Materials Science and Engineering C 37 (2014) 305–313
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Interaction of progenitor bone cells with different surface modifications of titanium implant Wen-Cheng Chen a,⁎, Ya-Shun Chen a, Chia-Ling Ko a,b, Yi Lin c, Tzu-Huang Kuo c, Hsien-Nan Kuo c a b c
Advanced Medical Devices and Composites Laboratory, Department of Fiber and Composite Materials, College of Engineering, Feng Chia University, Taichung 40724, Taiwan Dental Medical Devices and Materials Research Center, College of Dental Medicine, Kaohsiung Medical University, Kaohsiung 807, Taiwan Medical Device Development Division, Metal Industries Research & Development Centre, Kaohsiung 82151, Taiwan
a r t i c l e
i n f o
Article history: Received 25 May 2013 Received in revised form 2 December 2013 Accepted 8 January 2014 Available online 18 January 2014 Keywords: Titanium (Ti) Implant Surface modifications Roughness Cell mineralization
a b s t r a c t Changes in the physical and chemical properties of Ti surfaces can be attributed to cell performance, which improves surface biocompatibility. The cell proliferation, mineralization ability, and gene expression of progenitor bone cells (D1 cell) were compared on five different Ti surfaces, namely, mechanical grinding (M), electrochemical modification through potentiostatic anodization (ECH), sandblasting and acid etching (SLA), sandblasting, hydrogen peroxide treatment, and heating (SAOH), and sandblasting, alkali heating, and etching (SMART). SAOH treatment produced the most hydrophilic surface, whereas SLA produced the most hydrophobic surface. Cell activity indicated that SLA and SMART produced significantly rougher surfaces and promoted D1 cell attachment within 1 day of culturing, whereas SAOH treatment produced moderate roughness (Ra = 1.26 μm) and accelerated the D1 cell proliferation up to 7 days after culturing. The ECH surface significantly promoted alkaline phosphatase (ALP) expression and osteocalcin (OCN) secretion in the D1 cells compared with the other surface groups. The ECH and SMART-treated Ti surfaces resulted in maximum ALP and OCN expressions during the D1 cell culture. SLA, SAOH, and SMART substrate surfaces were rougher and exhibited better cell metabolic responses during the early stage of cell attachment, proliferation, and morphologic expressions within 1 day of D1 cell culture. The D1 cells cultured on the ECH and SMART substrates exhibited higher differentiation, and higher ALP and OCN expressions after 10 days of culture. Thus, the ECH and SMART treatments promote better ability of cell mineralization in vitro, which demonstrate their great potential for clinical use. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Ti implants have significant functions in orthopedics, bone reconstruction, and dentistry because of their biocompatibility, corrosion resistance, low weight, and exceptional mechanical properties [1]. Ti is widely used in manufacturing oral implants. However, if Ti surface is not treated properly, the implant material might cause biological interactions that lead to infections and gum inflammation [2,3]. The consequences are dangerous and possibly lead to increased treatment risks, costs, and medical resource consumption [4,5]. Aside from biocompatibility, implants should adapt and respond according to the biological environment after implantation [6]. Ti surface should be significantly modified to improve their physical and chemical properties for promoting faster osseointegration with greater efficiency. The modified Ti surfaces should enhance the attachment, proliferation, and differentiation of progenitor bone cells while the implant is in contact with the
⁎ Corresponding author at: Department of Fiber and Composite Materials, Feng Chia University, 100 Wenhwa Rd., Seatwen, Taichung 40724, Taiwan. Tel.: + 886 4 24517250x3413; fax: +886 4 24514625. E-mail addresses: [email protected], [email protected] (W.-C. Chen). 0928-4931/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2014.01.022
surrounding tissues to accelerate bone attachment and provide anchor strength to the mandibular and maxillary alveolar bones [7–10]. Several commercialized products have been derived using different physical surface modification methods, such as machine grinding (M) and sand/grit blasting with micrometer-rough and nanometer-rough Ti surfaces (e.g., in laser ablated method), which should be performed while fabricating a system for increased utilization [3,11]. Other important factors for Ti surfaces include increasing hydrophilicity and preventing the stress-shielding effects of stress cycles, as well as load and temperature changes [12–17]. Therefore, this study focused on improving and evaluating different plate Ti surface modifications. The methods used in this study were the application of micro-rough Ti surfaces via Al2O3 sandblasting and acid etching (SLA), heat treatment and alkali treatment (SMART), hydrogen peroxide and heat treatment (SAOH), and potentiostatic anodization in sulfate electrolytes through constant electric current supply (ECH) [4,18–21]. The products with these surface modifications are clinically used commercial products. For example, the products with roughened surfaces obtained by M and sand/grit blasting are SLA from Straumann AG, TiOblast™ and OssoSpeed™ from Astra Tech, and Anker from Alliance. The other products with roughened surfaces were obtained through a combination of SLA, heat, alkali, hydrogen peroxide, and potentiostatic anodization, such as OSSEOTITE and Nano Tite™ from Biomet 3i, whereas RBM and
306
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Bio Tite-H are from DIO SM dental implants. These commercialized products with various Ti surfaces were developed to generate roughened surfaces to guarantee stable mechanical tissue–implant interfaces and to reduce recovery time. The degree of osseointegration in the early stages of wound healing (1 month to 3 months) is critical for determining the mechanical stability of the implant. The recovery time using Ti oral implants without surface modifications is generally ~ 12 weeks after implantation. With surface modifications, the recovery time can be reduced to 6 weeks to 8 weeks, demonstrating the clinical advantage of shorter healing period [20,21]. Pure Ti implants are commercially used in current teeth reconstruction procedures, but their safety must first be considered in vitro using cells before proceeding with animal studies and clinical tests. This safety precaution helps adjust and evaluate surface modification techniques, as well as analyzing the efficacy of the techniques in vivo. Evaluating the effects of different surface modifications through in vitro tests contributes to the optimization of results. Despite the existence of various surface modifications, no specific standard has been established to determine the best method. The factors involved in various types of surface modifications and the combinations of surface modification protocols affect the interaction of implants with tissues. Therefore, this study evaluated the optimization of multiple surface modifications by examining the effects of these treatments on Ti specimen in vitro. Alkaline phosphatase (ALP) is a membrane-bound enzyme in osteoblasts. ALP enhances osteogenesis and mineralization; thus, this enzyme is an early marker for osteoprogenitor bone cells (D1) that can be measured via cell counting and staining. Osteocalcin (OCN) is a specific marker for the bone maturation of osteoblasts. OCN is also studied in D1 cells as a late-stage marker for cell differentiation [8,22]. Therefore, the effects of the surface modifications on cell affinity, proliferation, and ALP production were also analyzed. 2. Materials and methods 2.1. Substrate surface modifications Commercially pure Ti (c.p. Ti grade IV) was cast into 14.8 mm-diameter, 2 mm-thick circular disks. The substrates were successively smoothed with 400-, 800-, and 1200-grit sandpaper. The samples were then ultrasonically washed for 30 min in acetone. The same procedure was repeated with isopropanol and again with deionized water. Finally, the disks were dried at 40 °C in an incubator set. The samples comprised the control group, labeled as machined surface (M). Further modifications involved additional sample groups, namely, SLA, SAOH, and SMART. In SLA treatment, the samples were sandblasted for 30 s using an air compressor with 2 kg/cm2 to 3 kg/cm2 of powder (with Al2O3 particles with a mean size of 200 μm blasted over a distance of ~ 75 mm). Ultrasonic cleaning was performed to dry the disks, after which they were immersed in HCl (37%, Panreac, Barcelona, Spain), H2SO4 (95% to 98%, Panreac, Barcelona, Spain), and deionized water at 1:1:100. Subsequently, temperature etching was performed at 100 °C for 30 min. SAOH disks were processed in sulfuric acid (0.1 M, H2SO4) and hydrogen peroxide (8.8 M, H2O2) for ~20 min. The disks were successively heat-treated at 100 °C for 30 min [19]. SMART disks were immersed in NaOH (5.0 M) and etched at 100 °C for 1 h. The disks were cleaned with deionized water, dried, and immersed for 30 min in hydrochloric acid (0.1 M, HCl) for etching. The samples were cleaned, dried, and heat-treated again at 100 °C for 1 h in an oven. For the ECH treatment, the samples were processed for potentiostatic or galvanostatic anodization of Ti in a neutral solution with 0.1 M sulfate at a constant direct current density of ~200 A/m2. Two Pt plates were used as cathodes on both sides of Ti anode. The samples were electrochemically prepared. Surface oxides increased to the anodic forming voltage of 125 V, which is essential for surface chemical modification. The anodization resulted in a thickening of Ti oxide layer to several micrometers. After cleaning and drying, the samples were also heat-treated at 100 °C for 1 h in an
oven. All samples were cleaned and then autoclaved (121 °C/1.2 atm) for sterilization [4,19]. 2.2. Substrate surface analysis The surface wettability of the substrates was tested in triplicate using a contact angle meter (CAM-100, Creating Nano Technologies, Inc., Taiwan). The central line average surface roughness of Ra was measured using a roughness tester (SJ-301 Mitutoyo, Ltd., Japan). The topographies of different groups were analyzed under a scanning electron microscopy (SEM) system (Hitachi S-3000N, Hitachi, Tokyo, Japan). The samples were sequentially dehydrated in graded ethanol before they were coated with gold for SEM analysis. The substrate surface structures, cell morphology after cell attachment, and repopulation were examined. 2.3. Short-term cell attachment and long-term cell proliferation tests The bone marrow cells from mesenchymal stem cells (osteoprogenitor cells, D1) that were cloned from BALB/c mice were purchased from the American Type Culture Collection. The D1 cells were maintained in Dulbecco's modified Eagle's medium. The cells were supplemented with 10% fetal bovine serum in 37 °C incubators with 5% CO2. The cells were used before the eighth passage. The substrate disks were placed into 48-well plates, after which 50 μL of 5 × 103 cells was dripped onto the substrates for culturing. The cells were incubated for 1 h, 1 day, and 4 days. At different periods, the substrates were washed carefully with phosphate buffered saline (PBS) and were fixed with glutaraldehyde. The substrates were then gold-plated and screened using SEM to determine the interval required for cell attachment and proliferation. The substrates were placed in 48-well plates, after which 50 μL of 1 × 105 D1 cells was again dripped onto the disks for incubation for 4 h to allow cell attachment. Additional media were applied to the plates. The cells were incubated for 1, 4, 7, 10, and 14 days and were washed twice with PBS at each time point. Up to 500 μL of a medium with alamarBlue® solution from an alamarBlue® cell viability assay kit (AbD Serotec, US) was added into each well. The cells were then incubated for 4 h at 37 °C. The reaction medium was spectrophotometrically measured at 570 and 600 nm using an enzyme-linked immunosorbent assay (ELISA) microplate reader (UVM-340; ASYS Hitech GmbH, Eugendorf, Austria). The cell numbers were determined from a plot of absorbance (OD values) versus the respective D1 cells after adjustment using the alamarBlue® assay. Each experiment was performed five times (n = 5). 2.4. ALP quantification and staining ALP production on the surfaces of different sample groups was determined using p-nitrophenyl phosphate (pNPP) tablets (Sigma, USA). The pNPP and Tris-buffered saline tablets were placed in 20 mL deionized water and mixed. Testing was performed simultaneously with the same intervals as in the cell proliferation tests. At the end of each interval, Ti substrates were washed twice with PBS, after which 500 μL of the prepared solution was added into each well. This solution was subsequently incubated for 30 min. ALP activity was determined through absorbance measurements using ELISA reader at 405 nm. ALP staining was performed using tartrate-resistant acid phosphatase and ALP double-stain kit (Takara Bio Inc., Shiga, Japan) in accordance with the manufacturer's instructions. As in the previous experiment, 50 μL of 1 × 105 D1 cells was dripped onto the disks during incubation and cell attachment for 4 h and medium was subsequently added. The cells were incubated for 14 days, after which the cells were washed twice with PBS and fixed with citrate buffer (pH 5.4) containing 60% acetone and 10% methanol. After the surface cells were fixed and washed with distilled water, the substrate solution was added to
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
the samples and incubated at 37 °C. Subsequently, the test samples with ALP staining were washed thrice with distilled water and examined under a light microscope.
307
(Tukey's Honestly Significant Difference test) were employed to analyze significant differences. 3. Results and discussion
2.5. Alizarin Red stain (ARS)
3.1. Surface morphology and wettability
5
Up to 50 μL of 1 × 10 D1 cells was dripped onto the disks during incubation for 4 h for cell attachment. Additional media were added to incubate the cells. After 21 days, the cells were washed twice with PBS and fixed with 10% paraformaldehyde for 5 min to 10 min. The substrates were washed twice with deionized water and stained with 2% ARS (pH 4.1) for 15 min to stain the cells. The dye was discarded, and the substrates were washed twice with deionized water. The images of these substrates were captured using a digital camera and presented with imaging software. 2.6. ALP and OCN mRNA quantification For cell culture and RNA extraction, the substrate disks were placed in 48-well plates. Up to 50 μL of 1 × 105 D1 cells was dripped onto the disks during incubation and cell attachment for 4 h. Additional media were integrated to incubate the cells. After 1, 4, 7, 10, 14, and 21 days, the cells were washed twice with PBS. Lysis reagent (~ 1 mL) was added to each well. After thorough mixing and complete lysis of cells, the lysis reagent containing cell debris was added to the centrifuge tips, and ~0.2 mL of chloroform was added to each tip. After clear separation, the tips were centrifuged at 12,000 rpm for 15 min. The supernatant was discarded and isopropanol was added. The samples were left to stand for 10 min and subsequently centrifuged at 12,000 rpm for 10 min. The supernatant isopropanol was discarded, and 0.5 mL of 75% RNase-free EtOH was added for washing. The samples were recentrifuged at 12,000 rpm for 10 min. The EtOH was discarded. The RNA was left to stand for 2 min to allow the EtOH to evaporate. Finally, 20 μL of RNase-free water was added to dissolve the RNA. The complementary DNA (cDNA) Reverse Transcription Kit applied in the reverse transcription polymerase chain reaction (RT-PCR) was purchased from Applied Biosystems (USA). In a 0.2 mL centrifuge tip, 4.2 μL of nuclease-free water, 2 μL of 10 × reverse transcriptase (RT) buffer, 0.8 μL of 25× deoxyribonucleotide triphosphate, 2 μL of 10× RT random primers, 10 μL of total RNA, and 1 μL of RT were added into each sample. Each tip was then placed into a PCR machine. After completing the PCR cycles, 80 μL of RNase-free water was added into each sample. The samples were then stored at −80 °C. The RT-PCR cocktail was first prepared by mixing nuclease-free water, SYBR green, forward primer, and reverse primer. The PCR cocktail was distributed to each well, to which cDNA was then added. Each cDNA sample was run in triplicate to correct pipetting errors. The primer sequences used are shown in Table 1. All RT-PCR experiments were performed in triplicate (n = 3). 2.7. Statistical analysis Statistical analysis was performed through JMP 9.0 software (SAS Institute Inc., Cary, NC, US), and differences with p b 0.05 were considered statistically significant. A one-way ANOVA and a comparison procedure Table 1 Primers designed for real-time polymerase chain reaction.a Genes
Primer sequence
GAPDH forward GAPDH reverse ALP forward ALP reverse OCN forward OCN reverse
5′-CTG GAG AAA CCT GCC AAG TA-3′ 5′-TGT TGC TGT AGC CGT ATT CA-3′ 5′-AAC CCA GAC ACA AGC ATT CC-3′ 5′-GTC AGT CAG GTT GTT CCG ATT CAA-3′ 5′-GAG GGC AAT AAG GTA GTG AAC A-3′ 5′-AAG CCA TAC TGG TCT GAT AGC TCG-3′
a
Mission Biotech, Taipei.
In evaluating the properties of implant surface, wettability measurement is a typical strategy for assessing the hydrophilicity of material surfaces. The contact angle of the liquid on the substrate surface shows the wettability of such surface. Researchers have argued that surface roughness and wettability properties are related to the physical and chemical properties of the material surface. These factors affect the adhesion, proliferation, and differentiation of osteoblasts, and a higher surface energy or a more hydrophilic surface can enhance the osteoblast viability and differentiation [18,23]. The contact angle of the substrate samples, in which SLA was found to have the greatest contact angle value of above 90°, is shown in Fig. 1a. The contact angle confirms the hydrophobicity of SLA substrate samples. The surface-modified M, ECH, and SMART substrates were hydrophilic because their contact angles were between 70° and 80°. SAOH substrate had the smallest contact angle at around 50°. Compared with other groups, SAOH substrate was more hydrophilic because of the surface treatment of hydrogen peroxide and the heat treatment [24,25]. According to the Wenzel (1936) relation equation, cosθw = r cosθγ (where θw is the Wenzel contact angle, r is the roughness factor defined as actual surface/geometric surface, and θγ is the Young's contact angle); the wetting angle generally complies with the following relationship [26]: increased roughness decreases the contact angle if θγ b 90° and increases the contact angle if θγ N 90°. In our study, M had the lowest roughness (Fig. 1b). Although the group of M, ECH, and SLA substrates and that of SAOH, SMART, and SLA substrates showed proportional relationships between roughness and hydrophobicity (Fig. 2), the contact angles of M and ECH substrates were insensitive when Ti surfaces were less rough. SLA, SAOH, and SMART substrates with moderate roughness (Ra = 1 μm to 2 μm) are highly sensitive to contact angles and roughness. These groups have similar roughness (p N 0.05), but their hydrophilicities varied because of different surface treatments [27,28]. 3.2. Short-term cell attachment A good implant should deliver the desired interactions between the implant surface and the living tissues. Hence, the suitability for rapid cell attachment and cell proliferation of the sample substrates should be analyzed prior to testing for cell mineralization. In this work, standard cell attachment was first established for cell culturing to determine the accurate biological performance in terms of cell growth, activity, and mineralization in vitro. Fig. 3a shows the cells cultured on different samples after 1 h. The cell morphologies on SLA and SMART samples showed the beginning of filopodial extension. The filopodia were thin, actin-rich, plasma-membrane protrusions that act as antennae for cells to probe their environment [17,29]. SLA and SMART substrates had rougher surfaces compared with the other groups (Fig. 2). By contrast, the cells on the other three groups were spherical and did not extend the filopodia. Thus, these cells were not yet capable of migration and proliferation. This result indicates that the roughened surfaces treated with SLA and SMART at the micrometer scale help promote cell attachment. After 1 day of culture, the cells attached better onto the surfaces, thereby extending the trend from 1 h of culture. The cell morphologies in SLA, SAOH, and SMART groups were similar, and all their filopodia were extended outward. However, M and ECH groups still show spindle-shaped cells without obvious filopodial extension. The cell morphologies shown in Fig. 3c are from the substrates after 4 days of incubation. All groups demonstrated obvious cell growth and similar morphologies compared with the 1 day cultures. Again, SLA-, SAOH-, and SMART-modified substrates with higher roughness were almost entirely covered with a cell
308
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 1. a. Surface conditions and contact angle (mean ± standard deviation; unit: degree) under a light microscope. b. Surface morphologies under SEM and average roughness (mean ± standard deviation; unit: μm) as shown in different sample groups.
monolayer. M and ECH groups showed overlapping cells. The results prove that the migration ability of the progenitor bone cells D1 was the most prevalent in SLA, SAOH, and SMART substrates. The phenomena in which the samples were covered with a cell monolayer indicated that these cells adhered and spread perfectly on the culture specimen even at a low density, that is, a flattened morphology. Subsequent to the disorganization of a confluent cell monolayer and immediately after the cells had detached from one another, the major factor that might affect the cell proliferation was not limited to the surface conditions of specimens. Therefore, the results of overlapping cells prove the earlier stage for the migration of the progenitor bone cells with each other. These results indicate that although a hydrophilic surface
promotes cell attachment shortly, this surface also suppresses cell proliferation; this finding has been confirmed by the results of other studies [27,28]. The ECM also influences cytokine and growth factor production [25,30]. Considering the competitions for roughness and hydrophilic in concentrated cell populations in the present study, the roughness of the surface more largely and directly affects implant and cell biological interactions than wettability. 3.3. Cell proliferation after different culture time periods Cytoskeleton plays an important role in cell morphology, adhesion, growth, and signaling. The network of actin filaments in the cytoskeletal
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 2. The relative tendency of roughness and water contact angle exhibited by different sample groups.
structure is one of the most crucial contributors to cell morphology [31]. Therefore, the differences in cell morphology and proliferation confirm active cell cytochemical analysis. In this study, cell proliferation was
309
determined through an alamarBlue® assay based on OD values (Fig. 4). Similar to the aforementioned results, the D1 cell proliferation on the groups of ECH and SAOH substrates showed the least values and was insignificantly different (p N 0.05) within 1 day culture. By contrast, SLA group exhibited the greatest metabolic activity during D1 cell incubation, except for the value on 1 day culture. The number of cells peaked from day 7 to day 10 in all groups and decreased thereafter (p b 0.05). This phenomenon was due to the mineralization of D1 progenitor bone cells. The number of cells and their metabolic activity were inconsistent. The D1 cell on SMART substrate showed the highest cell proliferation within 1 day culture and was the first to decrease in the metabolic activity on day 10, which confirms that SMART surface modification encourages advanced D1 cell differentiation. SMART group was additionally treated with alkali. − Thus, the extra hydroxyl groups (\OH+ 2 , O , or both) possibly acted as functional groups. Ti implant surface may bind with peptides as targets or adsorb Ca2+ ions onto their surface. These adsorbed ions augmented the surface with amide peptide bonds. The hydrophilic properties of the surface significantly increased and interacted with cellular integrins, which are responsible for cellular responses [12,32]. Thus, the D1 cells cultured on SMART substrates have a higher probability to generate adequate anchorage upon differentiation [33].
Fig. 3. a. Micrographs of D1 cells cultured on different sample surface conditions after 1 h. b. Morphologies of D1 cells cultured on different sample surface conditions after 1 day of incubation. c. Images of D1 cells cultured on different sample surface conditions after 4 days of incubation.
310
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 3 (continued).
3.4. ALP quantification and staining In addition to proliferation analysis, the cells need to be subjected to mineralization analysis when studying different surface modification
Fig. 4. Cell metabolic activities of D1 cells seeded on different Ti surfaces after 1, 4, 7, 10, and 14 days of incubation (n = 5). The character symbols (a, b, and c) indicating each testing group within the same cell culture time are shown to be insignificantly different (p N 0.05).
methods. ALP activity is an early marker for osteogenesis, and its expression increases in the beginning of progenitor cell differentiation and throughout ECM maturation [34]. Therefore, the differences in cell morphology and the degrees of ECM maturation are also evident in ALP cytochemical analysis. The quantification of the total ALP proteins can help determine the relationship between cell activation and differentiation on the surface of different sample groups. These results are further corroborated by ALP colorimetric assay to cell proliferation (Fig. 5a). ALP activity was normalized to the total active cells and reached significantly higher levels within 14 days of ECH D1 cell culture compared with the other groups. Higher ALP expression is not caused by increased D1 cell proliferation and migration but rather by the combinations of surface condition. Such combinations further enhance D1 cell differentiation through the regulation of the recognition motif for integrins during cell adhesion and signaling [35]. The quantification of ALP per cell shows that the maximum value was reached after 14 days of per cell incubation. Thus, the decreased metabolic activity after day 10 (Fig. 4) shows that D1 cell differentiation was evident on day 14. According to the measurement of cell proliferation, the metabolic activity on the ECH substrate was incomparable with other groups; however, ALP secretion had the following increasing order: M b SAOH b SLA b SMART b ECH. Therefore, ALP expression was further enhanced by D1 cell differentiation, which was achieved through the ECH surface modification for the early-stage mineralization. ECH is followed by SMART and so on.
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
311
Fig. 6. ARS-stained substrates with D1 cells seeded on different Ti surface conditions after 1, 4, and 21 days of incubation.
Fig. 5. a. ALP activity of D1 cells seeded on different Ti surface conditions after 1, 4, 7, 10, and 14 days of incubation (n = 5). b. Light images of ALP-stained substrates (b) of D1 cells seeded on different Ti surface conditions after 1, 4, and 14 days of incubation.
adhering onto the surface and resulted in staining deficiency. ALP staining was the most intense on day 14 of incubation than other groups, which is consistent with the previous findings involving ALP quantification. Additionally, staining was the most intense in the ECH group, which is consistent with the previous discussions on the early mineralization of D1 cells through ALP. ARS staining of Ca matrix (Fig. 6) showed that the OCN was progressively stained with antibodies. The red staining increased with increasing culture duration. Consequently, the D1 cells in each group produced OCN as the late-stage marker for the differentiation. The OCN expression in the D1 cells in all groups increased, except in M group, and the cells were uniformly stained after 21 days of incubation. The most intense OCN staining was close to the center of the stained circle on day 21 of cell culture, particularly in the ECH group. 3.5. mRNA quantification for different periods
The ECH surface modifications include a neutral solution with sulfate anodization, which induces a thicker surface of TiO2 oxidation layers [36]. The following phosphate and phosphate reagent conjugations on the charge surface of TiO2 are suitable biological conditions for cell activity [3,30]. According to studies, the ECH and SMART surface modifications produce high ion adsorption rates compared with the other treatments, which indicates that they greatly promote cell differentiation [3,30,33,37]. Meanwhile, ALP staining (Fig. 5b) revealed the lack of roughness in M group, which decreased the likelihood of cells
ALP mRNA quantification revealed that the mRNA expression peaked on day 7, which suggests that D1 cells started differentiation during this period (Fig. 7a). The ECH group showed the highest ALP mRNA expression, followed by SMART, SLA, SAOH, and M groups. Therefore, the ECH surface modification provides the best conditions for the D1 cell early-stage ALP mineralization. Fig. 7b shows the OCN mRNA expression, and Table 2 presents the statistical analyses of ALP and OCN. The OCN mRNA levels in SMART group demonstrated an unexpected increase on day 7. The cross analysis in Fig. 4 shows that
312
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313 Table 2 One-way ANOVA of the mRNA levels of alkaline phosphatase (ALP) and osteocalcin (OCN) in progenitor D1 cells from the ECH, SLA, SAOH, and SMART groups compared with the same cell line in the M group on day 1 of culture. *: p b 0.05; **: p b 0.01. Cell culture days
Groups
p value of ALP
p value of OCN
1 day
M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART
Baseline 0.0104* 0.0084** 0.1942 0.3353 0.1799 0.1520 0.0673 0.3187 0.0983 0.0646 0.0022** 0.0002** 0.0014** 0.0004** 0.3961 0.1185 0.0011** 0.2991 0.2231 0.0386* 0.3635 0.0006** 0.2933 0.0326* 0.1327 0.3520 0.0119* 0.2756 0.0309*
Baseline 0.2714 0.1336 0.1193 0.1513 0.3887 0.4394 0.4076 0.2525 0.1809 0.4269 0.2372 0.1011 0.1243 0.0023** 0.1875 0.0031** 0.0216* 0.0013** 0.0014** 0.0632 0.0069** 0.0030** 0.0051** 0.0010** 0.0052** 0.0002** 0.0006** 0.0006** 0.0002**
4 days
7 days
10 days
14 days
21 days
expression levels per cell. Overall, the ECH surface modification resulted in better D1 cell attachment, proliferation, and differentiation. Fig. 7. RT-PCR mRNA analysis of (a) ALP, which indicates earlier mineralization of D1 cells (n = 3); (b) OCN indicating the late differentiation of the progenitor cell at different culture times in all groups (n = 3).
SMART group exhibited decreased metabolic activity after day 7. This result is consistent with previous results, which confirms that SMART surface modifications promote late-stage differentiation. Hara et al. (2012) [38] proved that the implants with rough surfaces (SLA in this study, which had the highest roughness) under the periosteum promote higher ALP and OCN gene expressions in rats. Yang et al. (2008) [39] state that the bioactivity of Ti and its alloy is enhanced by H2O2/HCl and heat treatment, as exhibited by SAOH in this study, which generated the smallest contact angle and the highest hydrophilicity. These studies confirm that surface treatment determines the ultimate biochemical reactions and mechanisms in cells. ALP activity is an early-stage marker of progenitor D1 cell differentiation to osteoblast cells. OCN mRNA is a late-stage differentiation marker, and its relative expression levels signify bone matrix maturation and the mineralization of progenitor bone cells cultured on substrate surfaces. Therefore, ALP and OCN quantification can be used to assess the initial performance of an implant and the biological reaction of tissues. SLA and SMART surface modifications, with their rougher surfaces, provided better conditions for cell adhesion that can lead to higher cell proliferation. The cell proliferation on SMART substrates decreased as early as day 10, which suggests earlier D1 cell differentiation. The greatest ALP activity was observed after 14 days of cell culture when ALP quantification and qualitative analyses were performed on each progenitor D1 cell cultured in the ECH group. The PCR analyses also revealed that the cell culture using the ECH substrate had the highest ALP and OCN mRNA
4. Conclusion The effects of each treatment on bone mineralization are determined through progenitor bone cell culturing on Ti substrates with different modifications and by studying ALP and OCN levels. Considering the competitions for roughness and hydrophilic in the early stage of D1 cell culture, the surfaces with larger roughness of SLA and SMART substrates fostered better early cell adhesive and proliferative abilities compared with the other groups. Aside from the additional cell attachment, the ECH surface conditions produced the highest ALP and OCN expression levels in D1 cells among the five testing groups. The ability of D1 cells to differentiate and mineralize depended on complex factors, faster cell adhesion, and proliferation. Our study proves that ECH and SMART had the better performance for promoting biological interactions and determining D1 cell behavior of differentiation and mineralization after 10 days of culture. Although D1 cell had the less potential for proliferate ability on the ECH treatments, the ECH and SMART might have better potential for achieving rapid osseointegration of dental implants onto bone surfaces in future study in vivo. Acknowledgment The authors are grateful to the grant from the Institute of Metal Industries Research & Development Centre for the support of this research under the contract no. 101-EC-17-A-04-04-0834. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2014.01.022.
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
References [1] M. Navarro, A. Michiardi, O. Castaño, J.A. Planell, J.R. Soc, Interface 5 (2008) 1137–1158. [2] A.B.. Novaes Jr., S.L. de Souza, R.R. de Barros, K.K.Y. Pereira, G. Iezzi, A. Piattelli, Braz. Dent. J. 21 (6) (2010) 471–481. [3] Á. Györgyey, K. Ungvári, G. Kecskeméti, J. Kopniczky, B. Hopp, A. Oszkó, I. Pelsöczi, Z. Rakonczay, K. Nagy, K. Turzó, Mater. Sci. Eng. C 33 (2013) 4251–4259. [4] L.L. Guéhennec, A. Soueidan, P. Layrolle, Y. Amouriq, Dent. Mater. 23 (2007) 844–854. [5] P.H. Chua, K.G. Neoh, E.T. Kang, W. Wang, Biomaterials 29 (2008) 1412–1421. [6] K. Kieswetter, Z. Schwartz, D.D. Dean, B.D. Boyan, Crit. Rev. Oral Biol. Med. 7 (1996) 329–345. [7] W.C. Chen, C.L. Ko, H.N. Kuo, D.J. Lin, H.Y. Wu, Y. Lo, C.W. Lou, J.H. Lin, Procedia Eng. 36 (2012) 173–178. [8] C.Y. Wang, B.H. Zhao, H.J. Ai, Y.W. Wang, Biomed. Mater. 3 (2008) 015004. [9] W.C. Chen, H.Y. Wu, H.S. Chen, Mater. Sci. Eng. C 33 (2013) 1143–1151. [10] J.D. Padmos, P. Duchesne, M. Dunbar, P. Zhang, J. Biomed. Mater. Res. Part A 95A (2010) 146–155. [11] Z. Schwartz, K. Kieswetter, D.D. Dean, B.D. Boyan, J. Periodontal Res. 32 (1997) 166–171. [12] P.M. Brett, J. Harle, V. Salih, R. Mihoc, I. Olsen, F.H. Jones, M. Tonetti, Bone 35 (2004) 124–133. [13] J.H. Park, R. Olivares-Navarrete, C.E. Wasilewski, B.D. Boyan, R. Tannenbaum, Z. Schwartz, Biomaterials 33 (2012) 5267–5277. [14] X. Hu, H. Shen, K. Shuai, E. Zhang, Y. Bai, Y. Cheng, X. Xiong, S. Wang, J. Fang, S. Wei, Appl. Surf. Sci. 257 (2011) 1813–1823. [15] J.S. Colombo, S. Satoshi, J. Okazaki, S. Crean, A.J. Sloan, R.J. Waddington, J. Dent. 40 (2012) 338–346. [16] B.S. Kim, J.S. Kim, Y.M. Park, B.Y. Choi, J. Lee, Mater. Sci. Eng. C 33 (2013) 1554–1560. [17] X. Chen, Y. Li, P.D. Hodgson, C. Wen, Mater. Sci. Eng. C 31 (2011) 1545–1552. [18] J. Lincks, B.D. Boyan, C.R. Blanchard, C.H. Lohmann, Y. Liu, D.L. Cochran, D.D. Dean, Z. Schwartz, Biomaterials 19 (1998) 2219–2232.
313
[19] X.X. Wang, S. Hayakawa, T. Tsuru, A. Osaka, J. Biomed. Mater. Res. 54 (2001) 172–178. [20] E. Nasatzky, J. Gultchin, Z. Schwartz, Refuat Hapeh Vehashinayim 20 (2003) 8–19. [21] E.W. Zhang, Y.B. Wang, K.G. Shuai, F. Gao, Y.J. Bai, Y. Cheng, X.L. Xiong, Y.F. Zheng, S.C. Wei, Biomed. Mater. 6 (2011) 025001. [22] P.T. de Oliveira, A. Nanci, Biomaterials 25 (2004) 403–413. [23] F. Rupp, L. Scheideler, N. Olshanska, M. de Wild, M. Wieland, J. Geis-Gerstorfer, J. Biomed, Mater. Res. A 76 (2006) 323–334. [24] E. Zhang, Y. Wang, F. Gao, S. Wei, Y. Zheng, Adv. Mater. Res. 79–82 (2009) 393–396. [25] K. Sano, K. Shiba, J. Am. Chem. Soc. 125 (2003) 14234–14235. [26] R.N. Wenzel, Ind. Eng. Chem. 28 (1936) 988–994. [27] G. Zhao, Z. Schwartz, M. Wieland, F. Rupp, J. Geis-Gerstorfer, D.L. Cochran, B.D. Boyan, J. Biomed. Mater. Res. A 74 (2005) 49–58. [28] W.C. Chen, C.L. Ko, Mater. Sci. Eng. C 33 (2013) 2713–2722. [29] P.K. Mattila, P. Lappalainen, Nat. Rev. Mol. Cell Biol. 9 (2008) 446–454. [30] J. Schwartz, M.J. Avaltroni, M.P. Danahy, B.M. Silverman, E.L. Hanson, J.E. Schwarzbauer, K.S. Midwood, E.S. Gawalt, Mater. Sci. Eng. C 23 (2003) 395–400. [31] Y.W. Heng, C.G. Koh, Int. J. Biochem. Cell Biol. 42 (2010) 1622–1633. [32] H.C. Kroese-Deutman, J. Van Den Dolder, P.H.M. Spauwen, J.A. Jansen, Tissue Eng. 11 (2005) 1867–1875. [33] B.H. Lee, Y.D. Kim, J.H. Shin, K.H. Lee, J. Biomed. Mater. Res. 61 (2002) 466–473. [34] S.J. Bidarra, C.C. Barrias, M.A. Barbosa, R. Soares, J. Amédée, P.L. Granja, Stem Cell Res. 7 (2011) 186–197. [35] J. Michael, L. Schönzart, I. Israel, R. Beutner, D. Scharnweber, H. Worch, U. Hempel, B. Schwenzer, Bioconjug. Chem. 20 (2009) 710–718. [36] J.W. Choi, S.J. Heo, J.Y. Koak, S.K. Kim, Y.J. Lim, S.H. Kim, J.B. Lee, J. Oral Rehabil. 33 (2006) 889–897. [37] J.M. Lee, J.I. Lee, Y.J. Lim, Appl. Surf. Sci. 256 (2010) 3086–3092. [38] T. Hara, K. Matsuoka, K. Matsuzaka, M. Yoshinari, T. Inoue, Bull. Tokyo Dent. Coll. 53 (2012) 45–50. [39] G.L. Yang, F.M. He, S.S. Zhao, X.X. Wang, S.F. Zhao, Int. J. Oral Maxillofac. Implants 23 (2008) 1020–1028.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Interaction of progenitor bone cells with different surface modifications of titanium implant Wen-Cheng Chen a,⁎, Ya-Shun Chen a, Chia-Ling Ko a,b, Yi Lin c, Tzu-Huang Kuo c, Hsien-Nan Kuo c a b c
Advanced Medical Devices and Composites Laboratory, Department of Fiber and Composite Materials, College of Engineering, Feng Chia University, Taichung 40724, Taiwan Dental Medical Devices and Materials Research Center, College of Dental Medicine, Kaohsiung Medical University, Kaohsiung 807, Taiwan Medical Device Development Division, Metal Industries Research & Development Centre, Kaohsiung 82151, Taiwan
a r t i c l e
i n f o
Article history: Received 25 May 2013 Received in revised form 2 December 2013 Accepted 8 January 2014 Available online 18 January 2014 Keywords: Titanium (Ti) Implant Surface modifications Roughness Cell mineralization
a b s t r a c t Changes in the physical and chemical properties of Ti surfaces can be attributed to cell performance, which improves surface biocompatibility. The cell proliferation, mineralization ability, and gene expression of progenitor bone cells (D1 cell) were compared on five different Ti surfaces, namely, mechanical grinding (M), electrochemical modification through potentiostatic anodization (ECH), sandblasting and acid etching (SLA), sandblasting, hydrogen peroxide treatment, and heating (SAOH), and sandblasting, alkali heating, and etching (SMART). SAOH treatment produced the most hydrophilic surface, whereas SLA produced the most hydrophobic surface. Cell activity indicated that SLA and SMART produced significantly rougher surfaces and promoted D1 cell attachment within 1 day of culturing, whereas SAOH treatment produced moderate roughness (Ra = 1.26 μm) and accelerated the D1 cell proliferation up to 7 days after culturing. The ECH surface significantly promoted alkaline phosphatase (ALP) expression and osteocalcin (OCN) secretion in the D1 cells compared with the other surface groups. The ECH and SMART-treated Ti surfaces resulted in maximum ALP and OCN expressions during the D1 cell culture. SLA, SAOH, and SMART substrate surfaces were rougher and exhibited better cell metabolic responses during the early stage of cell attachment, proliferation, and morphologic expressions within 1 day of D1 cell culture. The D1 cells cultured on the ECH and SMART substrates exhibited higher differentiation, and higher ALP and OCN expressions after 10 days of culture. Thus, the ECH and SMART treatments promote better ability of cell mineralization in vitro, which demonstrate their great potential for clinical use. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Ti implants have significant functions in orthopedics, bone reconstruction, and dentistry because of their biocompatibility, corrosion resistance, low weight, and exceptional mechanical properties [1]. Ti is widely used in manufacturing oral implants. However, if Ti surface is not treated properly, the implant material might cause biological interactions that lead to infections and gum inflammation [2,3]. The consequences are dangerous and possibly lead to increased treatment risks, costs, and medical resource consumption [4,5]. Aside from biocompatibility, implants should adapt and respond according to the biological environment after implantation [6]. Ti surface should be significantly modified to improve their physical and chemical properties for promoting faster osseointegration with greater efficiency. The modified Ti surfaces should enhance the attachment, proliferation, and differentiation of progenitor bone cells while the implant is in contact with the
⁎ Corresponding author at: Department of Fiber and Composite Materials, Feng Chia University, 100 Wenhwa Rd., Seatwen, Taichung 40724, Taiwan. Tel.: + 886 4 24517250x3413; fax: +886 4 24514625. E-mail addresses: [email protected], [email protected] (W.-C. Chen). 0928-4931/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2014.01.022
surrounding tissues to accelerate bone attachment and provide anchor strength to the mandibular and maxillary alveolar bones [7–10]. Several commercialized products have been derived using different physical surface modification methods, such as machine grinding (M) and sand/grit blasting with micrometer-rough and nanometer-rough Ti surfaces (e.g., in laser ablated method), which should be performed while fabricating a system for increased utilization [3,11]. Other important factors for Ti surfaces include increasing hydrophilicity and preventing the stress-shielding effects of stress cycles, as well as load and temperature changes [12–17]. Therefore, this study focused on improving and evaluating different plate Ti surface modifications. The methods used in this study were the application of micro-rough Ti surfaces via Al2O3 sandblasting and acid etching (SLA), heat treatment and alkali treatment (SMART), hydrogen peroxide and heat treatment (SAOH), and potentiostatic anodization in sulfate electrolytes through constant electric current supply (ECH) [4,18–21]. The products with these surface modifications are clinically used commercial products. For example, the products with roughened surfaces obtained by M and sand/grit blasting are SLA from Straumann AG, TiOblast™ and OssoSpeed™ from Astra Tech, and Anker from Alliance. The other products with roughened surfaces were obtained through a combination of SLA, heat, alkali, hydrogen peroxide, and potentiostatic anodization, such as OSSEOTITE and Nano Tite™ from Biomet 3i, whereas RBM and
306
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Bio Tite-H are from DIO SM dental implants. These commercialized products with various Ti surfaces were developed to generate roughened surfaces to guarantee stable mechanical tissue–implant interfaces and to reduce recovery time. The degree of osseointegration in the early stages of wound healing (1 month to 3 months) is critical for determining the mechanical stability of the implant. The recovery time using Ti oral implants without surface modifications is generally ~ 12 weeks after implantation. With surface modifications, the recovery time can be reduced to 6 weeks to 8 weeks, demonstrating the clinical advantage of shorter healing period [20,21]. Pure Ti implants are commercially used in current teeth reconstruction procedures, but their safety must first be considered in vitro using cells before proceeding with animal studies and clinical tests. This safety precaution helps adjust and evaluate surface modification techniques, as well as analyzing the efficacy of the techniques in vivo. Evaluating the effects of different surface modifications through in vitro tests contributes to the optimization of results. Despite the existence of various surface modifications, no specific standard has been established to determine the best method. The factors involved in various types of surface modifications and the combinations of surface modification protocols affect the interaction of implants with tissues. Therefore, this study evaluated the optimization of multiple surface modifications by examining the effects of these treatments on Ti specimen in vitro. Alkaline phosphatase (ALP) is a membrane-bound enzyme in osteoblasts. ALP enhances osteogenesis and mineralization; thus, this enzyme is an early marker for osteoprogenitor bone cells (D1) that can be measured via cell counting and staining. Osteocalcin (OCN) is a specific marker for the bone maturation of osteoblasts. OCN is also studied in D1 cells as a late-stage marker for cell differentiation [8,22]. Therefore, the effects of the surface modifications on cell affinity, proliferation, and ALP production were also analyzed. 2. Materials and methods 2.1. Substrate surface modifications Commercially pure Ti (c.p. Ti grade IV) was cast into 14.8 mm-diameter, 2 mm-thick circular disks. The substrates were successively smoothed with 400-, 800-, and 1200-grit sandpaper. The samples were then ultrasonically washed for 30 min in acetone. The same procedure was repeated with isopropanol and again with deionized water. Finally, the disks were dried at 40 °C in an incubator set. The samples comprised the control group, labeled as machined surface (M). Further modifications involved additional sample groups, namely, SLA, SAOH, and SMART. In SLA treatment, the samples were sandblasted for 30 s using an air compressor with 2 kg/cm2 to 3 kg/cm2 of powder (with Al2O3 particles with a mean size of 200 μm blasted over a distance of ~ 75 mm). Ultrasonic cleaning was performed to dry the disks, after which they were immersed in HCl (37%, Panreac, Barcelona, Spain), H2SO4 (95% to 98%, Panreac, Barcelona, Spain), and deionized water at 1:1:100. Subsequently, temperature etching was performed at 100 °C for 30 min. SAOH disks were processed in sulfuric acid (0.1 M, H2SO4) and hydrogen peroxide (8.8 M, H2O2) for ~20 min. The disks were successively heat-treated at 100 °C for 30 min [19]. SMART disks were immersed in NaOH (5.0 M) and etched at 100 °C for 1 h. The disks were cleaned with deionized water, dried, and immersed for 30 min in hydrochloric acid (0.1 M, HCl) for etching. The samples were cleaned, dried, and heat-treated again at 100 °C for 1 h in an oven. For the ECH treatment, the samples were processed for potentiostatic or galvanostatic anodization of Ti in a neutral solution with 0.1 M sulfate at a constant direct current density of ~200 A/m2. Two Pt plates were used as cathodes on both sides of Ti anode. The samples were electrochemically prepared. Surface oxides increased to the anodic forming voltage of 125 V, which is essential for surface chemical modification. The anodization resulted in a thickening of Ti oxide layer to several micrometers. After cleaning and drying, the samples were also heat-treated at 100 °C for 1 h in an
oven. All samples were cleaned and then autoclaved (121 °C/1.2 atm) for sterilization [4,19]. 2.2. Substrate surface analysis The surface wettability of the substrates was tested in triplicate using a contact angle meter (CAM-100, Creating Nano Technologies, Inc., Taiwan). The central line average surface roughness of Ra was measured using a roughness tester (SJ-301 Mitutoyo, Ltd., Japan). The topographies of different groups were analyzed under a scanning electron microscopy (SEM) system (Hitachi S-3000N, Hitachi, Tokyo, Japan). The samples were sequentially dehydrated in graded ethanol before they were coated with gold for SEM analysis. The substrate surface structures, cell morphology after cell attachment, and repopulation were examined. 2.3. Short-term cell attachment and long-term cell proliferation tests The bone marrow cells from mesenchymal stem cells (osteoprogenitor cells, D1) that were cloned from BALB/c mice were purchased from the American Type Culture Collection. The D1 cells were maintained in Dulbecco's modified Eagle's medium. The cells were supplemented with 10% fetal bovine serum in 37 °C incubators with 5% CO2. The cells were used before the eighth passage. The substrate disks were placed into 48-well plates, after which 50 μL of 5 × 103 cells was dripped onto the substrates for culturing. The cells were incubated for 1 h, 1 day, and 4 days. At different periods, the substrates were washed carefully with phosphate buffered saline (PBS) and were fixed with glutaraldehyde. The substrates were then gold-plated and screened using SEM to determine the interval required for cell attachment and proliferation. The substrates were placed in 48-well plates, after which 50 μL of 1 × 105 D1 cells was again dripped onto the disks for incubation for 4 h to allow cell attachment. Additional media were applied to the plates. The cells were incubated for 1, 4, 7, 10, and 14 days and were washed twice with PBS at each time point. Up to 500 μL of a medium with alamarBlue® solution from an alamarBlue® cell viability assay kit (AbD Serotec, US) was added into each well. The cells were then incubated for 4 h at 37 °C. The reaction medium was spectrophotometrically measured at 570 and 600 nm using an enzyme-linked immunosorbent assay (ELISA) microplate reader (UVM-340; ASYS Hitech GmbH, Eugendorf, Austria). The cell numbers were determined from a plot of absorbance (OD values) versus the respective D1 cells after adjustment using the alamarBlue® assay. Each experiment was performed five times (n = 5). 2.4. ALP quantification and staining ALP production on the surfaces of different sample groups was determined using p-nitrophenyl phosphate (pNPP) tablets (Sigma, USA). The pNPP and Tris-buffered saline tablets were placed in 20 mL deionized water and mixed. Testing was performed simultaneously with the same intervals as in the cell proliferation tests. At the end of each interval, Ti substrates were washed twice with PBS, after which 500 μL of the prepared solution was added into each well. This solution was subsequently incubated for 30 min. ALP activity was determined through absorbance measurements using ELISA reader at 405 nm. ALP staining was performed using tartrate-resistant acid phosphatase and ALP double-stain kit (Takara Bio Inc., Shiga, Japan) in accordance with the manufacturer's instructions. As in the previous experiment, 50 μL of 1 × 105 D1 cells was dripped onto the disks during incubation and cell attachment for 4 h and medium was subsequently added. The cells were incubated for 14 days, after which the cells were washed twice with PBS and fixed with citrate buffer (pH 5.4) containing 60% acetone and 10% methanol. After the surface cells were fixed and washed with distilled water, the substrate solution was added to
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
the samples and incubated at 37 °C. Subsequently, the test samples with ALP staining were washed thrice with distilled water and examined under a light microscope.
307
(Tukey's Honestly Significant Difference test) were employed to analyze significant differences. 3. Results and discussion
2.5. Alizarin Red stain (ARS)
3.1. Surface morphology and wettability
5
Up to 50 μL of 1 × 10 D1 cells was dripped onto the disks during incubation for 4 h for cell attachment. Additional media were added to incubate the cells. After 21 days, the cells were washed twice with PBS and fixed with 10% paraformaldehyde for 5 min to 10 min. The substrates were washed twice with deionized water and stained with 2% ARS (pH 4.1) for 15 min to stain the cells. The dye was discarded, and the substrates were washed twice with deionized water. The images of these substrates were captured using a digital camera and presented with imaging software. 2.6. ALP and OCN mRNA quantification For cell culture and RNA extraction, the substrate disks were placed in 48-well plates. Up to 50 μL of 1 × 105 D1 cells was dripped onto the disks during incubation and cell attachment for 4 h. Additional media were integrated to incubate the cells. After 1, 4, 7, 10, 14, and 21 days, the cells were washed twice with PBS. Lysis reagent (~ 1 mL) was added to each well. After thorough mixing and complete lysis of cells, the lysis reagent containing cell debris was added to the centrifuge tips, and ~0.2 mL of chloroform was added to each tip. After clear separation, the tips were centrifuged at 12,000 rpm for 15 min. The supernatant was discarded and isopropanol was added. The samples were left to stand for 10 min and subsequently centrifuged at 12,000 rpm for 10 min. The supernatant isopropanol was discarded, and 0.5 mL of 75% RNase-free EtOH was added for washing. The samples were recentrifuged at 12,000 rpm for 10 min. The EtOH was discarded. The RNA was left to stand for 2 min to allow the EtOH to evaporate. Finally, 20 μL of RNase-free water was added to dissolve the RNA. The complementary DNA (cDNA) Reverse Transcription Kit applied in the reverse transcription polymerase chain reaction (RT-PCR) was purchased from Applied Biosystems (USA). In a 0.2 mL centrifuge tip, 4.2 μL of nuclease-free water, 2 μL of 10 × reverse transcriptase (RT) buffer, 0.8 μL of 25× deoxyribonucleotide triphosphate, 2 μL of 10× RT random primers, 10 μL of total RNA, and 1 μL of RT were added into each sample. Each tip was then placed into a PCR machine. After completing the PCR cycles, 80 μL of RNase-free water was added into each sample. The samples were then stored at −80 °C. The RT-PCR cocktail was first prepared by mixing nuclease-free water, SYBR green, forward primer, and reverse primer. The PCR cocktail was distributed to each well, to which cDNA was then added. Each cDNA sample was run in triplicate to correct pipetting errors. The primer sequences used are shown in Table 1. All RT-PCR experiments were performed in triplicate (n = 3). 2.7. Statistical analysis Statistical analysis was performed through JMP 9.0 software (SAS Institute Inc., Cary, NC, US), and differences with p b 0.05 were considered statistically significant. A one-way ANOVA and a comparison procedure Table 1 Primers designed for real-time polymerase chain reaction.a Genes
Primer sequence
GAPDH forward GAPDH reverse ALP forward ALP reverse OCN forward OCN reverse
5′-CTG GAG AAA CCT GCC AAG TA-3′ 5′-TGT TGC TGT AGC CGT ATT CA-3′ 5′-AAC CCA GAC ACA AGC ATT CC-3′ 5′-GTC AGT CAG GTT GTT CCG ATT CAA-3′ 5′-GAG GGC AAT AAG GTA GTG AAC A-3′ 5′-AAG CCA TAC TGG TCT GAT AGC TCG-3′
a
Mission Biotech, Taipei.
In evaluating the properties of implant surface, wettability measurement is a typical strategy for assessing the hydrophilicity of material surfaces. The contact angle of the liquid on the substrate surface shows the wettability of such surface. Researchers have argued that surface roughness and wettability properties are related to the physical and chemical properties of the material surface. These factors affect the adhesion, proliferation, and differentiation of osteoblasts, and a higher surface energy or a more hydrophilic surface can enhance the osteoblast viability and differentiation [18,23]. The contact angle of the substrate samples, in which SLA was found to have the greatest contact angle value of above 90°, is shown in Fig. 1a. The contact angle confirms the hydrophobicity of SLA substrate samples. The surface-modified M, ECH, and SMART substrates were hydrophilic because their contact angles were between 70° and 80°. SAOH substrate had the smallest contact angle at around 50°. Compared with other groups, SAOH substrate was more hydrophilic because of the surface treatment of hydrogen peroxide and the heat treatment [24,25]. According to the Wenzel (1936) relation equation, cosθw = r cosθγ (where θw is the Wenzel contact angle, r is the roughness factor defined as actual surface/geometric surface, and θγ is the Young's contact angle); the wetting angle generally complies with the following relationship [26]: increased roughness decreases the contact angle if θγ b 90° and increases the contact angle if θγ N 90°. In our study, M had the lowest roughness (Fig. 1b). Although the group of M, ECH, and SLA substrates and that of SAOH, SMART, and SLA substrates showed proportional relationships between roughness and hydrophobicity (Fig. 2), the contact angles of M and ECH substrates were insensitive when Ti surfaces were less rough. SLA, SAOH, and SMART substrates with moderate roughness (Ra = 1 μm to 2 μm) are highly sensitive to contact angles and roughness. These groups have similar roughness (p N 0.05), but their hydrophilicities varied because of different surface treatments [27,28]. 3.2. Short-term cell attachment A good implant should deliver the desired interactions between the implant surface and the living tissues. Hence, the suitability for rapid cell attachment and cell proliferation of the sample substrates should be analyzed prior to testing for cell mineralization. In this work, standard cell attachment was first established for cell culturing to determine the accurate biological performance in terms of cell growth, activity, and mineralization in vitro. Fig. 3a shows the cells cultured on different samples after 1 h. The cell morphologies on SLA and SMART samples showed the beginning of filopodial extension. The filopodia were thin, actin-rich, plasma-membrane protrusions that act as antennae for cells to probe their environment [17,29]. SLA and SMART substrates had rougher surfaces compared with the other groups (Fig. 2). By contrast, the cells on the other three groups were spherical and did not extend the filopodia. Thus, these cells were not yet capable of migration and proliferation. This result indicates that the roughened surfaces treated with SLA and SMART at the micrometer scale help promote cell attachment. After 1 day of culture, the cells attached better onto the surfaces, thereby extending the trend from 1 h of culture. The cell morphologies in SLA, SAOH, and SMART groups were similar, and all their filopodia were extended outward. However, M and ECH groups still show spindle-shaped cells without obvious filopodial extension. The cell morphologies shown in Fig. 3c are from the substrates after 4 days of incubation. All groups demonstrated obvious cell growth and similar morphologies compared with the 1 day cultures. Again, SLA-, SAOH-, and SMART-modified substrates with higher roughness were almost entirely covered with a cell
308
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 1. a. Surface conditions and contact angle (mean ± standard deviation; unit: degree) under a light microscope. b. Surface morphologies under SEM and average roughness (mean ± standard deviation; unit: μm) as shown in different sample groups.
monolayer. M and ECH groups showed overlapping cells. The results prove that the migration ability of the progenitor bone cells D1 was the most prevalent in SLA, SAOH, and SMART substrates. The phenomena in which the samples were covered with a cell monolayer indicated that these cells adhered and spread perfectly on the culture specimen even at a low density, that is, a flattened morphology. Subsequent to the disorganization of a confluent cell monolayer and immediately after the cells had detached from one another, the major factor that might affect the cell proliferation was not limited to the surface conditions of specimens. Therefore, the results of overlapping cells prove the earlier stage for the migration of the progenitor bone cells with each other. These results indicate that although a hydrophilic surface
promotes cell attachment shortly, this surface also suppresses cell proliferation; this finding has been confirmed by the results of other studies [27,28]. The ECM also influences cytokine and growth factor production [25,30]. Considering the competitions for roughness and hydrophilic in concentrated cell populations in the present study, the roughness of the surface more largely and directly affects implant and cell biological interactions than wettability. 3.3. Cell proliferation after different culture time periods Cytoskeleton plays an important role in cell morphology, adhesion, growth, and signaling. The network of actin filaments in the cytoskeletal
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 2. The relative tendency of roughness and water contact angle exhibited by different sample groups.
structure is one of the most crucial contributors to cell morphology [31]. Therefore, the differences in cell morphology and proliferation confirm active cell cytochemical analysis. In this study, cell proliferation was
309
determined through an alamarBlue® assay based on OD values (Fig. 4). Similar to the aforementioned results, the D1 cell proliferation on the groups of ECH and SAOH substrates showed the least values and was insignificantly different (p N 0.05) within 1 day culture. By contrast, SLA group exhibited the greatest metabolic activity during D1 cell incubation, except for the value on 1 day culture. The number of cells peaked from day 7 to day 10 in all groups and decreased thereafter (p b 0.05). This phenomenon was due to the mineralization of D1 progenitor bone cells. The number of cells and their metabolic activity were inconsistent. The D1 cell on SMART substrate showed the highest cell proliferation within 1 day culture and was the first to decrease in the metabolic activity on day 10, which confirms that SMART surface modification encourages advanced D1 cell differentiation. SMART group was additionally treated with alkali. − Thus, the extra hydroxyl groups (\OH+ 2 , O , or both) possibly acted as functional groups. Ti implant surface may bind with peptides as targets or adsorb Ca2+ ions onto their surface. These adsorbed ions augmented the surface with amide peptide bonds. The hydrophilic properties of the surface significantly increased and interacted with cellular integrins, which are responsible for cellular responses [12,32]. Thus, the D1 cells cultured on SMART substrates have a higher probability to generate adequate anchorage upon differentiation [33].
Fig. 3. a. Micrographs of D1 cells cultured on different sample surface conditions after 1 h. b. Morphologies of D1 cells cultured on different sample surface conditions after 1 day of incubation. c. Images of D1 cells cultured on different sample surface conditions after 4 days of incubation.
310
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
Fig. 3 (continued).
3.4. ALP quantification and staining In addition to proliferation analysis, the cells need to be subjected to mineralization analysis when studying different surface modification
Fig. 4. Cell metabolic activities of D1 cells seeded on different Ti surfaces after 1, 4, 7, 10, and 14 days of incubation (n = 5). The character symbols (a, b, and c) indicating each testing group within the same cell culture time are shown to be insignificantly different (p N 0.05).
methods. ALP activity is an early marker for osteogenesis, and its expression increases in the beginning of progenitor cell differentiation and throughout ECM maturation [34]. Therefore, the differences in cell morphology and the degrees of ECM maturation are also evident in ALP cytochemical analysis. The quantification of the total ALP proteins can help determine the relationship between cell activation and differentiation on the surface of different sample groups. These results are further corroborated by ALP colorimetric assay to cell proliferation (Fig. 5a). ALP activity was normalized to the total active cells and reached significantly higher levels within 14 days of ECH D1 cell culture compared with the other groups. Higher ALP expression is not caused by increased D1 cell proliferation and migration but rather by the combinations of surface condition. Such combinations further enhance D1 cell differentiation through the regulation of the recognition motif for integrins during cell adhesion and signaling [35]. The quantification of ALP per cell shows that the maximum value was reached after 14 days of per cell incubation. Thus, the decreased metabolic activity after day 10 (Fig. 4) shows that D1 cell differentiation was evident on day 14. According to the measurement of cell proliferation, the metabolic activity on the ECH substrate was incomparable with other groups; however, ALP secretion had the following increasing order: M b SAOH b SLA b SMART b ECH. Therefore, ALP expression was further enhanced by D1 cell differentiation, which was achieved through the ECH surface modification for the early-stage mineralization. ECH is followed by SMART and so on.
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
311
Fig. 6. ARS-stained substrates with D1 cells seeded on different Ti surface conditions after 1, 4, and 21 days of incubation.
Fig. 5. a. ALP activity of D1 cells seeded on different Ti surface conditions after 1, 4, 7, 10, and 14 days of incubation (n = 5). b. Light images of ALP-stained substrates (b) of D1 cells seeded on different Ti surface conditions after 1, 4, and 14 days of incubation.
adhering onto the surface and resulted in staining deficiency. ALP staining was the most intense on day 14 of incubation than other groups, which is consistent with the previous findings involving ALP quantification. Additionally, staining was the most intense in the ECH group, which is consistent with the previous discussions on the early mineralization of D1 cells through ALP. ARS staining of Ca matrix (Fig. 6) showed that the OCN was progressively stained with antibodies. The red staining increased with increasing culture duration. Consequently, the D1 cells in each group produced OCN as the late-stage marker for the differentiation. The OCN expression in the D1 cells in all groups increased, except in M group, and the cells were uniformly stained after 21 days of incubation. The most intense OCN staining was close to the center of the stained circle on day 21 of cell culture, particularly in the ECH group. 3.5. mRNA quantification for different periods
The ECH surface modifications include a neutral solution with sulfate anodization, which induces a thicker surface of TiO2 oxidation layers [36]. The following phosphate and phosphate reagent conjugations on the charge surface of TiO2 are suitable biological conditions for cell activity [3,30]. According to studies, the ECH and SMART surface modifications produce high ion adsorption rates compared with the other treatments, which indicates that they greatly promote cell differentiation [3,30,33,37]. Meanwhile, ALP staining (Fig. 5b) revealed the lack of roughness in M group, which decreased the likelihood of cells
ALP mRNA quantification revealed that the mRNA expression peaked on day 7, which suggests that D1 cells started differentiation during this period (Fig. 7a). The ECH group showed the highest ALP mRNA expression, followed by SMART, SLA, SAOH, and M groups. Therefore, the ECH surface modification provides the best conditions for the D1 cell early-stage ALP mineralization. Fig. 7b shows the OCN mRNA expression, and Table 2 presents the statistical analyses of ALP and OCN. The OCN mRNA levels in SMART group demonstrated an unexpected increase on day 7. The cross analysis in Fig. 4 shows that
312
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313 Table 2 One-way ANOVA of the mRNA levels of alkaline phosphatase (ALP) and osteocalcin (OCN) in progenitor D1 cells from the ECH, SLA, SAOH, and SMART groups compared with the same cell line in the M group on day 1 of culture. *: p b 0.05; **: p b 0.01. Cell culture days
Groups
p value of ALP
p value of OCN
1 day
M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART M ECH SLA SAOH SMART
Baseline 0.0104* 0.0084** 0.1942 0.3353 0.1799 0.1520 0.0673 0.3187 0.0983 0.0646 0.0022** 0.0002** 0.0014** 0.0004** 0.3961 0.1185 0.0011** 0.2991 0.2231 0.0386* 0.3635 0.0006** 0.2933 0.0326* 0.1327 0.3520 0.0119* 0.2756 0.0309*
Baseline 0.2714 0.1336 0.1193 0.1513 0.3887 0.4394 0.4076 0.2525 0.1809 0.4269 0.2372 0.1011 0.1243 0.0023** 0.1875 0.0031** 0.0216* 0.0013** 0.0014** 0.0632 0.0069** 0.0030** 0.0051** 0.0010** 0.0052** 0.0002** 0.0006** 0.0006** 0.0002**
4 days
7 days
10 days
14 days
21 days
expression levels per cell. Overall, the ECH surface modification resulted in better D1 cell attachment, proliferation, and differentiation. Fig. 7. RT-PCR mRNA analysis of (a) ALP, which indicates earlier mineralization of D1 cells (n = 3); (b) OCN indicating the late differentiation of the progenitor cell at different culture times in all groups (n = 3).
SMART group exhibited decreased metabolic activity after day 7. This result is consistent with previous results, which confirms that SMART surface modifications promote late-stage differentiation. Hara et al. (2012) [38] proved that the implants with rough surfaces (SLA in this study, which had the highest roughness) under the periosteum promote higher ALP and OCN gene expressions in rats. Yang et al. (2008) [39] state that the bioactivity of Ti and its alloy is enhanced by H2O2/HCl and heat treatment, as exhibited by SAOH in this study, which generated the smallest contact angle and the highest hydrophilicity. These studies confirm that surface treatment determines the ultimate biochemical reactions and mechanisms in cells. ALP activity is an early-stage marker of progenitor D1 cell differentiation to osteoblast cells. OCN mRNA is a late-stage differentiation marker, and its relative expression levels signify bone matrix maturation and the mineralization of progenitor bone cells cultured on substrate surfaces. Therefore, ALP and OCN quantification can be used to assess the initial performance of an implant and the biological reaction of tissues. SLA and SMART surface modifications, with their rougher surfaces, provided better conditions for cell adhesion that can lead to higher cell proliferation. The cell proliferation on SMART substrates decreased as early as day 10, which suggests earlier D1 cell differentiation. The greatest ALP activity was observed after 14 days of cell culture when ALP quantification and qualitative analyses were performed on each progenitor D1 cell cultured in the ECH group. The PCR analyses also revealed that the cell culture using the ECH substrate had the highest ALP and OCN mRNA
4. Conclusion The effects of each treatment on bone mineralization are determined through progenitor bone cell culturing on Ti substrates with different modifications and by studying ALP and OCN levels. Considering the competitions for roughness and hydrophilic in the early stage of D1 cell culture, the surfaces with larger roughness of SLA and SMART substrates fostered better early cell adhesive and proliferative abilities compared with the other groups. Aside from the additional cell attachment, the ECH surface conditions produced the highest ALP and OCN expression levels in D1 cells among the five testing groups. The ability of D1 cells to differentiate and mineralize depended on complex factors, faster cell adhesion, and proliferation. Our study proves that ECH and SMART had the better performance for promoting biological interactions and determining D1 cell behavior of differentiation and mineralization after 10 days of culture. Although D1 cell had the less potential for proliferate ability on the ECH treatments, the ECH and SMART might have better potential for achieving rapid osseointegration of dental implants onto bone surfaces in future study in vivo. Acknowledgment The authors are grateful to the grant from the Institute of Metal Industries Research & Development Centre for the support of this research under the contract no. 101-EC-17-A-04-04-0834. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2014.01.022.
W.-C. Chen et al. / Materials Science and Engineering C 37 (2014) 305–313
References [1] M. Navarro, A. Michiardi, O. Castaño, J.A. Planell, J.R. Soc, Interface 5 (2008) 1137–1158. [2] A.B.. Novaes Jr., S.L. de Souza, R.R. de Barros, K.K.Y. Pereira, G. Iezzi, A. Piattelli, Braz. Dent. J. 21 (6) (2010) 471–481. [3] Á. Györgyey, K. Ungvári, G. Kecskeméti, J. Kopniczky, B. Hopp, A. Oszkó, I. Pelsöczi, Z. Rakonczay, K. Nagy, K. Turzó, Mater. Sci. Eng. C 33 (2013) 4251–4259. [4] L.L. Guéhennec, A. Soueidan, P. Layrolle, Y. Amouriq, Dent. Mater. 23 (2007) 844–854. [5] P.H. Chua, K.G. Neoh, E.T. Kang, W. Wang, Biomaterials 29 (2008) 1412–1421. [6] K. Kieswetter, Z. Schwartz, D.D. Dean, B.D. Boyan, Crit. Rev. Oral Biol. Med. 7 (1996) 329–345. [7] W.C. Chen, C.L. Ko, H.N. Kuo, D.J. Lin, H.Y. Wu, Y. Lo, C.W. Lou, J.H. Lin, Procedia Eng. 36 (2012) 173–178. [8] C.Y. Wang, B.H. Zhao, H.J. Ai, Y.W. Wang, Biomed. Mater. 3 (2008) 015004. [9] W.C. Chen, H.Y. Wu, H.S. Chen, Mater. Sci. Eng. C 33 (2013) 1143–1151. [10] J.D. Padmos, P. Duchesne, M. Dunbar, P. Zhang, J. Biomed. Mater. Res. Part A 95A (2010) 146–155. [11] Z. Schwartz, K. Kieswetter, D.D. Dean, B.D. Boyan, J. Periodontal Res. 32 (1997) 166–171. [12] P.M. Brett, J. Harle, V. Salih, R. Mihoc, I. Olsen, F.H. Jones, M. Tonetti, Bone 35 (2004) 124–133. [13] J.H. Park, R. Olivares-Navarrete, C.E. Wasilewski, B.D. Boyan, R. Tannenbaum, Z. Schwartz, Biomaterials 33 (2012) 5267–5277. [14] X. Hu, H. Shen, K. Shuai, E. Zhang, Y. Bai, Y. Cheng, X. Xiong, S. Wang, J. Fang, S. Wei, Appl. Surf. Sci. 257 (2011) 1813–1823. [15] J.S. Colombo, S. Satoshi, J. Okazaki, S. Crean, A.J. Sloan, R.J. Waddington, J. Dent. 40 (2012) 338–346. [16] B.S. Kim, J.S. Kim, Y.M. Park, B.Y. Choi, J. Lee, Mater. Sci. Eng. C 33 (2013) 1554–1560. [17] X. Chen, Y. Li, P.D. Hodgson, C. Wen, Mater. Sci. Eng. C 31 (2011) 1545–1552. [18] J. Lincks, B.D. Boyan, C.R. Blanchard, C.H. Lohmann, Y. Liu, D.L. Cochran, D.D. Dean, Z. Schwartz, Biomaterials 19 (1998) 2219–2232.
313
[19] X.X. Wang, S. Hayakawa, T. Tsuru, A. Osaka, J. Biomed. Mater. Res. 54 (2001) 172–178. [20] E. Nasatzky, J. Gultchin, Z. Schwartz, Refuat Hapeh Vehashinayim 20 (2003) 8–19. [21] E.W. Zhang, Y.B. Wang, K.G. Shuai, F. Gao, Y.J. Bai, Y. Cheng, X.L. Xiong, Y.F. Zheng, S.C. Wei, Biomed. Mater. 6 (2011) 025001. [22] P.T. de Oliveira, A. Nanci, Biomaterials 25 (2004) 403–413. [23] F. Rupp, L. Scheideler, N. Olshanska, M. de Wild, M. Wieland, J. Geis-Gerstorfer, J. Biomed, Mater. Res. A 76 (2006) 323–334. [24] E. Zhang, Y. Wang, F. Gao, S. Wei, Y. Zheng, Adv. Mater. Res. 79–82 (2009) 393–396. [25] K. Sano, K. Shiba, J. Am. Chem. Soc. 125 (2003) 14234–14235. [26] R.N. Wenzel, Ind. Eng. Chem. 28 (1936) 988–994. [27] G. Zhao, Z. Schwartz, M. Wieland, F. Rupp, J. Geis-Gerstorfer, D.L. Cochran, B.D. Boyan, J. Biomed. Mater. Res. A 74 (2005) 49–58. [28] W.C. Chen, C.L. Ko, Mater. Sci. Eng. C 33 (2013) 2713–2722. [29] P.K. Mattila, P. Lappalainen, Nat. Rev. Mol. Cell Biol. 9 (2008) 446–454. [30] J. Schwartz, M.J. Avaltroni, M.P. Danahy, B.M. Silverman, E.L. Hanson, J.E. Schwarzbauer, K.S. Midwood, E.S. Gawalt, Mater. Sci. Eng. C 23 (2003) 395–400. [31] Y.W. Heng, C.G. Koh, Int. J. Biochem. Cell Biol. 42 (2010) 1622–1633. [32] H.C. Kroese-Deutman, J. Van Den Dolder, P.H.M. Spauwen, J.A. Jansen, Tissue Eng. 11 (2005) 1867–1875. [33] B.H. Lee, Y.D. Kim, J.H. Shin, K.H. Lee, J. Biomed. Mater. Res. 61 (2002) 466–473. [34] S.J. Bidarra, C.C. Barrias, M.A. Barbosa, R. Soares, J. Amédée, P.L. Granja, Stem Cell Res. 7 (2011) 186–197. [35] J. Michael, L. Schönzart, I. Israel, R. Beutner, D. Scharnweber, H. Worch, U. Hempel, B. Schwenzer, Bioconjug. Chem. 20 (2009) 710–718. [36] J.W. Choi, S.J. Heo, J.Y. Koak, S.K. Kim, Y.J. Lim, S.H. Kim, J.B. Lee, J. Oral Rehabil. 33 (2006) 889–897. [37] J.M. Lee, J.I. Lee, Y.J. Lim, Appl. Surf. Sci. 256 (2010) 3086–3092. [38] T. Hara, K. Matsuoka, K. Matsuzaka, M. Yoshinari, T. Inoue, Bull. Tokyo Dent. Coll. 53 (2012) 45–50. [39] G.L. Yang, F.M. He, S.S. Zhao, X.X. Wang, S.F. Zhao, Int. J. Oral Maxillofac. Implants 23 (2008) 1020–1028.
