Materials Science and Engineering C 49 (2015) 623–631

Contents lists available at ScienceDirect

Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec

High porous titanium scaffolds showed higher compatibility than lower porous beta-tricalcium phosphate scaffolds for regulating human osteoblast and osteoclast differentiation Makoto Hirota a,⁎, Tohru Hayakawa b, Takaki Shima c, Akihiro Ametani c, Iwai Tohnai a a b c

Department of Oral and Maxillofacial Surgery, Yokohama City University Graduate School of Medicine, 3-9 Fuku-ura, Kanazawa-ku, Yokohama 236-004, Japan Department of Dental Engineering, Tsurumi University School of Dental Medicine, 2-1-3 Tsurumi, Tsurumi-ku, Yokohama, 230-8501, Japan Medical Device Department, HI-LEX Corporation, Inc., 1-12-28 Sakae-cho, Takaraduka 665-0845, Japan

a r t i c l e

i n f o

Article history: Received 26 May 2014 Received in revised form 30 November 2014 Accepted 4 January 2015 Available online 7 January 2015 Keywords: Scaffold Titanium Beta-tricalcium phosphate Human mesenchymal stem cells Osteoblast differentiation

a b s t r a c t We compared osteoblast and osteoclast differentiation when using beta-tricalcium phosphate (βTCP) and titanium scaffolds by investigating human mesenchymal stem cells (hMSCs) and osteoclast progenitor cell activities. hMSCs were cultured for 7, 14, and 21 days on titanium scaffolds with 60%, 73%, and 87% porosity and on βTCP scaffolds with 60% and 75% porosity. Human osteoclast progenitor cells were cultured with osteoblast for 14 and 21 days on 87% titanium and 75% βTCP scaffolds. Viable cell numbers with 60% and 73% titanium were higher than with 87% titanium and βTCP scaffolds (P b 0.05). An 87% titanium scaffold resulted in the highest osteocalcin production with calcification on day 14 (P b 0.01) in titanium scaffolds. All titanium scaffolds resulted in higher osteocalcin production on days 7 and 14 compared to βTCP scaffolds (P b 0.01). Osteoblasts cultured on 87% titanium scaffolds suppressed osteoclast differentiation on day 7 but enhanced osteoclast differentiation on day 14 compared to 75% βTCP scaffolds (P b 0.01). These findings concluded that high porosity titanium scaffolds could enhance progression of hMSC/osteoblast differentiation and regulated osteoclast differentiation cooperating with osteoblast differentiation for calcification as compared with lower porous βTCP. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Titanium and ceramics have a high affinity for bone. Based on this affinity, various clinical biomaterials have been developed. Titanium bone fixation and reconstruction plates that exploit the mechanical strength of titanium are widely applied. Dental implants made of titanium are the most widely used artificial bone substitutes, and this technology is well developed. Then, many other bone substitutes using ceramics have been developed, and in recent years, absorbent porous betatricalcium phosphate (βTCP) has become widely used as a material for bone replacement [1–3]. The porous nature of bone substitutes is essential for their osteoconductivity [4]. Even when using titanium, the surface characteristics of implants are important considerations for osteoconductivity and osteointegration [5]. In recent years, using dental implants with rough surfaces has become the norm. Rough surfaces have similar nano- or micro-porous structures and improve the affinity for bone tissue. Numerous techniques are used to manufacture the porous coatings on titanium surfaces, with porosity and pore size ranging from 35%–86% and 10–200 μm, respectively [6]. Titanium fiber scaffolds are porous materials made from pure titanium fibers with a diameter of 50 μm. ⁎ Corresponding author. E-mail address: [email protected] (M. Hirota).

http://dx.doi.org/10.1016/j.msec.2015.01.006 0928-4931/© 2015 Elsevier B.V. All rights reserved.

Research and development on bone tissue regeneration has resulted in the widespread use of titanium fiber scaffolds with a porosity of 86% and a pore size of 250 μm [7–9]. Osteoblasts proliferate and differentiate well on titanium fiber scaffolds with these porous structures [10]. The proliferation and differentiation of osteoblasts and their progenitors on porous materials are critical for bone tissue formation on a scaffold. Osteoblasts are osteogenic cells that are associated with bone formation through their production of osteoids and subsequent mineralization of the osteoid matrix. During osteoblast maturation, type I collagen is expressed as the first differentiation marker of osteoblasts, followed by alkaline phosphatase (ALP). Osteocalcin expression is induced during the final differentiation stage when calcification occurs [11]. Bone tissue is constantly replaced by osteoblasts and osteoclasts. Osteoprotegerin (OPG) inhibits osteoclast differentiation, whereas the receptor activator of NF-kappa B ligand (RANKL) enhances their differentiation [12]. High expression of tartrate-resistant acid phosphatase (TRAP) 5b and RANKL indicates the progression of osteoclast differentiation [12]. Very few studies have directly compared titanium and ceramic porosities. The differences between titanium and ceramic scaffolds could affect cell proliferation and the expression of osteoblast and osteoclast differentiation markers. Mesenchymal stem cells (MSCs) have been used in many studies of bone tissue regeneration on or around

624

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

biomaterials, such as ceramics and titanium, because their multipotent differentiation capability provides them with osteogenic potential [13–15]. Previous study revealed that block shape scaffold with low porosity could not show sufficient bone formation compared to those with higher porosity, in vivo [14]. Thus, in this study, to investigate the effects of different block materials and their porosity on osteoblast and osteoclast differentiation, we used titanium fiber and βTCP scaffolds with different porosities and compared the effects of their porosity on progression of osteoblast differentiation using human MSCs (hMSCs). In addition, we also assessed osteoclast differentiation markers' expression during coculture of osteoblasts and osteoclasts. 2. Materials and methods 2.1. Porous materials We used titanium fiber scaffolds (Hi-lex Co, Kobe, Japan) and βTCP scaffolds (Olympus Terumo Biomaterials, Tokyo, Japan) with different porosities in the form of porous blocks that measured 10 × 10 × 3 mm (Fig. 1). We used porous titanium scaffolds with porosities of 60% (Ti60), 73% (Ti73), and 87% (Ti87). The diameter of a

titanium fiber was 50 μm, and a sintered titanium fiber scaffold was processed to have an average internal pore size of 250 μm (range: Ti60; 100–150 μm, Ti73; 200–250 μm, Ti87; 300–350 μm). Porosity was determined by adjusting the titanium fiber scaffold weight per unit area for every desired porosity. We also used porous βTCP scaffolds with porosities of 60% (TCP60) and 75% (TCP75). Fine βTCP powder was produced by wet milling. A slurry mixture of CaCO3 and CaHPO 4 ·2H2 O (molar ratio of 1:2) in pure water was prepared in a pot mill for 24 h, and then dried at 80 °C. Crystals of calcium-deficient hydroxyapatite were converted to βTCP by calcination at 750 °C. This βTCP was then mixed to produce a foaming slurry, which was dried at room temperature for 1 day and at 40 °C for 1 additional day. Porous βTCP (purity of 99.9%) was obtained by sintering the material at 1050 °C. Porosity was determined by adjusting the dry weight of βTCP. The pore size distribution was 100–400 μm. A 90% porous body could not be made because the mechanical strength of βTCP is weak and its form cannot be maintained. The surface of each material was observed using a stereoscopic microscope (M80; Leica, Solms, Germany) for low magnification and a scanning electron microscopy (SEM) (VE9800; Keyence, Tokyo, Japan) for high magnification.

A

B

TCP60

TCP75

C

D

E

Ti60

Ti73

Ti87

F

G

100μm

H

100μm

J

I

100μm

100μm

100μm

Fig. 1. Porous materials used in this study. (A, F) βTCP with 60% porosity, (B, G) βTCP with 75% porosity, (C, H) titanium fiber scaffolds with 60% porosity, (D, I) titanium fiber scaffolds with 73% porosity, and (E, J) titanium fiber scaffolds with 87% porosity.

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

2.2. hMSC culture Cryopreserved normal human mesenchymal stem cells (hMSCs; Lonza AG, Basel, Switzerland) were thawed and placed in a 75 mL flask to produce 5000 cells mL−1. The cell culture medium used was a dedicated propagation culture medium (Lonza AG). Cells were cultured at 37 °C in an atmosphere of 5% CO2 until the required number of cells was obtained. Cells were detached by trypsin treatment, and then resuspended in growth medium. Cells were centrifuged at 220 ×g for 5 min and the supernatant was discarded. Cells were resuspended in a medium (Lonza AG) to induce osteoblast differentiation by hMSCs to generate 5 × 104 cells ∗ ml−1. This cell suspension was slowly dripped onto each material used (Ti60, Ti73, Ti87, TCP60, and TCP75) that had been placed in a 24-well plate. Three replicates of each sample were tested during each hMSC culture period. Plates were gently shaken using side-to-side agitation for 30 min. Then, 1 mL of growth medium was added along the sides of the wells. The medium was replaced the next day with 0.8 mL of differentiation inducing medium and the culture medium was then replaced every 3 days. The numbers of viable cells and their expressions of osteoblast differentiation markers were determined on days 7, 14, and 21 after seeding. After cell culture for 14 and 21 days, the scaffolds were observed with a stereomicroscope (M80, Leica) to detect crystal formation to evaluate calcification. 2.3. hMSC proliferation After hMSCs were placed in culture, the number of cells was determined on days 7, 14, and 21. Initially, porous materials were moved to a new 24-well plate to determine the number of cells that had adhered to the porous materials. Cell Count Reagent SF (Nacalai Tesque, Kyoto, Japan) and WST-8 [2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt] (Cell Counting Kit-8; Dojin, Kumamoto, Japan) were added to the reaction medium at a concentration of 10%. Then, 1 mL of this mixture was added to the wells and incubated at 37 °C for 3 h in an atmosphere of 5% CO 2 . The absorbance of viable cells in the reaction medium was determined using a microplate reader using measurement and control wavelengths of 450 and 595 nm, respectively. 2.4. Type I collagen, ALP, and osteocalcin production by hMSCs adhering to porous materials Type I collagen concentrations in culture supernatants were determined with a commercial kit (AC Biotechnologies, Yokohama, Japan) using the manufacturer's protocol. Briefly, 5 μL of a collagen standard solution and 45 μL of culture supernatant were added to a well and mixed by shaking for 1 min. After washing, 50 μL of HRP-labeled anti-human type I collagen antibody solution was added to each well and reacted for 60 min. Next, 50 μL of a color reaction solution (enzyme substrate solution) was added to each well and reacted at 37 °C for 15 min, after which 50 μL of stop solution was added. After mixing for 1 min, the absorbance of each well was measured using a plate reader set at a wavelength of 450 nm. Type I collagen concentration was determined from a standard curve. Alkaline phosphatase (ALP) concentrations in culture supernatants were determined using an ALP assay kit (Wako Jyunyaku, Osaka, Japan) using the manufacturer's protocol. Briefly, 20 μL of an ALP standard solution and 20 μL of culture supernatant were added to a well and mixed by shaking for 1 min. This mixture was reacted at 37 °C for 15 min to develop a color reaction. Next, 80 μL of stop solution was added and mixed for 1 min, after which the absorbance of each well was determined using a plate reader set at a wavelength of 405 nm. ALP concentration was determined from a standard curve. Osteocalcin concentrations in culture supernatants were determined using an osteocalcin kit (Takara Bio, Shiga, Japan) using the manufacturer's protocol. Briefly, 100 μL of osteocalcin standard solution

625

and 100 μL of culture supernatant were added to a well and mixed by shaking for 1 min. After washing, 100 μL of HRP-labeled anti-human osteocalcin antibody solution was added to each well and reacted for 120 min. Next, 100 μL of a color reaction solution was added to each well and reacted at 37 °C for 15 min, after which 100 μL of stop solution was added. After mixing for 1 min, the absorbance of each well was determined using a plate reader set at a wavelength of 450 nm. Osteocalcin concentration was determined from a standard curve. Osteocalcin concentration per viable cell was also determined. Osteocalcin concentrations of each sample on days 7, 14, and 21 were divided by the mean absorbance value at a wavelength of 450 nm, which was proportional to the number of cells on the day of observation. These were determined as indicators of osteocalcin concentration per hMSC for each sample. 2.5. Co-culture of human osteoblasts and osteoclast precursor cells Osteoclast differentiation was evaluated by co-culturing hMSCderived osteoblasts and osteoclast precursor cells on Ti87 and TCP75 scaffolds. We compared Ti87 with TCP75 because both scaffolds showed highest activity of osteoblast differentiation in their materials, although Ti73 has similar porosity to TCP75. Osteoblasts were obtained from hMSC culture as described above. Cells were sub-cultured once before use. Ti87 and TCP75 scaffolds were placed in the wells of a 24-well plate and 100 μL of osteoblasts adjusted to 4 × 105 mL−1 was added on top of these scaffolds. Likewise, according to the manufacturer's recommendation, 100 μL of osteoclast precursor cells (Takara Bio) at 1 × 106 mL−1 was added on top of these scaffolds, which were then agitated by gently shaking the plates from side-to-side for 30 min. α-MEM culture medium (Sigma, St. Louis, MO) was added to 1 mL of growth medium and this mixture was applied to the well walls. This plate was incubated at 37 °C in an atmosphere of 5% CO2. Culture medium was replaced every 3 days. Three samples each with Ti87 and TCP75 were cultured for 7 and 14 days. After culture, the production of osteoclast differentiation markers was determined. 2.6. TRAP, RANKL, and OPG production by human osteoblast and osteoclast precursor cells adhering to porous materials Tartrate-resistant acid phosphatase (TRAP) 5b concentrations in culture supernatants were determined using a mouse TRAP assay kit (COSMO BIO CO., LTD, Tokyo, Japan) using the manufacturer's protocol. An anti-TRAP antibody solution was added to each well and samples were shaken at 950 rpm at room temperature for 60 min. After washing, 25 μL of sample, 100 μL of TRAP standard solution, and 75 μL of 0.9% NaCl were added to each well. Next, 25 μL of a separation solution was added and samples were shaken at 950 rpm at room temperature for 60 min, after which 100 μL of color reaction solution was added to each well. This mixture was reacted at 37 °C for 2 h, after which 25 μL of stop solution was added. After mixing for 1 min, the absorbance of each well was determined using a plate reader set at a wavelength of 405 nm. TRAP5b concentration was determined from a standard curve. RANKL concentrations in culture supernatants were determined using a RANKL ELISA kit (Enzo Life Sciences, Tokyo, Japan) using the manufacturer's protocol. Briefly, 50 μL of TRANCE standard solution and 50 μL of culture supernatant were added to a well, mixed by shaking for 1 min, and then reacted at room temperature for 120 min. After washing, 100 μL of HRP-labeled anti-TRANCE polyclonal antibody solution was added to each well and reacted for 120 min. Next, 100 μL of a color reaction solution was added to each well and reacted at 37 °C for 2 h, after which 100 μL of stop solution was added. After mixing for 1 min, the absorbance of each well was determined using a plate reader set at a wavelength of 490 nm. RANKL concentration was determined from a standard curve. Osteoprotegenin (OPG) concentrations in culture supernatants were determined using an OPG ELISA kit (eBopscience, San Diego, CA) using

626

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

the manufacturer's protocol. Briefly, 50 μL of OPG standard solution and 50 μL of four times-diluted culture supernatant were added to a well, mixed by shaking for 1 min, and then reacted at room temperature for 120 min. After washing, 100 μL of HRP-labeled anti-OPG polyclonal antibody solution was added to each well and reacted at 37 °C for 120 min, after which 100 μL of stop solution was added. After mixing for 1 min, the absorbance of each well was determined using a plate reader set at a wavelength of 490 nm. OPG concentration was determined from a standard curve. 2.7. Statistical analysis The numbers of hMSCs, culture supernatant concentrations of type I collagen, ALP, and osteocalcin expression, and osteocalcin concentration per hMSC on days 7, 14, and 21 were compared among the different porous scaffolds used. TRAP, RANKL, and OPG concentrations were also compared among the different porous scaffolds used. Student's t-test was used to compare the results of two groups and multiple group comparisons were made by ANOVA followed by Bonferroni's post-hoc analysis for multiple comparisons. P b 0.05 was considered significant. 3. Results 3.1. hMSC proliferation As shown in Fig. 2, 60% and 73% porosity titanium scaffolds resulted in significantly higher hMSC numbers as compared to using βTCP scaffolds at all time points (P b 0.05). Using 87% titanium scaffolds resulted in significantly higher cell numbers as compared to using βTCP scaffolds on day 7 (P b 0.05) and on day 14 (P b 0.05). The number of viable cells on Ti60 and Ti73 scaffolds reached a plateau on day 14 of culture, whereas these numbers on Ti87 scaffolds gradually increased until day 21 of culture. At all time points, the numbers of viable cells on Ti60 and Ti73 scaffolds were significantly higher than on Ti87 scaffolds (P b 0.01 on days 7 and 14, and P b 0.05 on day 21). The numbers of viable cells on TCP60 and TCP75 scaffolds increased until day 21 of culture, although these numbers were lower than those on the titanium scaffolds throughout the culture period.

using βTCP scaffolds throughout the culture period (P b 0.01 on day 14, and P b 0.05 on days 7 and 21; Fig. 3A). Type I collagen production using titanium scaffolds significantly increased and peaked on day 14, after which it decreased by day 21 of culture. Type I collagen production when using βTCP scaffolds was low, although it gradually increased until day 21 of culture; however, these concentrations were lower than when using titanium scaffolds. On day 7, type 1 collagen production when using Ti60 and Ti73 scaffolds was significantly higher than that when using a Ti87 scaffold (P b 0.05; Fig. 3A). ALP concentrations in the supernatants of hMSCs cultured on titanium scaffolds were higher than when using βTCP scaffolds throughout the culture period. On βTCP, Ti60, and 73 scaffolds, ALP production by hMSCs/osteoblasts gradually increased, whereas that on Ti87 considerably increased until day 21 of culture. ALP production by hMSCs cultured on Ti87 was not only higher than that on TCP60 and TCP75, but was also higher than that on Ti60 and Ti73 on days 14 and 21 (P b 0.01; Fig. 3B). Thus, hMSCs cultured on a Ti87 scaffold maintained the highest osteogenic activity until day 21 of culture as compared with culture on βTCP and the other titanium scaffolds. As shown in Fig. 4A, osteocalcin production increased continuously when cultured on all scaffolds. Osteocalcin production for cells cultured on all titanium scaffolds was significantly higher than that on all βTCP scaffolds on days 7 and 14 (P b 0.01). On day 21, only osteocalcin production by cells cultured on TCP60 was significantly lower than with the other scaffolds (P b 0.05). Furthermore, osteocalcin production when using Ti87 was significantly higher than when using the other titanium scaffolds and βTCP scaffolds on day 14 (P b 0.01). Osteocalcin

A

*

25 20 15 10

3.2. Osteoblast differentiation by hMSCs

TCP75

Ti60

Ti73

**

** **

Absorbance(WST8)

Day 14

Day 21

0.8

**

TCP60

2.0 * *

1.5 1.0

ALP (units/μg protein)

**

*

Day 7

** *

*

0

B

Ti87

3.0 2.5

TCP60 TCP75 Ti60 Ti73 Ti87 2D

5

Type I collagen concentrations in the supernatants of hMSCs cultured on Ti60 and Ti73 scaffolds were significantly higher than when TCP60

**

30

TCP75 0.6

Ti60 Ti73

**

*

Ti87 0.4

2D

**

*

0.2

0.0

0.5

Day 7

Day 14

Day 21

0.0

Day 7

Day 14

Day 21

Fig. 2. Differences in viable cell numbers on different porous materials. The numbers of viable hMSCs when using Ti60 and Ti73 scaffolds were significantly higher than when using Ti87 and βTCP scaffolds on all observation days. Using Ti87 resulted in higher cell numbers than using βTCP scaffolds on day 7 and 14 (**P b 0.01; *P b 0.05).

Fig. 3. hMSC production of type I collagen (A) and ALP (B) with each scaffold and 2D culture (control). Type I collagen production by hMSCs cultured on Ti60 and Ti73 scaffolds was significantly higher than that by cells cultured on βTCP scaffolds throughout the culture period. When using TCP60, TCP75, Ti60, and Ti73 scaffolds, ALP production gradually increased, whereas that when using Ti87 considerably increased until day 21 of culture. ALP production when using Ti87 was higher than that when using both TCP60 and TCP75 and when using Ti60 and Ti73 (**P b 0.01; *P b 0.05).

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

A

TCP60

20

TCP75

Ti60

Ti73

Ti87

significantly higher on day 14 than that on TCP75 (P b 0.01), while RANKL production on Ti87 was significantly lower on day 7 than that on TCP75 (P b 0.05 l Fig. 6B). OPG production on Ti87 was significantly higher than that on TCP75 on days 7 and 14 (P b 0.01 on day 7, and P b 0.05 on day 14; Fig. 6C). The difference in OPG production was clearly evident, whereas the difference in RANKL production was minimal.

2D *

Osteocalcin(ng/ml)

**

16

**

**

12

4. Discussion 8

* *

4

0 Day 7

Day 14

Day 21

Osteocalcin in scaffold

B 14

TCP60

TCP75

Ti60

Ti73

Ti87

**

12 Osteocalcin （ng/ml）

627

**

10

** **

** **

8 6 4 2 0 Day 7

Day 14

Day 21

Osteocalcin per hMSC/osteoblast Fig. 4. hMSC production of osteocalcin and osteocalcin expression per viable cell. (A) Osteocalcin production increased continuously when using titanium scaffolds of all porosities. Osteocalcin production when using Ti87 was significantly higher than when using Ti 60, Ti73, TCP60, and TCP75 on day 14 of culture. On days 7 and 14, Ti scaffolds resulted in higher osteocalcin production than when using TCP scaffolds. On day 21, osteocalcin production when using TCP60 was lower than that with other scaffolds. (B) For osteocalcin expression per viable cell with each material used, TCP75 and Ti87 resulted in significantly higher osteocalcin expression per viable cell than did TCP60, Ti60, and Ti73 throughout the culture period (**P b 0.01; *P b 0.05).

production when using TCP75 was significantly higher than when using Ti60 at all time points (P b 0.05). As shown in Fig. 4B, osteocalcin expression per viable cell for each material used showed that TCP75 and Ti87 resulted in significantly higher values than those for TCP60, Ti60, and Ti73 throughout the culture period (P b 0.01). 3.3. Calcification in scaffolds Stereomicroscopic observations indicated crystal-like structures in an 87% titanium scaffold on day 14 (Fig, 5). In contrast, no other scaffolds (60% and 73% titanium, and 60% and 75% βTCP) showed any crystal formation on day 14. By day 21, all of the scaffolds used showed some crystal formation. These crystal-like structures were observed in the narrow micro-pore spaces in these scaffolds. On day 21, crystallike structures were more remarkable in Ti60 and Ti73 scaffolds as compared with those in a Ti87 scaffold (Fig. 5). 3.4. TRAP5b, RANKL, and OPG production in co-cultures of osteoblasts and osteoclast precursor cells TRAP5b production on Ti87 was significantly higher than that on TCP75 on day 14 (P b 0.05; Fig. 6A). RANKL production on Ti87 was

In this study, we investigated hMSC proliferation and progression of osteoblast and osteoclast differentiation when these cells were cultured on porous βTCP and titanium scaffolds. Titanium was clearly better than βTCP with regard to the numbers of osteoblasts at all culture times. Compared to βTCP scaffolds, twice as many osteoblasts were found when using Ti60 and Ti73 scaffolds on days 7 and 14 of culture. On day 21, these differences were smaller, although this may have been because cell proliferation on the Ti60 and Ti73 scaffolds had reached a plateau. βTCP and titanium both have a high affinity for MSCs and/or osteoblasts; however, it was clear that titanium had a higher affinity for cells when these two materials were compared with regard to hMSCs' proliferation. Osteocalcin production indicates cell cycle arrest during cell division [16,17]. Generally, cells that exit cell cycle do not proliferate. Thus, many hMSCs had rapidly differentiated to the final stage of osteoblasts rather than proliferating on porous βTCP scaffolds and did not remain in the cell cycle for cell proliferation. Porous βTCP had less activity for hMSC proliferation than porous titanium. Because βTCP is a soluble material, Ca++ and P+ ions may have eluted during the culture period and may have promoted osteoblast maturation and/or differentiation [18]. Despite the smaller number of viable cells as compared with using titanium, osteocalcin production when using TCP75 was the same as that when using titanium. That is, the functional activity of osteoblasts was possibly at its highest level when using TCP75. ALP production is an indicator of bone regeneration. It is thought that ALP activity will persist as long as osteoblasts remain viable, even when the stage of differentiation has progressed [13,14,16]. In this study, ALP production was high when using titanium materials, particularly Ti87, throughout the culture period. On days 14 and 21, ALP production by cells cultured on Ti87 was significantly higher than that by cells cultured on Ti60 and Ti73 as well as on TCP60 and TCP75. With βTCP scaffolds, most of the hMSCs rapidly differentiated to the final osteoblast stage while rarely producing ALP, whereas ALP was produced by cells cultured on titanium on days 14 and 21 of culture, particularly those cultured on Ti87. This indicated that most osteoblasts had not reached the final stage of differentiation during the early stage of culture on titanium. Osteoblast differentiation by hMSCs was also enhanced by titanium, but not by βTCP. Osteocalcin production, which is the final stage differentiation marker of osteoblasts, by cells cultured on TCP60 and TCP75 was significantly lower than that by cells cultured on all titanium scaffolds until day 14. In particular, osteocalcin production on Ti87 was higher than other titanium scaffolds. Moreover, Ti87 showed evidence of calcification on day 14, whereas no other scaffold showed evidence for calcification on the same day, which indicated that a higher porosity titanium scaffold enhanced osteoblast differentiation. Calcification was observed on all scaffolds on day 21. The calcification observed on Ti60 and Ti73 seemed to be greater than that on Ti87. Crystal-like structures, which reflect calcification, were mainly seen in the narrow micro-pore spaces in each scaffold. Ti60 and Ti73 had much narrow pore areas as compared to Ti87 because their porosities were less than that of Ti87. The micro-pores in Ti87 were clearly larger and more irregular than those in Ti60 and Ti73, which suggests that it was difficult for crystals to spread out into a Ti87 scaffold. That is, large, irregular micro-pores might be disadvantageous for cells to proliferate into these scaffolds. Hence, those cells seeded on Ti87 began to differentiate during an early period in culture. As a

628

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

Fig. 5. Stereomicroscopic findings for each sample after cell culture on day 14 and day 21. Stereomicroscopic images are for the following scaffolds: (A, B) TCP 60, (C, D) TCP 75, (E, F) Ti60, (G, H) Ti73, and (I, J) Ti87. (A, C E, G, I). Stereoscopic findings on day 14 showed crystal-like structures in an 87% titanium scaffold only (I), whereas no crystal-like structures were observed in the other scaffolds (A, C, E, and G, corresponding to 60% and 73% titanium and 60% and 75% βTCP, respectively). (B, D, F, H, J) On day 21, all scaffolds showed crystal-like structures. Crystal-like structures were observed in the narrow spaces in these scaffolds. On day 21, crystal-like structures were more remarkable in Ti60 (F) and Ti73 (H) scaffolds compared with a Ti87 scaffold (I). Bars = 500 μm.

consequence, early, enhanced osteocalcin production on Ti87 could be observed. A previous study also reported that a large pore size in porous titanium materials was conducive to early osteoblast differentiation by MSCs [13]. Cells cultured on TCP75 barely exhibited osteocalcin production as compared to that on the titanium scaffolds even by day 21. However,

given the small number of cells that were established on βTCP scaffolds, osteocalcin expression per viable cell was significantly higher when using TCP75 as compared with using Ti60 and Ti73 scaffolds on all observation days. This expression when using TCP75 was equivalent to that when using Ti87 throughout the culture period. These results suggested that using titanium scaffolds was advantageous for hMSC

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

A

TCP75

B

Ti87

Ti87

30

0.6

25

*

RANKL (pg/mL)

0.55

TRAP5b (U/L)

TCP75

629

0.5

0.45

*

**

20 15 10 5 0

0.4 Day 7

Day 7

Day 14

RANKL expression

TRAP5b expression

C

TCP75 300

Day 14

Ti87

**

OPG (pg/mL)

250 200 150 100

* 50 0 Day 7

Day 14

Osteoprotegerin expression Fig. 6. TRAP5b, RANKL, and OPG production in co-cultures of osteoblasts and osteoclast precursor cells. (A) TRAP5b production by cells co-cultured on Ti87 was significantly higher than for those co-cultured on TCP75 at day 14 of culture. (B) RANKL production by cells co-cultured on Ti87 was significantly higher than for those co-cultured on TCP75 at day 14 of culture, whereas RANKL production using Ti87 was significantly lower than when using TCP75 on day 7. (C) OPG production by cells co-cultured on Ti87 was significantly higher than for those co-cultured on TCP75 at days 7 and 14 of culture. The difference in OPG production was clearly evident, whereas that in RANKL production was minimal (**P b 0.01; *P b 0.05).

proliferation as compared to using βTCP scaffolds and that using higher porosity scaffolds was advantageous for a single hMSC to progress osteoblast maturation and/or differentiation stage. Regarding high porosity titanium scaffolds, our results suggest that large, irregular micro-pores should be improved for cell proliferation and with increased calcification that can spread out into the entire scaffold. Regarding higher porosity βTCP scaffolds, these could be enhanced to increase cell proliferation and differentiation by increasing the cell concentrations initially seeded into these scaffolds. Ti87 and TCP75 were used to observe osteoclast differentiation in the present study because both scaffolds showed significantly higher osteoblast activity of osteocalcin expression than that of other scaffolds, Ti60, Ti73 and TCP60. Osteoclast differentiation was significantly accelerated on titanium scaffolds. However, previous studies reported contradicting results for the progression [19] and suppression [20] of

osteoclast differentiation on Ti87. In the present study, both progression and suppression were also observed when using Ti87. However, the timing of these phenomena was different. Then, the expression amount of the differentiation markers was also different. The amount of RANKL and TRAP in Ti87, a positive marker of osteoclast differentiation, was stable through the culture period, whereas the amount of OPG in Ti87, a negative marker of osteoclast differentiation, reduced during the culture. First, osteoclast differentiation was suppressed on day 7 when using Ti87, after which crystal-like structures appeared on a Ti87 scaffold on day 14 when osteoclast differentiation had progressed when using Ti87. These findings can be reasonably explained if there is cross-talk between osteoblast and osteoclast activities. For βTCP, not crystal-like structure but βTCP itself might affect osteoclast differentiation because of its resorbability, suggesting osteoclast differentiation was progressed on TCP75 from the beginning of culture unlike on Ti87.

630

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631

Porosities of the materials that we analyzed in the present study ranged 60–87% that was close to trabecular bone porosity [21]. The cortical bone has a higher porosity, 3–10%. And, shape of materials could influence osteoblast activities. That is to say, when use low porosity size and granule shape material, response of osteoblast could be different from the present results. As only in vitro study was performed, further investigation is expected to analyse bone-scaffold interaction using lower and/or granule shape in vitro and in vivo. Comparative study using same porous titanium scaffolds reported no distinct differences on the cortical bone response among 60, 73 and 87% porosity titanium scaffolds that were implanted in rabbit tibiae at 12 weeks after implantation, while only 60% porous scaffolds showed bone formation in medullary cavity [22]. Other in vivo study using same 87% porous titanium scaffold of the present study reported that cortical bone response was seen at 3 weeks after implantation [23]. To evaluate the effect of different pore size, including in vivo effect for bone formation, implantation period should be considered in the further investigation. Ranges of both scaffold pore size were not so different. Pore size plays an important role in bone formation [6,24,25]. The minimum pore size required to regenerate mineralized bone tissue is considered to be 50–100 μm [14]. It was reported that osteogenic differentiation in vitro was not affected by pore size [6]. On the other hand, it was reported that osteogenic differentiation of hMSC in hydroxyapatite scaffolds with 200 μm pore size was higher than with 500 μm pores, but proliferation was higher in the 500 μm scaffolds [26]. In the present study, higher porous scaffolds showed relatively higher osteogenic differentiation in each material. Considering that pore space would increase with porosity, pore size as well as porosity could affect hMSC proliferation and differentiation. Pore interconnection of scaffold as well as pore size can influence the bone deposition, the blood vessel invasion and the kinetics of bone formation [6,27]. In resorbable materials, interconnection density and pore density are more important than pore size [14]. Then, as long as the scaffold is fully interconnected the influence of pore size for bone ingrowth is less, while bone ingrowth should be faster in a non-fully interconnected scaffold [14]. TCP scaffolds used in the present study were not fully interconnected unlike titanium scaffold. These differences of interconnectivity could affect cell proliferation and differentiation of hMSCs seeded on the scaffold in the present study. Not only porosity and pore size but also surface chemistry and topography of scaffolds influence behavior of bone cells [28,29]. Rougher and more complex textures of material surfaces have an advantage in terms of cell spreading compared to smoother surfaces [29]. Hydroxyapatite coatings can be prepared for surface modification to differentially regulate osteoblast and osteoclast activity [28]. In the present study, titanium scaffold had smooth surface and phosphate scaffold had relatively rough surface. Although the subject was focused on porosity, influences of scaffold material surface topography on cell behaviors should be considered in further investigations. Recently, it is proposed that in vivo mechanical loading be used to increase bone formation in the scaffold instead of pre-seeding the scaffold with cells or delivering growth factors [30]. Although no direct comparison of mechanical strength of scaffolds, the TCP scaffold using the present study can't maintain the mechanical property when its porous is over 75%. By mechanotransduction aspects, titanium scaffold could have an advantage for bone tissue formation in vivo. The scaffold shape may influence osteoblast activity to create new bone. Further investigation about the scaffold shape should be addressed, especially for in vivo study. Both titanium and βTCP have high affinities for bone and are extremely useful biological materials for bone reconstruction. Accelerated cell proliferation without differentiation and differentiation without progression due to over-accelerated proliferation are both unfavorable for bone reconstructive and regenerative treatments. It is hoped that future research will build on these findings to aid in the development of new biological materials for bone reconstruction.

5. Conclusion In our comparison of osteoblast and osteoclast activities when using titanium and βTCP scaffolds, both these scaffolds have unique properties that can be enhanced by changing the conditions for their porosity. Low porosity titanium enhanced hMSC/osteoblast proliferation. Higher porosity scaffold enhanced progression of a hMSC/osteoblast differentiation compared to lower scaffold. High porosity βTCP scaffolds enhanced progression of a hMSC/osteoblast differentiation and βTCP itself might be a stimulating factor for osteoblast to induce osteoclast differentiation, while higher porosity titanium regulated osteoclast differentiation cooperating with osteoblast differentiation for calcification. Although it was difficult to conclude their superiority or inferiority from present results, it was believed that hMSCs/osteoblast compatibility was influenced by surface chemistry as well as pore architectures, including surface topography. Within the limitation of this study, high porosity titanium scaffolds could enhance progression of hMSC/osteoblast differentiation and regulated osteoclast differentiation cooperating with osteoblast differentiation for calcification as compared with lower porous βTCP. References [1] K. Kurashina, H. Kurita, Q. Wu, A. Ohtsuka, K. Kobayashi, Ectopic osteogenesis with biphasic ceramics of hydroxyapatite and tricalcium phosphate in rabbits, Biomaterials 23 (2002) 407–412. [2] R.D. Gaasbeek, H.G. Toonen, R.J. van Heerwaarden, P. Buma, Mechanism of bone incorporation of β-TCP bone substitute in open wedge tibial osteotomy in patients, Biomaterials 26 (2005) 6713–6719. [3] I.R. Zerbo, A.L.J.J. Bronckers, G. de Lange, E.H. Burger, Localisation of osteogenic and osteoclastic cells in porous β-tricalcium phosphate particles used for human maxillary sinus floor elevation, Biomaterials 26 (2005) 1445–1451. [4] A.C. Jones, C.H. Arns, A.P. Sheppard, D.W. Hutmacher, B.K. Milthorpe, M.A. Knackstedt, Assessment of bone ingrowth into porous biomaterials using MICROCT, Biomaterials 28 (2007) 2491–2504. [5] D.D. D'Lima, S.M. Lemperle, P.C. Chen, R.E. Holmes, C.W. Colwell Jr., Bone response to implant surface morphology, J. Arthroplasty 13 (1998) 928–934. [6] V. Karageorgiou, D. Kaplan, Porosity of 3D biomaterial scaffolds and osteogenesis, Biomaterials 26 (2005) 5474–5491. [7] J.A. Jansen, A.F. von Recum, J.P. van der Waerden, K. de Groot, Soft tissue response to different types of sintered metal fiber-web materials, Biomaterials 13 (1992) 959–968. [8] J.W. Vehof, P.H. Spauwen, J.A. Jansen, Bone formation in calcium-phosphate-coated titanium mesh, Biomaterials 21 (2000) 2003–2009. [9] J.W. Vehof, J. van den Dolder, J.E. de Ruijter, P.H. Spauwen, J.A. Jansen, Bone formation in Ca–P-coated and non-coated titanium fiber mesh, J. Biomed. Mater. Res. A 64 (2003) 417–426. [10] M. Hirota, T. Hayakawa, A. Ametani, Y. Kuboki, M. Sato, I. Tohnai, Apatite coating titanium fiber web accelerates human osteoblast functional activity, Int. J. Oral Maxillofac. Implants 26 (2011) 245–250. [11] E. Boanini, P. Torricelli, M. Gazzano, R. Giardino, A. Bigi, Alendronate-hydroxyapatite nanocomposites and their interaction with osteoclasts and osteoblast-like cells, Biomaterials 29 (2008) 790–796. [12] N. Bucay, I. Sarosi, C.R. Dunstan, S. Morony, J. Tarpley, C. Capparelli, S. Scully, H.L. Tan, W. Xu, D.L. Lacey, W.J. Boyle, W.S. Simonet, Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification, Genes Dev. 12 (1998) 1260–1268. [13] H.L. Holtorf, N. Datta, J.A. Jansen, A.G. Mikos, Scaffold mesh size affects the osteoblastic differentiation of seeded marrow stromal cells cultured in a flow perfusion bioreactor, J. Biomed. Mater. Res. 74A (2005) 171–180. [14] P. Kasten, I. Beyen, P. Niemeyer, R. Luginbu, M. Bohner, W. Richter, Porosity and pore size of β-tricalcium phosphate scaffold can influence protein production and osteogenic differentiation of human mesenchymal stem cells: an in vitro and in vivo study, Acta Biomater. 4 (2008) 1904–1915. [15] L. Lin, K.L. Chow, Y. Leng, Study of hydroxyapatite osteoinductivity with an osteogenic differentiation of mesenchymal stem cells, J. Biomed. Mater. Res. 89A (2009) 326–335. [16] U. Mayr-Wohlfart, J. Fiedler, K.P. Günther, W. Puhl, S. Kessler, Proliferation and differentiation rates of a human osteoblast-like cell line (SaOS-2) in contact with different bone substitute materials, J. Biomed. Mater. Res. 57 (2001) 132–139. [17] Y.C. Shi, L. Worton, L. Esteban, P. Baldock, C. Fong, J.A. Eisman, E.M. Gardiner, Effects of continuous activation of vitamin D and Wnt response pathways on osteoblastic proliferation and differentiation, Bone 41 (2007) 87–96. [18] I.R. Zerbo, S.A. Zijderveld, A. de Boer, A.L. Bronckers, G. de Lange, C.M. ten Bruggenkate, E.H. Burger, Histomorphometry of human sinus floor augmentation using a porous beta-tricalcium phosphate: a prospective study, Clin. Oral Implants Res. 15 (2004) 724–732. [19] T. Lange, A.F. Schilling, F. Peters, F. Haag, M.M. Morlock, J.M. Rueger, M. Amling, Proinflammatory and osteoclastogenic effects of beta-tricalciumphosphate and hydroxyapatite particles on human mononuclear cells in vitro, Biomaterials 30 (2009) 5312–5318.

M. Hirota et al. / Materials Science and Engineering C 49 (2015) 623–631 [20] S. Makihira, Y. Mine, H. Nikawa, T. Shuto, E. Kosaka, M. Sugiyama, R. Hosokawa, Immobilized-OPG-Fc on a titanium surface inhibits RANKL-dependent osteoclast differentiation in vitro, J. Mater. Sci. Mater. Med. 21 (2010) 647–653. [21] T.M. Keaveny, E.F. Morgan, G.L. Niebur, O.C. Yeh, Biomechanics of trabecular bone, Annu. Rev. Biomed. Eng. 3 (2001) 307–333. [22] T. Amemiya, K. Nakaoka, Y. Hamada, T. Hayakawa, Effect of porosity of titanium web on cortical bone response, J. Hard Tissue Biol. 21 (2012) 103–108. [23] M. Hirota, T. Hayakawa, A. Ametani, Y. Monden, Y. Noishiki, I. Tohnai, Use of the molecular precursor method to facilitate thin hydroxyapatite coating of titanium fiber web scaffold and enhance bone formation: an experimental study in rat cranial bone defects, Int. J. Oral Maxillofac. Implants 25 (2010) 888–892. [24] J.P. Li, P. Habibovica, M. van den Doelc, C.E. Wilson, J.R. de Wijn, C.A. van Blitterswijk, K. de Groota, Bone ingrowth in porous titanium implants produced by 3D fiber deposition, Biomaterials 28 (2007) 2810–2820. [25] L.G. Sicchieri, G.E. Crippa, P.T. de Oliveira, M.M. Beloti, A.L. Rose, Pore size regulates cell and tissue interactions with PLGA-CaP scaffolds used for bone engineering, J. Tissue Eng. Regen. Med. 6 (2012) 155–162.

631

[26] T. Mygind, M. Stiehler, A. Baatrp, H. Li, X. Zou, A. Flyvbjerg, M. Kassem, C. Buenger, Mesenchymal stem cell ingrowth and differentiation on coralline hydroxyapatite scaffolds, Biomaterials 6 (2007) 1036–1047. [27] M. Mastrogiacomo, S. Scaglione, R. Martinetti, L. Dolcini, F. Beltrame, R. Cancedda, R. Quarto, Role of scaffold internal structure on in vivo bone formation in macroporous calcium phosphate bioceramics, Biomaterials 27 (2006) 3230–3237. [28] D.O. Costa, P.D. Prowse, T. Chrones, S.M. Slims, D.W. Hamilton, A.S. Rizkalla, S.J. Dixon, The differential regulation of osteoblast and osteoclast activity by surface topography of hydroxyapatite coatings, Biomaterials 34 (2013) 7215–7226. [29] B. Mueller, J.E. Reseland, H.J. Haugen, H. Tiainen, Cell growth on pore-graded biomimetic TiO2 bone scaffolds, J. Biomater. Appl. (2015) (in press). [30] D.P. Pioletti, Integration of mechanotransduction concepts in bone tissue engineering, Comput. Methods Biomech. Biomed. Eng. 10 (2013) 1050–1055.