Immobilization of naringin onto chitosan substrates by using ozone activation

Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

Contents lists available at ScienceDirect

Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb

Immobilization of naringin onto chitosan substrates by using ozone activation Chung Hsing Li a , Jing Wei Wang b , Ming Hua Ho b,∗ , Jia Lin Shih c , Sheng Wen Hsiao d , Doan Van Hong Thien e a

Division of Orthodontics & Pediatric Dentistry, Dental Department, Tri-Service General Hospital, Taipei, Taiwan Department of Chemical Engineering, National Taiwan University of Science and Technology, Taipei 10617, Taiwan c Division of Oral and Maxillofacial Surgeons, Dental Department, Tri-Service General Hospital, Taipei, Taiwan d R&D Center for Membrane Technology, Chung Yuan University, Chungli, Taiwan e Department of Chemical Engineering, Can Tho University, Can Tho City, Viet Nam b

a r t i c l e

i n f o

Article history: Received 31 July 2013 Received in revised form 1 November 2013 Accepted 7 November 2013 Available online 15 November 2013 Keywords: Chitosan Naringin Ozone Surface modification Osteoconduction

a b s t r a c t Ozone oxidation can easily produce peroxides containing active free radicals that can be used for the surface modification of biomaterials. This process is highly efficient and nontoxic. In this research, naringin, an HMG-CoA reductase inhibitor that can promote bone formation, was immobilized onto a chitosan film using ozone activation. First, a chitosan film was treated by ozone to produce peroxides; these peroxides were then quantified and their amount was optimized by an iodide assay. For the in vitro delivery of naringin, a chitosan–naringin substrate was immersed in phosphate-buffered saline to quantify the released amount of naringin. It was found that the immobilized naringin was slowly released over the course of two weeks, where its concentration in the medium was controlled by this delivery process. The results of cell culture showed that cell viability and early osteogenic differentiation, as measured by alkaline phosphatase expression, were promoted with the immobilized naringin on chitosan substrates. The expression of osteogenic proteins, including type-I collagen, bone siloprotein, and osteocalcin, were also enhanced. According to the results of Smad1 and Smad6 phosphorylation, immobilized naringin on ozonated chitosan substrates would be able to initiate bone morphogenetic protein-Smad signaling by activating receptor Smad and by suppressing inhibitory Smad. The results in this research demonstrated that the naringin–chitosan substrate produced by biocompatible ozone activation was highly osteoconductive without cytotoxicity. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Naringin, a bioeffective compound derived from traditional Chinese herbal medicine, can promote proliferation and alkaline phosphatase (ALPase) expression in osteoblasts [1]. Naringin induces the synthesis of bone morphogenetic protein-2 (BMP-2) [2,3], which in turn is able to stimulate bone formation [4]. With BMP-2, bone mineralized density is increased by suppressing the activity of HMG-CoA reductase [5], promoting bone regeneration [6]. However, BMPs are very expensive and easily denatured. Thus, naringin, an inducer for BMP-2 synthesis in vivo, has recently attracted interest as an economical and highly stable pharmaceutical alternative to BMP. Although naringin has the potential to accelerate bone regeneration, naringin in high concentration is cytotoxic [4,7,8], which

∗ Corresponding author. Tel.: +886 2 27301255; fax: +886 2 27301255. E-mail addresses: [email protected], [email protected] (M.H. Ho). 0927-7765/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2013.11.006

is revealed by an increase in apoptosis [9]. In addition, naringin is unstable and easily oxidized after oral administration [10]. Controlled delivery from encapsulated naringin has been often applied to decrease its cytotoxicity while maintaining its effectiveness in vivo [11]. Another possibility would be to immobilize naringin on biomaterials, which may make naringin sustainably effective in osteoinduction without serious cytotoxicity [12]. Unfortunately, few studies have explored the immobilization of naringin on substrates derived from tissue engineering, especially on biocompatible and biodegradable biopolymers. Many surface-modification processes have been reported for biopolymers. Creating highly active peroxides as surface modifications have been used previously to graft biomolecules [13–15]. On polymers’ surfaces, ozone oxidation can easily produce peroxides, such as hydroperoxides, which can generate free radicals. By introducing these highly active groups, bioactive molecules can be further immobilized onto materials’ surfaces. For example, typeI collagen (COL1) and RGD peptides were grafted onto ozonated poly-lactide scaffolds, thereby enhancing the material’s cell affinity

2

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

[16]. Compared with other methods for surface modification, ozone activation is efficient and simple. Moreover, ozone modification greatly reduces the need for nonbiocompatible organic solvents in the preparation process. Thus, ozone treatment was applied in this research to immobilize naringin onto chitosan, a biodegradable polymer that usually serves as substrate in cell culture and scaffolds in tissue engineering. 2. Materials and methods 2.1. Materials Chemicals used in this study were sourced from several reputable firms: chitosan was purchased from Sigma–Aldrich with a deacetyl degree of 85%; acetic acid, sodium hydroxide (NaOH), ethanol, 3-(4,5-dimethythiazol-2yl)-2,5-diphenyltetrazolium bromide (MTT), and phosphate-buffered saline (PBS) were also purchased from Sigma–Aldrich; DPPH used in the study was sourced from Sigma-Aldrich; 1-methyl-2-pyrrolidinone (NMP) was purchased from Tedia Company, Inc.; ethanol was acquired from Seoul Chem. Ind. For cell cultures, ␣-minimum essential medium (␣MEM), sodium ␤-glycerophosphate, ascorbic acid, and dexamethasone were purchased from Sigma–Aldrich. Fetal bovine serum (FBS), penicillin–streptomycin–amphotercin, and trypsin–EDTA solution were purchased from Gibco-BRL. Finally, reagents for ALPase, including sodium carbonate, pnitrophenylphosphate, MgCl2 , ALPase, DNA quantification kit, and silver nitrate, were purchased from Sigma–Aldrich. All solvents used in this study were of analytical quality; distilled and deionized water was used throughout this study. 2.2. Preparation of chitosan substrates Chitosan substrates were fabricated by a solvent-casting process. Chitosan powder was dissolved in an acetic acid aqueous solution (1 M) to form a 2 wt% chitosan solution. After the chitosan solution was cast on a culture dish and dried at 50 ◦ C for 2 days, the chitosan substrates were immersed in a NaOH/ethanol aqueous solution, which was prepared by dissolving 40 g NaOH powder in 1000 mL of a 50 vol% ethanol aqueous solution. The chitosan film was then washed with PBS several times and dried at room temperature to remove the remaining liquid. This solvent-casting process allowed us to obtain dense chitosan substrates. The thickness and roughness of the chitosan films used in this research were 1 ± 0.18 mm and 12 ± 2 nm, as measured by atomic force microscope (Nanoscope III, Bruker Inc., USA). 2.2.1. Ozone treatment The chitosan films were placed in a flask, through which a stream of O3 was continuously bubbled, for a certain period. A modified iodide assay was used to measure peroxides produced on the surface of ozone-treated films [16]. 2.2.2. Naringin immobilization and controlled release After ozone activation, thermal induction was employed to graft naringin onto the ozonated chitosan substrate. The ozone-activated chitosan substrate and naringin solutions (0, 0.1, 5, 10, and 25 wt% naringin in NMP) were placed into a flask saturated with nitrogen, and then placed into an isothermal oil bath at 60 ◦ C for thermal induction. The chitosan substrates with grafted naringin were subsequently washed with deionized water three times and sterilized with UV before release profiles or cell cultures were evaluated. For comparison, some chitosan substrates were also immersed into a naringin solution for 12 h without any ozone activation, where naringin was adsorbed onto chitosan surfaces directly.

Chitosan substrates with naringin immobilization were immersed in PBS at 37 ◦ C. After immersions for 2, 4, 6, 8, 12, 24, 72, and 120 h, the buffer solution was analyzed by high-performance liquid chromatography (HPLC) to measure the concentration of released naringin. HPLC was performed with a Shimadzu (Kyoto, Japan) DGU-14A system equipped with a model LC-10AT-VP liquid chromatography pump, an autoinjector, and a diode-array detector. Shimadzu software was used to calculate peak areas. Compounds were separated on a Spherisorb ODS1 column from Waters Instruments (MA, USA). The detection wavelength was 285 nm. 2.3. Cell culture and analysis UMR-106, osteoblast-like cells originally isolated from a rat osteosarcoma, were used for this study. The cells were cultured in ␣MEM, supplemented with 10% FBS and 100 U/mL penicillin–streptomycin–amphotercin, at 37 ◦ C in 5% CO2 . UMR cells suspended in the culture medium (5 × 105 cells/ml) were added to Petri dishes containing the modified or unmodified chitosan films. The culture medium was renewed every two days during cell culturing. After various periods of incubation, the dishes were rinsed with PBS buffer in preparation for the analysis of cell viability, ALPase activity, and expressions of osteogenic proteins. 2.3.1. Analysis of cell viability Cell viability was evaluated by MTT assay, in which yellow MTT was reduced to yield a purple formazan product by mitochondrial succinate dehydrogenase. Spectrophotometric measurement of MTT-formazan at 570 nm allowed the quantitation of cell viability. The substrates with cultured cells were transferred into new 24-well plates containing 1 mL of a working solution. After 6 h of incubation, the working solutions were replaced by DMSO. An ELISA plate reader (MQX200R, Bio-Tek Instrument Inc., USA) was used at a wavelength of 570 nm to determine absorbance. 2.3.2. Analysis of osteogenic differentiation Early cell differentiation was determined by ALPase activity. An ALPase buffer containing 1 mg/mL PNPP was prepared by dissolving MgCl2 (1 mM), ZnCl2 (1 mM), and glycine (0.1 M) in deionized water. After removing the culture medium, the scaffolds with UMR106 cells were washed twice with PBS, immersed into 0.4 mL of a lysis buffer for 15 min, and finally put into 1.2 mL of an ALPase buffer for 30 min. The reaction was stopped by the addition of 300 mL of 3 N NaOH. Absorbance was read at 405 nm using an ELISA plate reader. After the cells cultured to 90% confluence, total RNA was extracted using Trizol reagent, and single strand cDNA synthesis was performed by using ReverTra qPCR RT kit. Real-time quantative polymerase chain reaction (PCR) was conducted using an SYBR green kit. Tested genes’ expression levels were normalized to GAPDH levels. The PCR primer sequences were as follows: bone siloprotein (BSP), sense: 5 -AATGAAAA CGAAGAAAGCGAAG-3 , antisense: 5 -ATCATAGCCATCGTAGCCTTGT-3 , osteocalcin (OCN), sense: 5 -ATGAGAGCCCTCACACTCCTC3 , antisense: 5 -GCCGTAGAAGCG CCGATAGGC-3 , COL1, sense: antisense: 5 -TGCACTTTT5 -ACAGCCGCTTCACCTACAGC-3 , GGTTTTTGGTCAT-3 . For Smad1 and Smad6 analysis, protein extractions of lysed cells were fractionated by 10% SDS-PAGE, electroblotted onto a HybondP membrane, which was probed with antibodies to Smad1 and P-Smad1, or with antibodies to Smad6 and P-Smad6. Then, samples were developed using an ECF Western blotting kit (Amersham Pharmacia Biotech) and visualized using a Typhoon 9410 Imager (Amersham Biosciences, Piscataway, NJ).

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

peroxides conc. (gmol/cm2)*10-7

(a)

3

6

4

2

0 0

2

4 6 Flow rate (L/min)

8

10

peroxides conc. (gmol/cm2)*10-7

(b) 4

2

0 0

60

120

Time (min) Fig. 1. Peroxide concentrations on chitosan substrates. The peroxides were generated by ozone activation with different ozone flow rates (a), and with different ozone activation times (b). In (a), the ozone flow rates were varied from 0 to 8 L/min when the ozone treatment time was fixed as 60 min. In (b), the ozone treatment periods were varied from 0 to 120 min when the ozone flow rate was fixed as 4 L/min in (b).

3. Results and discussion 3.1. Optimization of the ozonation process The peroxide content of the ozone-treated chitosan samples were determined by iodide assay, and the effects of ozone flow rate and treatment time on the amount of peroxide generated are shown in Fig. 1. The outcome in Fig. 1(a) indicated that accumulated peroxides would increase when ozone flow rate was continuously increased from 0 to 4 L/min. Once the flow rate exceeded 4 L/min, the amount of peroxides generated reached a plateau. This plateau is the result of the variable crystallinity of chitosan films, which has been discussed in our previous study [16]. Based on Fig. 1(a), ozone flow rate was set to 4 L/min for subsequent experiments. Fig. 1(b) shows that the amount of peroxides generated increased with the duration of ozone treatment ranging from 0 to 60 min. The amount of peroxides did not increase any further when the treatment period exceeded 60 min. Based on Fig. 1(b), the ozonation reaction time was set to 60 min in later experiments. Fig. 1 shows that the density of generated peroxides on the chitosan substrates was approximately 10−7 g mol/cm2 . This peroxide content is higher than that resulting from ozone activation among various polymers [13,16], and is enough for grafting bioactive molecules [13,15,16]. These results support the claim that ozonation can be applied in the surface modification of chitosan substrates. The density of immobilized naringin was approximately 1.9 × 10−8 , 2.5 × 10−8 , 4 × 10−8 , and 8.1 × 10−8 g mol/cm2 on ozonated chitosan substrates when naringin concentrations used in the modification process were 0.1, 5, 10, and 25 wt%, respectively. On chitosan substrates without ozone treatment, the naringin

Fig. 2. (a) IR spectra of chitosan substrates with naringin modification, including direct naringin adsorption without ozone activation (spectrum A) and naringin immobilization with ozonation (spectrum B). (b) IR spectrum of chitosan substrates without naringin modification (spectrum C).

density was approximately 1.6 × 10−8 , 2.1 × 10−8 , 3.4 × 10−8 , and 7.5 × 10−8 g mol/cm2 when naringin concentrations used in the modification process were 0.1, 5, 10, and 25 wt%, respectively. There were no significant differences in naringin density on ozonated and non-ozonated chitosan substrates. The immobilization efficiency of naringin was defined as the ratio of naringin immobilized on chitosan substrates to naringin used in the modification process. Immobilization efficiency was 42%, 21%, 11%, and 8% when naringin concentrations used in modification process were 0.1, 5, 10, and 25 wt%, respectively. Fig. 2(a) shows that the peak at approximately 1600 cm−1 was changed in the spectrum of naringin on ozonated chitosan (spectrum B) because of the aromatic in naringin [17]. On the contrary, the results from IR spectra indicate that naringin adsorbed to chitosan surfaces was possibly washed away in the pretreatment before IR analysis, because there was no difference between spectra A and C at approximately 1600 cm−1 in Fig. 2(a) and (b). The appearance of naringin on ozonated chitosan after washing would suggest that the intermolecular forces between ozonated chitosan and naringin were significantly higher. 3.2. In vitro release of naringin To quantify the effect of ozone activation on naringin immobilization, chitosan substrates with immobilized naringin were immersed in PBS, and the concentrations of naringin released were analyzed and quantified after several immersion periods. In this analysis, naringin was immobilized onto chitosan substrates by two processes. (1) To immobilize naringin after chitosan was ozonated, which was designated by “O”. (2) To immerse chitosan substrate directly into a naringin solution without any ozonation.

4

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

Fig. 3. Release rate (a) and final steady-state concentration (b) of naringin from chitosan substrates in 120 h. The concentrations of naringin solution used in surface modification were 0.1, 5, 10 and 25 wt%. “A” indicated the modification without ozonation and naringin was physically adsorbed onto chitosan substrates. “O” indicated that chitosan substrates were treated with ozone, and then immersed into naringin solution for modification. Signicances of differences between samples with and without ozonation were indicated by star signs (*p < 0.05, **p < 0.01, Student’s t-test).

In the second immobilization process, naringin was attached onto chitosan mainly because of adsorption, which was designated by “A” in this study. The released concentration quickly increased in the first stage, from 0 to 24 h, at which point the concentration gradually reached steady state. Such a two-stage release profile is the most general case in controlled release [18]. The release rates in the first stage and steady concentration in the second stage were calculated and presented in Fig. 3(a) and (b). Fig. 3(a) shows that the release rates of naringin in samples without ozone activation would be significantly higher than those with ozonation. Since high concentrations of free naringin were toxic to cells or tissues locally, the results revealed that ozonation would minimize the cytotoxicity of naringin. The slow and sustained release of naringin from ozone-activated chitosan substrates is possibly because of the strong intermolecular forces between naringin and chitosan. The peroxides generated from ozone treatment were highly reactive, through radicalization, and thus reacted with naringin. Therefore, naringin was firmly incorporated onto the chitosan substrates, decreasing release rate. The steady concentration in the second stage in Fig. 3(b) shows that after a long release period, the final concentration of released naringin increased with naringin concentration initially applied in the modification process. However, with the same modification concentration, ozone treatment did not influence the steady-state concentration of naringin. That is, ozonation mainly reduces initial burst release in the first

Fig. 4. Normalized cell viability on chitosan substrates after cell culture for 1, 3 and 5 days. In (a), chitosan substrates were modified with narining but without ozone pretreatment. In (b), chitosan substrates were modified with naringin after ozonated. All the cell viabilities were devided by the that on chitosan without any modification which was the control group and indicated as “Chitosan”. The concentrations of naringin solution used in surface modification were 0, 0.1, 5, 10, 25 wt% in NMP, respectively. Signicances of differences between samples with and without modification were marked by black circles (• p < 0.05, •• p < 0.01, Student’s t-test). Significances of differences between indicated groups were marked by star sighs (*p < 0.05, **p < 0.01, Student’s t-test).

stage, but does not affect the steady-state concentration in the second stage of naringin delivery. 3.3. Cell viability The attached UMR cells’ viability on various chitosan substrates is shown in Fig. 4. In Fig. 4(a), naringin was directly adsorbed onto the chitosan substrates without ozone pretreatment. The results indicate that the naringin adsorbed on chitosan substrates was not effective in promoting cell viability. On the first day of culturing, chitosan with naringin adsorption resulted in lower cell viability than that of cells cultured on chitosan substrates without any modification. According to the results shown in Fig. 3, the release rate in the first stage was quite high for naringin-adsorbed chitosan substrates. It was inferred that most of the naringin was quickly released at the beginning, which resulted in cytotoxicity on the first day of culturing, especially when a high naringin concentration was applied (A-25%). Then, naringin was removed when the medium was renewed, so that the exposure to naringin was not sufficient to influence cultured cells. Thus, was almost ineffective on cell viability on the third and fifth day of culture. However, if high naringin concentration was applied, such as A-25% in Fig. 3,

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

3.4. Osteogenic differentiation Fig. 5(a) depicts the ALPase expression of UMR cells cultured on naringin-modified chitosan substrates without ozone pretreatment. In contrast, Fig. 5(b) shows ALPase expression on naringin-modified chitosan substrates that have been pretreated by ozone. In Fig. 5, the continued expression and reasonable value of the early osteoblastic markers, ALPase activity, indicates that the osteogenic phenotype was maintained [20]. In general, longer culture times result in higher ALPase activities. In Fig. 5(a), the UMR cells expressed the highest amount of ALPase when cultured on the chitosan substrate without naringin. This suggests that osteoblastic phenotypes were not promoted, but were instead suppressed by the naringin adsorbed on the chitosan substrates. Naringin would not affect bone regeneration if it was just added or coated to a biomaterial directly. Moreover, naringin would actually decrease the osteoconductivity of biomaterials. The results in Fig. 5(b) contradict those in Fig. 5(a), showing that immobilized naringin with a suitable concentration (O-5%) enhanced the expression of ALPase. Since ALPase is an important marker of early osteogenic differentiation of bone cells, Fig. 5(b) demonstrates that by ozone activation, the naringin modification would promote the osteoconduction of chitosan substrates. In addition, naringin was clearly effective only when the concentration used in modification was 5 wt%, suggesting that the amount of naringin must be controlled precisely to enhance the osteoinduction abilities of chitosan. The RT-PCR results in Fig. 6 show that the osteogenic expression of UMR cells, including BSP, COL1, and OCN, was greatly increased on ozonated naringin–chitosan substrates with naringin concentration of 5 wt% (O-5%). The expression of OCN and BSP was also promoted slightly at O-0.1% and O-10% (p < 0.05). In contrast,

Fig. 5. ALPase expressions on chitosan substrates after cell culture for 1, 3 and 5 days. In (a), chitosan substrates were modified with narining but without ozone pretreatment. In (b), chitosan substrates were modified with naringin after ozonated. The concentrations of naringin used in surfacemodification were 0, 0.1, 5, 10, 25 wt%. “Chitosan” indicated the chitosan substrate without any modification. Signicances of differences between samples with and without modification were marked by black circles (• p < 0.05, •• p < 0.01, Student’s t-test). Significances of differences between indicated groups were marked by star sighs (*p < 0.05, **p < 0.01, Student’s t-test).

Chitosan A-0.1% A-5% A-10% O-0.1% O-5% O-10%

2000

Relative mRNA level

cytotoxicity of naringin appeared, because free naringin was not completely removed by medium renewal. Fig. 4(b) shows that naringin was immobilized after the chitosan was ozonated. These results indicate that cell viability on chitosan would be significantly improved by O-5% and O-0.1%. After the first day of cell culturing, naringin did not yet exhibit a clear effect on cell viability, which confirms previous research [19]. On the third day, viability of UMR cells increased to 180% and 280% on O-0.1% and O-5% of its former value, respectively. These results and those in Fig. 3 indicate that ozone activation was able to immobilize naringin successfully; thus, greatly promote the biocompatibility of chitosan substrates. Nevertheless, if the naringin concentration used in modification exceeded 10 wt% (O-10% and O-25%), the effect of naringin on promoting chitosan’s biocompatibility was no longer significant. This is possibly because the amount of the released naringin was sufficient for it to show cytotoxicity. On the fifth day, the enhancement in cell viability due to naringin immobilization was no longer evident on all chitosan substrates, which resulted from contact inhibition because of confluent cell density. Since 5 wt% and 10 wt% seem to be the critical modification concentrations for naringin effects, the resulting released naringin concentrations were further evaluated from Fig. 3. For both O-5% and O-10%, the release rates of naringin from ozonated chitosan were approximately 0.12 ppm per hour, which was similar to the release rate at O-0.1%. However, cell viability for O-0.1% was different from the other two concentrations. Fig. 3 indicates that the steady-state concentrations for O-0.1%, O-5%, and O-10% were 0.8, 3.84, and 8.05 ppm, respectively, where this tendency basically corresponded to the resulting cell viability. Accordingly, this implied that the difference in steady-state concentration of naringin released from ozonated chitosan was the dominant factor in cell viability, and the critical steady concentration would lie from 3.84 to 8.05 ppm.

5

1500

1000

500

0 OCN

BSP

COL1

8 days after culture Fig. 6. OCN, BSP and COL1 expressions on naringin-modified chitosan substrates with and without ozone pretreatment after cell culture for 8 days. The concentrations of naringin solution used in surface modification were 0, 0.1, 5, 10, 25 wt% in NMP, respectively. “Chitosan” indicated the chitosan substrate without any modification. “A” indicated naringin was adsorbed onto chitosan without ozone pretreatment. “O” indicated chitosan was ozonated before naringin immobilization. Signicances of differences compared with A-0.1% were marked by black circles (• p < 0.05, •• p < 0.01, Student’s t-test). Significances of differences compared with chitosan were marked by star sighs (*p < 0.05, **p < 0.01, Student’s t-test).

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

absorbed naringin on chitosan was not able to enhance osteogenic differentiation, as revealed by decreases in BSP, COL1, and OCN in those samples. Moreover, as naringin concentration increased, osteogenic differentiation was further decreased; thus, A-10% < A5% < A-0.1% was revealed from the expression of BSP, COL1, and OCN. These results demonstrated that through the ozonation process proposed in this research, the immobilized naringin significantly improved the osteoconduction of chitosan, which cannot be achieved by adsorbed naringin. This was because naringin immobilization on ozone pretreated chitosan stimulates osteoblastic cells continuously and moderately. In contrast, the naringin directly adsorbed onto chitosan would be released at the very beginning, providing insufficient time to induce osteogenic differentiation.

(a) 0.6

0.5

P-Smad1 : Smad1

6

0.4

0.3

0.2

0.1

3.5. Smad phosphorylation 0.0 Chitosan A-0.1

A-5%

A-10% O-0.1% O-5%

Chitosan A-0.1

A-5%

A-10% O-0.1% O-5%

O-10%

(b) 0.5

P-Smad6 : Smad6

Phosphorylation of Smad1 and Smad6 was examined to further investigate the effects of immobilized naringin; this kind of testing has not been conducted in any previous studies for naringin. Smad1, one of the receptor Smad proteins, acts in response to BMPs [21], while Smad6 (inhibitory Smad) is structurally divergent from the other Smads and antagonizes signal transduction. The analysis of Smad1 and Smad6 phosphorylation is illustrated in Fig. 7. Fig. 7(a) shows that phosphorylation of Smad1 was increased by naringin, whether naringin was immobilized by ozone pretreatment or by direct adsorption, compared with Smad1 phosphorylation on unmodified chitosan substrates. Since Smad1 plays an important role in transducing BMP receptor molecules and in mediating BMP signaling, the results provide evidence that naringin is an effective promoter of bone formation, because naringin enhances osteogenic responses by BMP signaling. Our findings support the claim in Yu et al. that naringin would induce BMP-2 expression by the PI3K, Akt, and AP-1 pathways [22]. Fig. 7(a) also indicates that the naringin immobilized by ozonation was more influential on Smad1 activation than naringin adsorbed onto chitosan without ozone pretreatment. Among the various naringin concentrations used in modification, O-5% was most effective, and corresponded to the results for cell viability and osteogenic differentiation in Figs. 5 and 6. The experimental results of this study reveal that naringin enhanced Smad1 phosphorylation more effectively if naringin was immobilized after chitosan ozonation. The results in Fig. 7(a) might derive from the ability of naringin immobilized on ozonated chitosan to support continuous osteoblastic cell stimulation by providing high naringin concentrations but below cytotoxic levels. Fig. 7(b) shows the Smad6 phosphorylation of UMR cells. The results suggest that O-5% and O-10%, especially the former, suppressed the expression of Smad6 efficiently. Smad6 interferes with phosphorylation of receptor Smads and negatively regulates TGF␤ signaling [23]. The suppression of TGF-␤ signaling would inhibit osteogenic differentiation; in other words, the inhibition of Smad6 would promote osteogenic differentiation. Thus, the results in Fig. 7(b) suggest that the naringin immobilization on ozonated chitosan would be osteoconductive by inhibiting Smad6 expression. On the other hand, Smad6 phosphorylation on A-5% and A-10% was only slightly suppressed, and A-0.1% and A-0.5 did not influence Smad6 activation. These results indicate that adsorbed naringin was ineffective in suppressing Smad6 phosphorylation. Fig. 7 reveals that O-5% highly promoted osteogenic differentiation, not only because of Smad receptor activation but also because of the suppression of inhibitory Smad. These factors also explain why the osteogenic expression of UMR at O-5%, as shown in Fig. 6, was much higher than at other levels of ozonation. With ozone treatment, naringin would be immobilized on chitosan with stronger intermolecular forces, causing naringin to be released more slowly and influence cells continuously. Such

0.4

0.3

0.2

0.1

0.0 O-10%

Fig. 7. Smad1 (a) and Smad6 (b) phosphorylation on naringin-modified chitosan substrates with and without ozone pretreatment after cell culture for 3 days. PSmad1 and P-Smad6 are the phospho-Smad1 and phospho-Smad6, respectively. Brand intensities from blots were measured by densitomery using Scion Image to determine the ratio of P-Smad to Smad. “A” indicated chitosan substrates were modified without ozonation and with naringin adsoption. “O” indicated chitosan substrates were treated with ozone, and then immersed into naringin solution for modification.

a continuous contact between immobilized naringin and cells probably activates focal adhesion kinase (FAK). FAK activation has been shown to be necessary for the transcription of the Smad family, which is essential for the BMP-Smad pathway in osteogenic differentiation [24], and is probably related to the signaling of extracellular signal-regulated kinase [25]. In contrast, directly adsorbed naringin and free naringin would stimulate osteoblastic cells significantly only for a very short period. Thus, there is not enough time for naringin to suppress Smad6. For these reasons, naringin immobilized on ozonated chitosan was more osteoconductive in this research. 4. Conclusions In this study, chitosan–naringin complex substrates were successfully synthesized by ozone activation. From the modified iodide assay, ozone flow rate and activation time were optimized to produce the maximum amount of peroxides on chitosan substrates, onto which naringin was immobilized. Ozone activation reduced the release rate of naringin from the substrates. In the in vitro

C.H. Li et al. / Colloids and Surfaces B: Biointerfaces 115 (2014) 1–7

experiments, the cell affinity and osteoconductivity of chitosan were significantly enhanced by the immobilization of naringin by ozonation of its surface. Using the optimal naringin concentration in the modification process (5 wt%), osteogenic differentiation was highly promoted, indicated by the results of ALPase, OCN, BSP, and COL1 expression. Observations of Smad1 and Smad6 phosphorylation indicated that naringin immobilization by ozone treatment would activate receptor Smad1 and suppress inhibitory Smad6. The results of this research demonstrate that naringin immobilization by ozonation increased the osteoconductivity of chitosan substrates efficiently, enhancing the potential of chitosan in bone tissue engineering. Acknowledgements This work was financially supported by National Science Council, Taiwan (NSC 101-2221-E-011-094-MY3) and by National Taiwan University of Science and Technology. References [1] L.L. Chen, L.H. Lei, P.H. Ding, Q. Tang, Y.M. Wu, Arch. Oral Biol. 56 (2011) 1655. [2] J.B. Wu, Y.C. Fong, H. Tsai, Y.F. Chen, M. Tsuzuki, C.H. Tang, Pharmacol. Eur. J. 588 (2008) 333. [3] V. Habauzit, S.M. Sacco, A. Gil-Izquierdo, A. Trzeciakiewicz, C. Morand, D. Barron, S. Pinaud, E. Offord, M.N. Horcajada, Bone 49 (2011) 1108. [4] R.W.K. Wong, A.B.M. Rabie, Hong Kong Dent. J. 4 (2007) 15. [5] M.R. Urist, Science 150 (1965) 893.

7

[6] M. Geiger, R.H. Li, W. Friess, Adv. Drug Deliv. Rev. 55 (2003) 1613. [7] L. Li, Z. Zeng, G. Cai, Phytomedicine 18 (2011) 985. [8] Z. Peng, K.R. Dai, S.G. Yan, W.Q. Yan, Z. Chao, D.Q. Chen, X. Bo, Z.W. Xu, Pharmacol. Eur. J. 607 (2009) 1. [9] P. Pu, D.M. Gao, S. Mohamed, J. Chen, J. Zhang, X.Y. Zhou, N.J. Zhou, J. Xie, H. Jiang, Arch. Biochem. Biophys. 518 (2012) 61. [10] S.L. Hsiu, T.Y. Huang, Y.C. Hou, D.H. Chin, P.D. Chao, Life Sci. 70 (2002) 1481. [11] M.R. Lauro, F. DeSimone, F. Sansone, P. Iannelli, R.P. Aquino, J. Drug Del. Sci. Tech. 17 (2007) 119. [12] A. Semalty, M. Semalty, D. Singh, M.S.M. Rawat, Phenom. Macrocyclic Chem. 67 (2010) 253. [13] H. Suh, Y.S. Hwang, J.E. Lee, C.D. Han, J.C. Park, Biomaterials 22 (2001) 219. [14] H. Suh, M.J. Song, M. Ohata, Y.B. Kang, S. Tsutsumi, Tissue Eng. 9 (2003) 635. [15] M.H. Ho, L.T. Hou, C.Y. Tu, H.J. Hsieh, J.Y. Lai, W.J. Chen, D.M. Wang, Macromol. Biosci. 26 (2006) 5617. [16] M.H. Ho, L.T. Hou, C.Y. Tu, H.J. Hsieh, J.Y. Lai, W.J. Chen, D.M. Wang, Macromol. Biosci. 7 (2007) 467. [17] A.S. Roy, D.R. Tripathy, A. Chatterjee, S. Dasgupta, J. Biophys. Chem. 1 (2010) 141. [18] A.M. Hillery, A.W. Lloyd, J. Swarbrick, Drug Delivery and Targeting for Pharmacists and Pharmaceutical Scientisis, Taylor and Francis, New York, 2001. [19] W.Y. Pang, X.L. Wang, S.K. Mok, W.P. Lai, H.K. Chow, P.C. Leung, X.S. Yao, M.S. Wong, Br. J. Pharmacol. 159 (2010) 1693. [20] G.S. Stein, J.B. Lian, J.L. Stein, A.J. van Wijnen, M. Montecino, Physiol. Rev. 76 (1996) 593. [21] R.D. Farhadieh, M.P. Gianoutsos, Y. Yu, W.R. Walsh, J. Craniofac. Surg. 15 (2004) 714. [22] Y. Yu, J.L. Yang, P.J. Chapman-Sheath, W.R. Walsh, J. Biomed. Mater. Res. 60 (2000) 392. [23] T. Imamura, M. Takase, A. Nishihara, E. Oeda, J. Hanai, M. Kawabata, K. Miyazono, Nature 389 (1997) 622. [24] Y. Tamura, Y. Takeuchi, M. Suzawa, S. Fukumoto, M. Kato, K. Miyazono, T. Fujita, J. Bone Miner. Res. 16 (2001) 1772. [25] N. Ghosh-Choudhury, S.L. Abboud, R. Nishimura, A. Celeste, L. Mahimainathan, G.G. Choudhury, Biol. Chem. J. 277 (2002) 33361.