to that on the smooth surface in the absence ofWnt3a.

3.5. Effect of exogenous Dkk1 or Wnt3a on osteogenesis-relatedgene expressions

In the absence and presence of exogenous Dkk1 for cells on theMNTs and exogenous Wnt3a for cells on the smooth surface for 7days, the osteogenesis-related gene expressions are monitored byreal time PCR (Fig. 5). The ALP and BMP mRNA expressions are

Fig. 2. Wnt pathway gene expressions by MG63 cells after incubation of 3 and 7 days of culture on the samples (a p < 0.05 compared to smooth surface, b p < 0.05 compared toR-5).

W. Wang et al. / Biomaterials 33 (2012) 7993e80027996

obviously enhanced by the MNTs, especially R-20, and the Runx2and ColI expressions are also slightly promoted by the MNTs. Theexogenous Wnt3a significantly increases the expressions ofosteogenesis-related genes on the smooth surface to levelscomparable to those on the MNTs in the absence of Dkk1. Dkk1significantly ablates the enhanced osteogenesis-related geneexpressions by theMNTs to be similar to or even slightly lower thanthose on the smooth surface.

3.6. ALP staining

The cell ALP product in the presence and absence of exoge-nous Wnt3a or Dkk1 is stained (Fig. 6). The MNTs inducesignificantly higher ALP amounts than the smooth surface.Wnt3a significantly increases the cell ALP product on the smoothsurface and Dkk1 largely attenuates the enhanced cell ALPproduct by the MNTs.

3.7. Collagen secretion

Cell collagen secretion in the absence and presence of exoge-nous Dkk1 orWnt3a is quantified by Sirius Red staining (Fig. 7). TheMNTs lead to obviously more collagen secretion than the smoothsurface. Exogenous Wnt3a dramatically promotes collagen secre-tion by one fold on the smooth surface. On the other hand, theelevated collagen secretion by the MNTs is greatly attenuated bythe exogenous Dkk1 and this effect is more evident on R-20.

3.8. Cell viability

In the presence and absence of exogenous Wnt3a or Dkk1, thecell viability on the samples during the first 7 days of incubation isassessed (Fig. 8). The MNTs induce no obvious difference in the cellviability compared to the smooth surface. The exogenous Wnt3ashows no effect on the cell vitality on the smooth surface, while the

Fig. 3. Western blot and semi-quantitative analysis of b-catenin signaling activation inMG63 cells on the samples after incubation for 7 days. a�tubulin is used as a controlfor equal loading. a p < 0.05 compared to each untreated group.

W. Wang et al. / Biomaterials 33 (2012) 7993e8002 7997

exogenous Dkk1 produces differential effects on the cell vitality interms of different nanotubular diameters. Reduced cell viability isobserved from R-5 in response to Dkk1, while the cell viability onR-20 is not affected by Dkk1.

3.9. Cell apoptosis analysis

The proportion of apoptotic cells on each surface is measured byflow cytometer in the absence and presence of exogenousWnt3a orDkk1 for 3 days (Fig. 9). The MNTs do not lead to obvious cell

Fig. 4. Western blot and semi-quantitative analysis of nuclear b-catenin levels in theMG63 cells cultured on the samples for 7 days. The cells on the MNTs are treated withexogenous Dkk1 and those on the smooth surface are treated with exogenous Wnt3a.a�tubulin is used as a control for equal loading. a p < 0.05 compared to each untreatedgroup.

apoptosis compared to the smooth surface. The exogenous Wnt3aor Dkk1 do not influence cell apoptosis on the smooth surface orthe MNTs.

4. Discussion

The proper implant surface topographies such as the MNTs havebeen found to deliver enhanced osteogenic properties [5,6,22], butthe biological mechanisms responsible for these findings are stillnot well understood. In this study, we find that the MNTs enhanceMG63 cell differentiation in terms of up-regulating theosteogenesis-related gene expressions and enhancing the ALP andcollagen product. These effects are related to the enhancement inthe Wnt3a expression as well as inhibition in the expressions ofWnt/b-catenin pathway inhibitors including sFRP1, sFRP2, Dkk1and Dkk2 and consequent b-catenin signaling activation. On thesmooth surface, the exogenous Wnt3a significantly enhances b-catenin signaling and cell differentiation. The exogenous Dkk1obviously attenuates enhanced b-catenin signaling and cell differ-entiation by the MNTs. Hence, the topography of the biomaterialscan enhance the expressions of Wnt protein and its receptor whilesimultaneously inhibiting the Wnt pathway inhibitor expressionsto activate the Wnt/b-catenin pathway and promote osteoblastdifferentiation (Fig. 10).

The MNTs significantly enhance MG63 cell differentiation interms of the higher mRNA expressions of Runx2, ALP, BMP and ColIas well as the more ALP and collagen product. Runx2 is a tran-scription factor essential to osteoblast differentiation [23]. The ALPregulate phosphate metabolism via hydrolyzation of phosphateesters and is an early marker for osteoblast differentiation [24].BMP that belongs to the TGF-b superfamily is essential to osteo-genic differentiation and bone formation [25]. ColI is the main ECMprotein in bones [26] and one of the most widely recognizedbiochemical markers in osteoblast differentiation. Up-regulation ofthe expressions of these genes demonstrates the promoting effectsof the MNTs on osteoblast differentiation. This is further corrobo-rated by the larger amounts of ALP and collagen product on theMNTs. The present results are in line with our previous observationthat the MNTs significantly promote primary osteoblast differen-tiation [6].

The Wnt/b-catenin pathway is an important regulator of boneformation through action on cells of the osteoblast lineage andessentially each step of the osteogenic process can be affected bythis pathway [8]. The Wnt/b-catenin pathway is stimulated byWntproteins, which binding to the Frizzled (FZD) receptor and the co-receptor LRP5/6 leads to activation of Dishevelled and thus inhi-bition of a complex comprising Axin, glycogen synthase kinase 3b(GSK3b), and adenomatous polyposis coli. Consequently, GSK3b isunable to phosphorylate b-catenin and instead, b-catenin accu-mulates in the cytoplasm, translocates into the nucleus to reactwith the transcription factor T cell factor (TCF), and to activatetarget genes [18]. There is a number of endogenous Wnt antago-nists including the Dkk family and sFRPs. Dkk1 and Dkk2 bind toLRP5/6 and prevent the formation of the WnteFZDeLRP complexto inhibit the canonical Wnt signaling pathway [19,20]. sFRPspossess a cysteine-rich domain similar to FZD and they act either bybinding directly to the Wnt proteins or forming dimers with FZD toform non-functional complexes thereby inhibiting the Wnt/b-cat-enin pathway [21]. We study whether the expressions of theseWnt/b-catenin pathway modulators are influenced by the MNTs.The Wnt receptor LRP6 that is required for bone formation [27] isup-regulated by theMNTs. Interestingly, the expression of theWnt/b-catenin pathway activatorWnt3a is enhanced by theMNTs, whilethat of the non-canonical Wnt pathway activator Wnt5a is notaffected. On the contrary, the mRNA levels of the Wnt antagonists

Fig. 5. Real time PCR analysis of Runx2, ALP, BMP and Col1 expressions in MG63 cells after 7 days of incubation on the samples. The cells on the MNTs were treated with exogenousDkk1 and those on the smooth surface were treated with exogenous Wnt3a. a p < 0.05 compared to each untreated group.

Fig. 6. ALP staining of the MG63 cells after culturing for 7 days in the osteogenic differentiation medium. The cells on the MNTs are treated with exogenous Dkk1 and those on thesmooth surface are treated with exogenous Wnt3a.

W. Wang et al. / Biomaterials 33 (2012) 7993e80027998

Fig. 7. Optical images and colorimetric quantification of collagen secretion by MG63 cells cultured in osteogenic medium for 14 days. The cells on the MNTs are treated withexogenous Dkk1 and those on the smooth surface are treated with exogenous Wnt3a. a p < 0.05 compared to each untreated group.

Fig. 8. MTT assay for MG63 cell viability after culturing for 3 and 7 days on differentsamples. The cells on the MNTs are treated with exogenous Dkk1 and those on thesmooth surface are treated with exogenous Wnt3a. a p < 0.05 compared to eachuntreated group.

W. Wang et al. / Biomaterials 33 (2012) 7993e8002 7999

sFRP1, sFRP2, Dkk1 and Dkk2 are all depressed. The Western blotassay results verify the activation of b-catenin signaling. Hence, theMNTs promote osteoblast differentiation by, at least partly, the dualeffects of enhancing the expressions of the Wnt protein andreceptor and inhibiting the Wnt inhibitor expressions to activate b-catenin signaling. These results are consistent with the previousfindings of higher LRP5 expression and decreased Dkk1 expressioninMC3T3 cells cultured on silicon incorporated porous TiO2 coating[15]. However, on microstructured titanium surfaces, reducedWnt3a expression, increased non-canonical Wnt pathway ligandWnt5a, and increased Dkk2 secretion by osteoblasts have beenreported [2,28e30]. The contradiction appears to arise from thedifference in sample topography. Compared to the microstructuredtitanium surfaces, the MNTs in our study have nanostructured cuesand the nanocues have been shown to significantly induce b-cat-enin signaling [31,32].

The biomaterials not only affect cell functions directly throughcells/biomaterials interaction, but also modulate the cell microen-vironment by influencing the cell secreting profiles to affect the cellbehavior indirectly [17]. Our present results indicate that the MNTs

Fig. 9. Analysis of MG63 cell apoptosis on the samples in the presence and absence of Dkk1 for cells on the MNTs andWnt3a for cells on the smooth surface. The lower left quadrantcontains viable cells, the upper left quadrant contains PI-positive cells, and the two right quadrants contain annexin-V positive cells. The apoptotic cells are located in the two rightquadrants.

Fig. 10. Schematic diagram showing the details of the Wnt/b-catenin that mediates the effect of the topography on osteoblast differentiation. The topographical cue up-regulatesthe Wnt3a expression and inhibits the Dkk1/2 and sFRP1/2 expressions, which in turn activates the cell Wnt/b-catenin signaling in the autocrine/paracrine modes to promote celldifferentiation.

W. Wang et al. / Biomaterials 33 (2012) 7993e80028000

W. Wang et al. / Biomaterials 33 (2012) 7993e8002 8001

may modulate the Wnt modulators in the microenvironmentaround the cell consequently leading to activation of the Wnt/b-catenin pathway through the autocrine/paracrine modes. Actually,it has been demonstrated that the Wnt autocrine/paracrine loopmediates the effect of BMP-2 in pre-osteoblastic cells [10]. Forverification, we study whether the exogenous Wnt3a can enhancecell differentiation on the smooth surface. Wnt3a increases theb-catenin signaling activity on the smooth surface to a level slightlyhigher than those on the MNTs. Consequently, osteoblast differ-entiation is also significantly enhanced byWnt3a. At the same time,we study whether the Wnt inhibitor Dkk1 influences theenhancing effect of the MNTs on osteoblast differentiation. Asexpected, Dkk1 attenuates the enhanced b-catenin signalingactivity on the MNTs, and this is in line with the widely reportedeffect of Dkk1 [33]. Furthermore, the enhanced expressions of theosteogenesis-related genes, ALP product, and collagen secretion bythe MNTs are significantly reduced by Dkk1. The data definitelyconfirm our hypothesis demonstrating that the osteoblast differ-entiation promoting effect of the MNTs is mediated by the cellsecreted Wnt modulators in terms of enhancing Wnt proteinsecretion and inhibiting product of Wnt/b-catenin pathway inhib-itors (Fig. 10).

Since the Wnt/b-catenin pathway is reported to have regulatingeffects on cell viability and cell apoptosis [34e36], we investigatethe effects of the MNTs on them and the role of the Wnt/b-cateninpathway in these events. The MNTs do not obviously alter the cellviability. We monitor the changes in the cell viability aftermanipulating the b-catenin signaling activity by exogenous Wnt3aor Dkk1. The exogenous Wnt3a does not affect the cell viability onthe smooth surface and Dkk1 does not produce any obviousdifference in the cell viability on R20 but surprisingly, it causessignificant decrease in the cell vitality on R-5. Olivares-Navarreteet al. have found that Dkk1 has no effect on the MG63 cell numberon microstructured surface [30]. Dkk1 shows a surface dependenteffect on osteoblast viability and only has effects on nanotubes ofa smaller tube size. Park et al. have reported that nanotubes withincreasing tube size induce higher rates of cell apoptosis [37].However, our results show that on all the samples, the cellapoptosis rates are small and no significantly difference is observedfrom the MNTs and smooth surface. These results are consistentwith by our recent report that nanotubes support mesenchymalstem cell proliferation and osteogenic differentiation withoutinducing obvious cell death [5]. It is suggested that the differentserum concentrations in the cell culture in the different studiesmay account for the inconsistent results (2% used by Park et al and10% by us) [5,37]. In our studies, the 10% serum used in the cellculture leads to abundant proteins adsorbed onto the nanotubesthereby supporting cell functions without cell apoptosis [5]. Wnt3aor Dkk1 show no influence in cell apoptosis on the smooth surfaceor the MNTs and the small cell apoptosis rate reflects the goodcytocompatibility of the MNTs.

In this study, we attempt to gain deeper insight into themolecular mechanism associated with the biological effects of theimplant surface topography by uncovering the role of Wnt/b-cat-enin pathway in this process. This is expected to enrich ourknowledge about biomaterials modification or biofunctionalizationin order to accomplish better clinical performance. For example,Wnt3a may be loaded onto the implant surface and released toenhance osteoblast differentiation. In addition, lithium (Li) ionshave been reported to activate the Wnt/b-catenin by inhibitingGSK-3b and enhance osteoblast differentiation [38,39]. They havebeen incorporated into scaffolds to improve the biological perfor-mance [40]. The nanotubes are particularly ideal with respect toloading and delivering inorganic bioactive elements since they arestable and function at low doses thereby generating long-lasting

activity [41]. Hence, Li doped nanotubular structures withcontrolled Li release behavior may render better biological effects.

5. Conclusion

The MNTs enhance MG63 cell differentiation and the mecha-nism is related to the enhanced expressions of Wnt3a and Wntreceptor LRP6, inhibited expressions of Wnt/b�catenin pathwayinhibitors, and consequent b-catenin signaling activation. Theexogenous Wnt3a can significantly enhance b-catenin signalingactivation and cell differentiation on the smooth surface, and theexogenous Dkk1 attenuates the enhancement of them by theMNTs.The results verify that the topography of the biomaterials canregulate cell secretion of the Wnt modulators to activate the Wnt/b-catenin pathway in autocrine/paracrine modes therebypromoting osteoblast differentiation.

Acknowledgments

This work was supported by National Natural Science Founda-tion of China Nos. 81070862 and 31170915 and Hong Kong ResearchGrants Council (RGC) General Research Funds (GRF) No. CityU112510. L. Z. Zhao also thanks the grants from The Fourth MilitaryMedical University.

References

[1] Schwartz Z, Lohmann CH, Sisk M, Cochran DL, Sylvia VL, Simpson J, et al. Localfactor production by MG63 osteoblast-like cells in response to surfaceroughness and 1,25-(OH)2D3 is mediated via protein kinase C- and proteinkinase A-dependent pathways. Biomaterials 2001;22:731e41.

[2] Olivares-Navarrete R, Hyzy SL, Park JH, Dunn GR, Haithcock DA,Wasilewski CE, et al. Mediation of osteogenic differentiation of humanmesenchymal stem cells on titanium surfaces by a Wnt-integrin feedbackloop. Biomaterials 2011;32:6399e411.

[3] Wang L, Zhao G, Olivares-Navarrete R, Bell BF, Wieland M, Cochran DL, et al.Integrin b1 silencing in osteoblasts alters substrate-dependent responses to1,25-dihydroxy vitamin D3. Biomaterials 2006;27:3716e25.

[4] Mendonca G, Mendonca DB, Aragao FJ, Cooper LF. Advancing dental implantsurface technologyefrom micron- to nanotopography. Biomaterials 2008;29:3822e35.

[5] Zhao L, Liu L, Wu Z, Zhang Y, Chu PK. Effects of micropitted/nanotubulartitania topographies on bone mesenchymal stem cell osteogenic differentia-tion. Biomaterials 2012;33:2629e41.

[6] Zhao L, Mei S, Chu PK, Zhang Y, Wu Z. The influence of hierarchical hybridmicro/nano-textured titanium surface with titania nanotubes on osteoblastfunctions. Biomaterials 2010;31:5072e82.

[7] Gaur T, Lengner CJ, Hovhannisyan H, Bhat RA, Bodine PV, Komm BS, et al.Canonical WNT signaling promotes osteogenesis by directly stimulatingRunx2 gene expression. J Biol Chem 2005;280:33132e40.

[8] Liu F, Kohlmeier S, Wang CY. Wnt signaling and skeletal development. CellSignal 2008;20:999e1009.

[9] Tamura M, Nemoto E, Sato MM, Nakashima A, Shimauchi H. Role of the Wntsignaling pathway in bone and tooth. Front Biosci (Elite Ed) 2010;2:1405e13.

[10] Rawadi G, Vayssiere B, Dunn F, Baron R, Roman-Roman S. BMP-2 controlsalkaline phosphatase expression and osteoblast mineralization by a Wntautocrine loop. J Bone Miner Res 2003;18:1842e53.

[11] Sunters A, Armstrong VJ, Zaman G, Kypta RM, Kawano Y, Lanyon LE, et al.Mechano-transduction in osteoblastic cells involves strain-regulated estrogenreceptor alpha-mediated control of insulin-like growth factor (IGF) I receptorsensitivity to ambient IGF, leading to phosphatidylinositol 3-kinase/AKT-dependent Wnt/LRP5 receptor-independent activation of beta-cateninsignaling. J Biol Chem 2010;285:8743e58.

[12] Galli C, Macaluso GM, Piemontese M, Passeri G. Titanium topography controlsFoxO/beta-catenin signaling. J Dent Res 2011;90:360e4.

[13] Galli C, Passeri G, Ravanetti F, Elezi E, Pedrazzoni M, Macaluso GM. Roughsurface topography enhances the activation of Wnt/b-catenin signaling inmesenchymal cells. J Biomed Mater Res A 2010;95:682e90.

[14] Vlacic-Zischke J, Hamlet SM, Friis T, Tonetti MS, Ivanovski S. The influence ofsurface microroughness and hydrophilicity of titanium on the up-regulationof TGFb/BMP signalling in osteoblasts. Biomaterials 2011;32:665e71.

[15] Wang Q, Hu H, Qiao Y, Zhang Z, Sun J. Enhanced performance of osteoblasts bysilicon incorporated porous TiO2 coating. J Mater Sci Technol 2012;28:109e17.

[16] Galli C, Piemontese M, Lumetti S, Ravanetti F, Macaluso GM, Passeri G. Actincytoskeleton controls activation of Wnt/b-catenin signaling in mesenchymal

W. Wang et al. / Biomaterials 33 (2012) 7993e80028002

cells on implant surfaces with different topographies. Acta Biomater 2012;8:2963e8.

[17] Schwartz Z, Olivares-Navarrete R, Wieland M, Cochran DL, Boyan BD. Mech-anisms regulating increased production of osteoprotegerin by osteoblastscultured on microstructured titanium surfaces. Biomaterials 2009;30:3390e6.

[18] Miller JR. The Wnts. Genome Biol 2002;3. REVIEWS3001.[19] Li X, Liu P, Liu W, Maye P, Zhang J, Zhang Y, et al. Dkk2 has a role in terminal

osteoblast differentiation and mineralized matrix formation. Nat Genet 2005;37:945e52.

[20] van der Horst G, van der Werf SM, Farih-Sips H, van Bezooijen RL, Lowik CW,Karperien M. Downregulation of Wnt signaling by increased expression ofDickkopf-1 and -2 is a prerequisite for late-stage osteoblast differentiation ofKS483 cells. J Bone Miner Res 2005;20:1867e77.

[21] Kawano Y, Kypta R. Secreted antagonists of the wnt signalling pathway. J CellSci 2003;116:2627e34.

[22] Wennerberg A, Albrektsson T. Effects of titanium surface topography on boneintegration: a systematic review. Clin Oral Implants Res 2009;20:172e84.

[23] Ducy P, Zhang R, Geoffroy V, Ridall AL, Karsenty G. Osf2/Cbfa1: a transcrip-tional activator of osteoblast differentiation. Cell 1997;89:747e54.

[24] Whyte MP. Hypophosphatasia and the role of alkaline phosphatase in skeletalmineralization. Endocr Rev 1994;15:439e61.

[25] Hoffmann A, Gross G. BMP signaling pathways in cartilage and bone forma-tion. Crit Rev Eukaryot Gene Expr 2001;11:23e45.

[26] Viguet-Carrin S, Garnero P, Delmas PD. The role of collagen in bone strength.Osteoporos Int 2006;17:319e36.

[27] Bodine PV, Komm BS. Wnt signaling and osteoblastogenesis. Rev EndocrMetab Disord 2006;7:33e9.

[28] Olivares-Navarrete R, Hyzy SL, Hutton DL, Dunn GR, Appert C, Boyan BD, et al.Role of non-canonical Wnt signaling in osteoblast maturation on micro-structured titanium surfaces. Acta Biomater 2011;7:2740e50.

[29] Chakravorty N, Ivanovski S, Prasadam I, Crawford R, Oloyede A, Xiao Y. ThemicroRNA expression signature on modified titanium implant surfacesinfluences genetic mechanisms leading to osteogenic differentiation. ActaBiomater 2012;8:3516e23.

[30] Olivares-Navarrete R, Hyzy S, Wieland M, Boyan BD, Schwartz Z. The roles ofWnt signaling modulators Dickkopf-1 (Dkk1) and Dickkopf-2 (Dkk2) and cell

maturation state in osteogenesis on microstructured titanium surfaces.Biomaterials 2010;31:2015e24.

[31] Schernthaner M, Reisinger B, Wolinski H, Kohlwein SD, Trantina-Yates A,Fahrner M, et al. Nanopatterned polymer substrates promote endothelialproliferation by initiation of b-catenin transcriptional signaling. Acta Biomater2012;8:2953e62.

[32] Yu W, Zhang Y, Xu L, Sun S, Jiang X, Zhang F. Microarray-based bioinformaticsanalysis of osteoblasts on TiO2 nanotube layers. Colloids Surf B Biointerfaces2012;93:135e42.

[33] Liu N, Shi S, Deng M, Tang L, Zhang G, Ding B, et al. High levels of b-cateninsignaling reduce osteogenic differentiation of stem cells in inflammatorymicroenvironments through inhibition of the noncanonical Wnt pathway.J Bone Miner Res 2011;26:2082e95.

[34] Bodine PV. Wnt signaling control of bone cell apoptosis. Cell Res 2008;18:248e53.

[35] Xia JJ, Pei LB, Zhuang JP, Ji Y, Xu GP, Zhang ZP, et al. Celecoxib inhibits b-catenin-dependent survival of the human osteosarcoma MG-63 cell line. J IntMed Res 2010;38:1294e304.

[36] Bonewald LF, Johnson ML. Osteocytes, mechanosensing and Wnt signaling.Bone 2008;42:606e15.

[37] Park J, Bauer S, von der Mark K, Schmuki P. Nanosize and vitality: TiO2nanotube diameter directs cell fate. Nano Lett 2007;7:1686e91.

[38] O’Brien WT, Harper AD, Jove F, Woodgett JR, Maretto S, Piccolo S, et al.Glycogen synthase kinase-3b haploinsufficiency mimics the behavioral andmolecular effects of lithium. J Neurosci 2004;24:6791e8.

[39] Galli C, Piemontese M, Lumetti S, Manfredi E, Macaluso GM, Passeri G. GSK3b-inhibitor lithium chloride enhances activation of Wnt canonical signaling andosteoblast differentiation on hydrophilic titanium surfaces. Clin Oral ImplantsRes doi:10.1111/j.1600-0501.2012.02488.x, in press.

[40] Han P, Wu C, Chang J, Xiao Y. The cementogenic differentiation of peri-odontal ligament cells via the activation of Wnt/b-catenin signallingpathway by Liþ ions released from bioactive scaffolds. Biomaterials 2012;33:6370e9.

[41] Zhao L, Wang H, Huo K, Cui L, Zhang W, Ni H, et al. Antibacterial nano-structured titania coating incorporated with silver nanoparticles. Biomate-rials 2011;32:5706e16.