- Email: [email protected]
DNA-lipid-apatite composite layers enhance gene expression of mesenchymal stem cells
Materials Science and Engineering C 33 (2013) 512–518
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
DNA-lipid-apatite composite layers enhance gene expression of mesenchymal stem cells Xiupeng Wang a,⁎, Atsuo Ito a, Xia Li a, Yu Sogo a, Motohiro Hirose a, Ayako Oyane b, Hideo Tsurushima c a b c
Human Technology Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan Nanosystem Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Central 4, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8562, Japan Department of Neurosurgery, Faculty of Medicine, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8575, Japan
a r t i c l e
i n f o
Article history: Received 19 April 2012 Received in revised form 27 July 2012 Accepted 25 September 2012 Available online 3 October 2012 Keywords: Apatite Gene transfer Fibroblast Mesenchymal stem cell (MSC) Bone morphogenetic protein-2 (BMP-2)
a b s t r a c t DNA-apatite composite layer (D-Ap layer) and DNA-lipid-apatite composite layer (DLp-Ap layer) were prepared on ceramic hydroxyapatite disk and scaffold using supersaturated calcium phosphate solutions supplemented with 0.5–5 μg/mL plasmid and 0–10 μL/mL lipid transfection reagent FuGENE®. Both in vitro and in vivo studies were carried out using mesenchymal stem cells (MSCs) and two kinds of gene (luciferase and bone morphogenetic protein (BMP)-2) for demonstrating potential application of the gene transfer system using the D-Ap and DLp-Ap layers in bone tissue engineering. In the in vitro study using luciferase gene, the DLp-Ap layers showed 1–2 orders of magnitudes higher gene transfer efficiency to MSCs than the D-Ap layer. In the in vivo study using BMP-2 gene, DLp-Ap layer slightly increased BMP-2 protein concentration than D-Ap layer, thereby enhancing their osteogenic differentiation than D-Ap layer. The present gene transfer system using the DLp-Ap layers, with the advantages of good biocompatibility, bone-bonding ability, and efficacy in in vitro and in vivo gene transfer to MSCs, would be useful in bone tissue engineering. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Tissue engineering has emerged as an important approach to grow new tissues and organs in vitro or in vivo by using combinations of cells, biomaterials, and biologically active molecules [1–4]. For the in vivo tissue engineering, a suitable delivery system of biologically active molecules such as protein and gene is required to control cell behavior. Unfortunately, direct protein delivery can suffer from protein instability. Moreover, inadequate post-translational modifications of recombinant protein may limit protein bioactivity [5,6]. Gene transfer is preferred over the protein delivery to the cell as, typically, (1) protein is expressed by endogenous cells, (2) in situ protein expression increases the effectiveness of recombinant protein, and (3) production of DNA is more cost-effective than recombinant protein [4,7–13]. Among the gene transfer systems developed hitherto, non-viral calcium phosphate-DNA composite layer gene transfer system shows advantages of good biocompatibility, bone-bonding ability, and areaspecificity, and hence great potential in tissue engineering especially for the bone tissue [14,15]. Moreover, a gene transfer system using a DNA-apatite composite layer (D-Ap layer) showed excellent gene transfer efficiency as efficient as an optimized commercial lipid transfection
⁎ Corresponding author. Tel.: +81 29 861 6528; fax: +81 29 861 6149. E-mail address: [email protected] (X. Wang). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2012.09.023
reagent in vitro [14,15]. To further improve the gene transfer efficiency, cell adhesion molecules were immobilized in the D-Ap layer, which resulted in 1–2 orders of magnitude increase in gene transfer efficiency [16–19]. Recently, immobilization of lipid transfection reagents in the D-Ap layer [15,20] was shown to be even more effective than that of cell adhesion molecules in improving the gene transfer efficiency in vitro. Efficiency of the gene transfer system using the DNA-lipid-apatite composite layer (DLp-Ap layer) was higher than that mediated using a commercial lipid-based transfection reagent applied using the manufacture's recommended optimum condition [21]. However, these previous gene transfer studies were carried out in vitro using cell lines, and efficiency of gene transfer to mesenchymal stem cells (MSCs), which are useful in bone tissue engineering, has yet been evaluated neither in vitro nor in vivo. The gene transfer efficiency to stem/primary cells is generally low compared with that to cell lines. In this study, DLp-Ap layers were prepared on ceramic hydroxyapatite (HA) disk and scaffold using a lipid transfection reagent FuGENE® and two kinds of gene (luciferase and BMP-2). For the first time, the gene transfer efficiency of the DLp-Ap layer and the D-Ap layer as a control was tested in vitro using MSCs and fibroblastic NIH3T3 cell line for comparison. The HA scaffolds with Ap, D-Ap, and DLp-Ap layers and those seeded with MSCs were implanted in the dorsal subcutaneous tissue to test in vivo gene expression and osteogenic differentiation of cells under skin and MSCs, respectively.
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
2. Experimental 2.1. HA disk and HA scaffold preparation For HA disk preparation, stoichiometric and pure HA powders (Ca/P = 1.67, Advance Co., Ltd., Japan) supplemented with 3% polyvinyl alcohol and 1% polyethylene glycol were molded into disk shape. Then the molded powders were sintered at 1150 °C for 1 h to prepare HA disks (1 mm thickness and 13 mm diameter). The HA disk was used as a substrate for in vitro assay. For HA scaffold preparation, HA powders (3 wt.%) were dispersed in binder solution to prepare a HA suspension. Polyurethane sponges with cylindrical shape were then immersed in the HA suspension and stirred at room temperature for 2 h. Then the polyurethane sponges were picked up and dried at 60 °C for 1 day. The raw HA powders contained in the sponges were sintered at 1150 °C for 1 h to obtain cylindrical porous HA scaffolds (5 mm height and 5 mm diameter). The HA scaffold was used as a substrate for in vivo assay. 2.2. Ap, D-Ap and DLp-Ap layer preparation Ap, D-Ap and DLp-Ap layers were formed on HA disk and HA scaffold by a method described elsewhere [17,18]. Briefly, a coating solution for an apatite layer (Ap layer), i.e. a supersaturated calcium phosphate solution (CP solution) was prepared by dissolving NaCl (final concentration = 142 mM), K2HPO4·3H2O (1.50 mM), 1 M HCl (40 mM), and CaCl2 (3.75 mM) (Nacalai Tesque, Japan) in ultrapure water and then buffering at pH 7.40 at 25.0 °C with tris(hydroxymethyl)aminomethane (final concentration = 50 mM) and a necessary amount of 1 M HCl (Nacalai Tesque, Japan). Coating solutions for DLp-Ap layers were prepared by supplementing the CP solution with 0.5, 1, 2, and 5 μg/mL plasmid and 1, 2, 4, and 10 μL/mL FuGENE® (Roche, Germany). Plasmid pGL3 (Promega, USA) including luciferase cDNA was used for the in vitro study, since luciferase has been widely used as a reporter gene. Plasmid pCI-neo (Promega, USA) including BMP-2 cDNA was used for the in vivo study, since BMP-2 is effective for bone formation [22]. Before immersion in the coating solutions, the HA disk and HA scaffold were sterilized at 160 °C for 3 h in a dry sterilizer (Model SG600, Yamato Scientific, Co., Ltd., Japan). After being cooled to room temperature, the disk and scaffold were aseptically immersed in 3 mL of the coating solution at 25 °C for 24 h. 2.3. Analysis of Ap, D-Ap and DLp-Ap layers The HA disks and HA scaffolds taken out from the coating solutions were gently washed with ultrapure water and freeze-dried for scanning electron microscope (SEM) observation, analysis by X-ray diffractometer (XRD), and DNA leaching test. The surface morphology of the layers formed on the HA disks and scaffolds was observed by SEM (Model XL30, FEI Company, USA) at an accelerating voltage of 10 kV. The crystalline phase of the layers detached from the HA disks by sticking Scotch® double sided tape (Sumitomo 3 M, Japan) was analyzed by powder X-ray diffractometer (XRD, Rigaku, Japan) using Cu Kα X-ray at 40 kV and 300 mA. Because of the low DNA concentration in the coating solution and interference by FuGENE®, we failed to quantify the amount of DNA and FuGENE® in the D-Ap and DLp-Ap layers by the conventional UV spectroscopic technique [16–19]. The DNA leaching test followed by real-time polymerase chain reaction (rt-PCR) analysis was conducted for the HA scaffolds with Ap, D-Ap and DLp-Ap layers under the same condition to confirm that BMP-2 cDNA was successfully immobilized in the D-Ap and DLp-Ap layers. The HA scaffolds with Ap, D-Ap and DLp-Ap layers were respectively immersed in 0.5 mL of ultrapure water for 2 h. Then, the obtained solutions were used for rt-PCR analysis to detect any BMP-2 cDNA. The rt-PCR analysis was performed with 40 cycles
513
of denaturation at 95 °C for 10 s and annealing at 56 °C for 30 s, using an rt-PCR detection system (MiniOpticon, Bio-Rad) with the following primers: forward primer 5′-AAGGCACCCTTTGTATGTGG-3′ and reverse primer 5′-CATGCCTTAGGGATTTTGGA-3′. Amount of BMP-2 gene in the leaching solutions from the D-Ap and DLp-Ap layers was normalized to that from the Ap layer. 2.4. In vitro gene transfer to NIH3T3 cells The HA disks with D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water, and placed in each well of a cell culture plate. Then 1 mL of 5×104 cells/mL suspension of fibroblastic NIH3T3 cells (NIH3T3-3-4, Riken Bioresource Center, Japan) was poured onto the HA disk. Dulbecco's modified essential medium supplemented with 10% fetal bovine serum (FBS) was used as a culture medium. After 3 days of culture, the cells were washed three times with phosphate buffered saline (PBS(−)). After washing with PBS(−), HA disks with the cells were moved to wells of new cell culture plate one by one. Then an aliquot of 200 μL of cell culture lysis reagent (Promega, USA) was added to each well and the mixture was kept at −80 °C for 10 min. After vortexing, the supernatant was obtained by centrifuging the cell lysate at 12,000 g for 1 min. The luciferase activity was evaluated by measuring the luciferase luminescence of the supernatant using a luminometer (Gene Light 55, Microtec, Japan) employing a luciferase assay kit (Promega, USA). 2.5. In vitro gene transfer to MSCs MSCs were obtained from both femora from Fischer 344 syngeneic rats (SLC, Inc. Japan). Both ends of the rat femora were cut away from the epiphysis. Bone marrow was flushed out using 15 mL of culture medium (minimal essential medium with Earle's salts and L-glutamine) containing 10% FBS and 1% antibiotics solution (100 U/mL penicillin G, 100 μg/mL streptomycin sulfate, and 0.25 μg/mL Amphotericin B) and then dispersed through an 18-gauge needle. The marrow suspension was poured in a 75 cm2 tissue culture polystyrene flask and cultured until reaching about 90% confluence. Culture medium was replaced with fresh one every 3 days. The HA disks with D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water, and placed in each well of a cell culture plate. Then 1 mL of 5 × 104 cells/mL MSC suspension was poured onto the HA disk. After 3 days of culture, the luciferase activity was evaluated with the same procedure described in Section 2.4. 2.6. In vivo gene transfer to cells under skin The HA scaffolds with Ap, D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water and implanted into the dorsal subcutaneous tissue of Fischer 344/N rats (SLC, Inc. Japan) [23,24]. After an intraperitoneal injection of pentobarbital (60 mg/kg body weight, Somnopentil; Dainippon Sumitomo Pharma Co., Ltd., Japan), preoperative shaving, washing and sterilization of skin surface with 70% ethanol, three subcutaneous pockets were made by skin incision followed by blunt dissection of both the right and left flanks. The pockets were 15 to 20 mm apart from the midline. One pocket was filled with one scaffold. The skin incisions were then sutured using disposable skin staplers (Precise™, 3 M Health Care Ltd., Japan). One, two, three, and four weeks after the implantation, the rats were sacrificed with an overdose administration of pentobarbital. During the implantation, cells under skin (most of them were considered as fibroblasts) migrated into the macropores of the scaffolds. The scaffolds were harvested together with the cells within their macropores and used for protein assays. In more detail, the harvested scaffolds were immersed in 300 μL of 0.1% Triton X-100 and the cells within the scaffolds were
514
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
homogenized. The supernatant was obtained by centrifuging the cell suspension at 12,000 g for 1 min. The supernatant was analyzed for ALP activity, protein BMP-2 and protein OCN contents using a Laboassay™ ALP kit (Wako Pure Chemicals, Japan), a BMP-2 Quantikine® kit (R&D Systems, Inc., USA) and a rat OCN EIA kit (Biomedical Technologies, Inc., USA) in accordance with the manufacturers' instructions, respectively. 2.7. In vivo gene transfer to MSCs The HA scaffolds with Ap, D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water and seeded with MSCs (5 × 10 5 cells/scaffold). The scaffolds were allowed to stand in a humidified atmosphere of 5% CO2 at 37 °C for 3 h for MSCs attachment. Then, the scaffolds with MSCs were implanted into the dorsal subcutaneous tissue of Fischer 344/N rats, harvested, and examined as described in Section 2.6. All the animal experiments and feeding were carried out in accordance with the guidelines of the Ethical Committee of the National Institute of Advanced Industrial Science and Technology (AIST), Japan, and the National Institutes of Health, USA, for the care and use of laboratory animals (NIH Pub. No. 85-23 Rev. 1985). 3. Results
about 100–300 μm (Fig. 1, A, inset), while the HA disk showed a macroscopically dense surface (data not shown). Before coating, round shape primary particles of HA with diameter of about 500–1000 nm were observed on the HA scaffold surface (Fig. 1, A). After coating with Ap layer, a continuous and homogeneous layer consisting of fine flakes of 100–200 nm thickness and 2–5 μm length was formed on the HA scaffold surface (Fig. 1, B). The D-Ap layer showed similar surface morphology with the Ap layer (Fig. 1, C and E). Both D-Ap layers on HA scaffold and HA disk showed similar surface morphology. The flakes consisting of the DLp-Ap layer (Fig. 1, D and F) greatly decreased in size as compared with those consisting of the Ap and D-Ap layers. The flakes on the DLp-Ap layer showed about 50–100 nm thickness and 1–2 μm length (Fig. 1, D and F). The DLp-Ap layers on HA scaffold and HA disk showed similar surface morphology. Representative XRD patterns of the Ap, D-Ap, and DLp-Ap layers are shown in Fig. 2. Note that, the Ap, D-Ap, and DLp-Ap layers were mildly detached from the HA disks before XRD analysis. No interference peaks from the HA disk were observed in the XRD patterns. All the coatings consisted of low-crystalline HA, as proved by the broad peaks at 31.8°, 32.2° and 32.9° (ICDD No. 09–432). The rt-PCR analysis was carried out to confirm whether cDNA was successfully immobilized in the D-Ap and DLp-Ap layers (Fig. 3). The BMP-2 gene in the leaching solutions from the D-Ap and DLp-Ap layers was much higher than from the Ap layer. The result means D-Ap and DLp-Ap layers contain plasmid including cDNA of BMP-2.
3.1. Formation of Ap, D-Ap, and DLp-Ap layers on HA disk and HA scaffold 3.2. In vitro luciferase gene expression of NIH3T3 cells and MSCs Representative SEM images of the HA scaffold/disk before and after coating with Ap, D-Ap, and DLp-Ap layers are shown in Fig. 1. The as-prepared HA scaffold showed macropores with diameter of
To check the efficiency of gene transfer to the NIH3T3 cells and MSCs on the D-Ap and DLp-Ap layers and to optimize the preparation
Fig. 1. Representative SEM images of as-prepared HA scaffold (A, inset: low magnification image of HA scaffold); HA scaffold with Ap layer (B); HA scaffold with D-Ap layer (C, 1 μg/mL DNA in coating solution); HA scaffold with DLp-Ap layer (D, 2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution); HA disk with D-Ap layer (E, 1 μg/mL DNA in coating solution); HA disk with DLp-Ap layer (F, 2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution).
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
apatite: * Intensity (a.u.)
**
*
DLp-Ap D-Ap Ap
Luciferase activity (Count)
1E7
28
30
32
34
36
A
DLp-Ap
1E6
1E5
1E4
1E3
100 26
515
D-Ap
10
2θ (o)
1
0.5
Fig. 2. Representative XRD patterns of Ap, D-Ap (1 μg/mL DNA in coating solution), and DLp-Ap (2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution) layers scraped off from the HA disks.
Increasing Fold (Times)
10000
8000
6000
4000
2000
0 Ap
D-Ap
DLp-Ap
Fig. 3. Relative amount of BMP-2 gene in ultrapure water after immersion of HA scaffolds with the Ap, D-Ap and DLp-Ap layers (quantified by rt-PCR).
5
DLp-Ap
Luciferase activity (Count)
1000
100
D-Ap
10
1
3.3. In vivo BMP-2 protein concentration of cells under skin and MSCs The HA scaffolds with Ap, D-Ap, and DLp-Ap layers (Section 2.6) and those seeded with MSCs (Section 2.7) were implanted into the subcutaneous tissue of rats to evaluate in vivo BMP-2 expression to cells under skin (mostly fibroblasts) and MSCs. Both cells under skin (Fig. 5, A) and MSCs (Fig. 5, B) on the D-Ap and DLp-Ap layers showed improved BMP-2 protein concentration than those on the Ap layer. Whereas, only slightly improved BMP-2 protein concentration was observed on DLp-Ap layer than D-Ap layer. BMP-2 protein concentration on the DLp-Ap layer was 1.8 fold that on the D-Ap layer for MSCs at the implantation period of 1 week (Fig. 5, B).
2
B 10000
condition of the DLp-Ap layer, luciferase assay was performed. As shown in Fig. 4, the DLp-Ap layers showed 4 and 1–2 orders of magnitudes higher gene transfer efficiency as compared with the D-Ap layer for the NIH3T3 cells (Fig. 4, A) and MSCs (Fig. 4, B), respectively. Among the DLp-Ap layers, that prepared in the coating solution with 2 μg/mL DNA and 4 μL/mL FuGENE showed the highest gene transfer efficiency to MSCs, and this was used in the following in vivo experiments.
1
DNA concentration (μg/mL)
0.5
1
2
5
DNA concentration (μg/mL) Fig. 4. In vitro gene transfer efficiency to NIH3T3 cells (A) and MSCs (B) on the D-Ap (control) and DLp-Ap layers prepared in the coating solutions with different DNA concentrations.
3.4. In vivo ALP activity and OCN protein concentration of cells under skin and MSCs To confirm that BMP-2 enhances osteogenic differentiation of cells under skin and MSCs in vivo, the osteogenic differentiation markers (ALP and OCN) were investigated at the protein level. Both cells under skin (Fig. 6, A) and MSCs (Fig. 6, B) on the D-Ap and DLp-Ap layers showed an increase in ALP activity, an early marker of osteogenic differentiation as compared with those on the Ap layer, at the implantation period of 2 weeks. Improvement in ALP activity on the DLp-Ap layer in comparison with that on the D-Ap layer was found for the cells under skin and MSCs, at the implantation period of 1 and 2 weeks. Similar to the results on ALP activity (Fig. 6), both cells under skin (Fig. 7, A) and MSCs (Fig. 7, B) on the D-Ap and DLp-Ap layers showed slightly increased OCN protein concentration, a late marker of osteogenic differentiation, as compared with those on the Ap layer. Slight improvement in OCN protein concentration was found for either of the two cells on the DLp-Ap layer compared with those on the D-Ap layer. 4. Discussion The Ap, D-Ap and DLp-Ap layers were formed on HA disks and HA scaffolds in the coating solutions, i.e., the CP solution supplemented
516
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
A
1w 2w 3w 4w
40
BMP-2 (pg/sample)
35 30
600
ALP activity (nmol/min/sample)
45
25 20 15 10
A
1w 2w 3w 4w
500 400 300 200 100
5 0
0 Ap
B
1w 2w 3w 4w
40
BMP-2 (pg/sample)
35 30
Ap
DLp-Ap
600
ALP activity (nmol/min/sample)
45
D-Ap
25 20 15 10
B
D-Ap
DLp-Ap
D-Ap
DLp-Ap
1w 2w 3w 4w
500 400 300 200 100
5 0
0 Ap
D-Ap
DLp-Ap
Ap
Fig. 5. BMP-2 protein concentration of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
Fig. 6. ALP activity of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
with various concentrations of plasmid (0–5 μg/mL) and lipid transfection reagent FuGENE® (0–10 μL/mL) (Figs. 1–3). The DLp-Ap layer is likely to be formed via the following process. Sintered HA particles at the surface of the disk and scaffold induce secondary nucleation of HA on their surfaces in the coating solution, as the solution is highly supersaturated with respect to HA. The HA nuclei grow into flakes, merged, and finally formed a continuous and homogeneous layer. During the spontaneous growth of HA crystals on the HA disk and HA scaffold surfaces, plasmid and lipid (FuGENE®), in the form of either discrete molecule or complex, are adsorbed onto the growing crystal surface through electrostatic interaction, taking into account our previous reports on protein-apatite composite layers [25–33]. The obviously decreased HA crystal size of the DLp-Ap layer (Fig. 1, D) than that of the Ap layer (Fig. 1, B) is probably due to the surface adsorption of plasmid and lipid during the coating process, which inhibits crystal growth of HA. As a result of such a coprecipitation process, DNA and lipid are immobilized in the DLp-Ap layer, compared with a simple adsorption process on the Ap layer surface, although we failed to quantify the amount of DNA and lipid in the layers. The present gene transfer system using the D-Ap and DLp-Ap layers was effective in not only in vitro gene transfer to cell line as was reported previously [21] but also in vitro and in vivo gene transfer to MSCs (stem cells, primary cells). It should be noted that the gene transfer efficiency to stem/primary cells is generally low compared with that to cell lines. MSCs are multipotent stem cells that can differentiate into a variety of cell types, including chondrocytes, osteoblasts, adipocytes, fibroblasts, marrow stroma cells, and other cells
of mesenchymal origin. MSCs have enormous therapeutic potential for tissue repair owing to their wide source in bone marrow and high capacity for self-renewal while maintaining multipotency [7–10,34]. Although any therapeutic effect of the MSCs seeded on the scaffold was unapparent in this animal model, the present gene transfer system using the D-Ap and DLp-Ap layers, with potential to control behavior of MSCs, would be useful in tissue engineering. Potential of the present gene transfer system to control cell behavior was demonstrated for cells under skin (Figs. 5–7, A) and MSCs (Figs. 5–7, B) by the in vivo results. The in vivo results showed that immobilization of BMP-2 gene into the Ap layer (resulting in the D-Ap and DLp-Ap layers) increased BMP-2 protein concentration (Fig. 5), thereby increasing ALP activity and OCN protein concentration (Figs. 6 and 7). These findings indicate that BMP-2 expressed by gene transfer from the D-Ap and DLp-Ap layers retained its biological activity and enhanced osteogenic differentiation of these cells adhering to the layers in vivo. Generally, non-viral gene transfer systems are less efficient than viral systems, and have difficulty in attaining gene expression at therapeutically effective level in vivo. The present gene transfer system using the D-Ap and DLp-Ap layers is a non-viral system effective in enhancing cell differentiation in vivo. Therefore, the present gene transfer system has advantages of high efficiency in addition to good biocompatibility and bone-bonding ability owing to the HA matrix. The high efficiency should be due to the sustained release of gene from the D-Ap and DLp-Ap layers, which acts to increase the local concentration of gene and/or lipid within the vicinity of the cell [4,11,35]. It is hypothesized that the sustained gene release is mediated by the gradual dissolution of the HA matrix
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
3.0
A
OCN (ng/sample)
DLp-Ap layer and the D-Ap layer as a control was tested in vitro and in vivo using MSCs and fibroblastic NIH3T3/cells under skin. The DLp-Ap layers showed 1–2 orders of magnitudes higher gene transfer efficiency to MSCs than the D-Ap layer in vitro. The DLp-Ap layer slightly increased BMP-2 protein concentration than D-Ap layer, thereby enhancing their osteogenic differentiation than D-Ap layer in vivo. The present gene transfer system is a non-viral system with the advantages of good biocompatibility, bone-bonding ability, and the potential to control cell differentiation of MSCs, hence it would be useful in bone tissue engineering.
1w 2w 3w 4w
2.5
2.0
1.5
1.0
Acknowledgment
0.5
0.0 Ap
3.0
B
D-Ap
DLp-Ap
1w 2w 3w 4w
2.5
OCN (ng/sample)
517
2.0
References
1.5
1.0
0.5
0.0 Ap
We thank Dr. Hajime Ohgushi, Dr. Mika Tadokoro, Dr. Koji Hattori, Dr. Youichi Katou and Mr. Yushin Yazaki from the National Institute of Advanced Industrial Science and Technology (AIST) for technical assistance. This study was supported in part by KAKENHI (Grant-in-Aid for Young Scientists B; 22700499) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.
D-Ap
DLp-Ap
Fig. 7. OCN protein concentration of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
serving as a reservoir for gene. This hypothesis is supported by our previous reports that the gene release is accompanied by partial dissolution of the HA matrix and this plays a critical role in controlling gene transfer efficiency and timing on the layer [36]. Immobilization of lipid transfection reagent FuGENE® to the D-Ap layer (resulting in the DLp-Ap layer) remarkably improved efficiency of gene transfer in vitro, although this improvement was less noticeable in vivo. The cells grow in a much more complex environment in vivo than in vitro. 1) Many factors influence in vivo cellular response, e.g. a complex extracellular matrix, soluble bioactive factors, and products of homo- and hetero-typical cell–cell interactions. 2) The cells grow in a circulation environment in vivo, thus the DNA concentration in vivo is lower than the still case in vitro. 3) The serum in vivo generally reduces the gene transfer efficiency of conventional lipids due to the interaction of serum components with the DNA–lipid complexes, which causes disintegration of the complexes and degradation of DNA molecules [37–40]. 4) Inflammatory response occurs immediately to implantation in vivo, which is different with in vitro case. Therefore, the optimized gene transfection condition obtained based on the in vitro results may not apply to the in vivo case. The in vivo gene transfection condition should be further optimized to prepare more effective composite apatite layer for their in vivo applications. 5. Conclusions DLp-Ap and D-Ap layers were prepared using supersaturated calcium phosphate solutions supplemented with 0.5–5 μg/mL plasmid and 0–10 μL/mL FuGENE®. The gene transfer efficiency of the
[1] R. Langer, D.A. Tirrell, Nature 428 (2004) 487–492. [2] S. Mitragotri, J. Lahann, Nat. Mater. 8 (2009) 15–23. [3] D.S. Benoit, M.P. Schwartz, A.R. Durney, K.S. Anseth, Nat. Mater. 7 (2008) 816–823. [4] L.D. Shea, E. Smiley, J. Bonadio, D.J. Mooney, Nat. Biotechnol. 17 (1999) 551–554. [5] Y.C. Huang, C. Simmons, D. Kaigler, K.G. Rice, D.J. Mooney, Gene Ther. 12 (2005) 418–426. [6] J. Fang, Y.Y. Zhu, E. Smiley, J. Bonadio, J.P. Rouleau, S.A. Goldstein, et al., Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 5753–5758. [7] H. Hosseinkhani, T. Azzam, H. Kobayashi, Y. Hiraoka, H. Shimokawa, A.J. Domb, et al., Biomaterials 27 (2006) 4269–4278. [8] H. Hosseinkhani, M. Hosseinkhani, F. Tian, H. Kobayashi, Y. Tabata, Biomaterials 27 (2006) 5089–5098. [9] H. Hosseinkhani, M. Hosseinkhani, F. Tian, H. Kobayashi, Y. Tabata, Biomaterials 27 (2006) 4079–4086. [10] H. Hosseinkhani, M. Yamamoto, Y. Inatsugu, Y. Hiraoka, S. Inoue, H. Shimokawa, et al., Biomaterials 27 (2006) 1387–1398. [11] J. Bonadio, E. Smiley, P. Patil, S. Goldstein, Nat. Med. 5 (1999) 753–759. [12] D. Lew, S.E. Parker, T. Latimer, A.M. Abai, A. Kuwahararundell, S.G. Doh, et al., Hum. Gene Ther. 6 (1995) 553–564. [13] S.E. Parker, H.L. Vahlsing, L.M. Serfilippi, C.L. Franklin, S.G. Doh, S.H. Gromkowski, et al., Hum. Gene Ther. 6 (1995) 575–590. [14] H. Shen, J. Tan, W.M. Saltzman, Nat. Mater. 3 (2004) 569–574. [15] B.B. Sun, K.K. Tran, H. Shen, Biomaterials 30 (2009) 6386–6393. [16] A. Oyane, M. Murayama, A. Yamazaki, Y. Sogo, A. Ito, H. Tsurushima, J. Biomed. Mater. Res. A 92 (2010) 1038–1047. [17] A. Oyane, H. Tsurushima, A. Ito, Gene Ther. 14 (2007) 1750–1753. [18] A. Oyane, H. Tsurushima, A. Ito, J. Gene Med. 12 (2010) 194–206. [19] X. Wang, A. Oyane, H. Tsurushima, Y. Sogo, X. Li, A. Ito, Biomed. Mater. 6 (2011) 045004. [20] D.H. Kohn, L.N. Luong, K.M. McFalls, Biomaterials 30 (2009) 6996–7004. [21] A. Oyane, Y. Yazaki, H. Araki, Y. Sogo, A. Ito, A. Yamazaki, H. Tsurushima, J. Mater. Sci. Mater. Med. 23 (2012) 1011–1019. [22] E.A. Wang, V. Rosen, J.S. Dalessandro, M. Bauduy, P. Cordes, T. Harada, et al., Proc. Natl. Acad. Sci. U. S. A. 87 (1990) 2220–2224. [23] H. Ohgushi, M. Okumura, S. Tamai, E.C. Shors, A.I. Caplan, J. Biomed. Mater. Res. 24 (1990) 1563–1570. [24] Y. Dohi, H. Ohgushi, S. Tabata, T. Yoshikawa, K. Dohi, T. Moriyama, J. Bone Miner. Res. 7 (1992) 1173–1180. [25] X. Wang, A. Ito, Y. Sogo, X. Li, H. Tsurushima, A. Oyane, Acta Biomater. 5 (2009) 2647–2656. [26] X. Li, X.P. Wang, A. Ito, Y. Sogo, K. Cheng, A. Oyane, Mater. Sci. Eng. C 29 (2009) 216–221. [27] A. Oyane, M. Uchida, K. Onuma, A. Ito, Biomaterials 27 (2006) 167–175. [28] A. Oyane, Y. Yokoyama, M. Kasahara, N. Ichinose, M. Saito, A. Ito, Mater. Sci. Eng. C 29 (2009) 1681–1686. [29] X. Wang, A. Ito, X. Li, Y. Sogo, A. Oyane, Biofabrication 3 (2011) 022001. [30] X.P. Wang, A. Ito, Y. Sogo, X. Li, A. Oyane, J. Biomed. Mater. Res. A 92A (2010) 1181–1189. [31] X.P. Wang, A. Ito, Y. Sogo, X. Li, A. Oyane, Acta Biomater. 6 (2010) 962–968. [32] X.P. Wang, X. Li, K. Onuma, A. Ito, Y. Sogo, K. Kosuge, et al., J. Mater. Chem. 20 (2010) 6437–6445. [33] A. Oyane, T. Ootsuka, K. Hayama, Y. Sogo, A. Ito, Acta Biomater. 7 (2011) 2969–2976. [34] G. Ciapetti, L. Ambrosio, G. Marletta, N. Baldini, A. Giunti, Biomaterials 27 (2006) 6150–6160. [35] J.M. Fang, Y.Y. Zhu, E. Smiley, J. Bonadio, J.P. Rouleau, S.A. Goldstein, et al., Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 5753–5758.
518
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
[36] Y. Yazaki, A. Oyane, Y. Sogo, A. Ito, A. Yamazaki, H. Tsurushima, Biomaterials 32 (2011) 4896–4902. [37] J.P. Yang, L. Huang, Gene Ther. 4 (1997) 950–960. [38] J.P. Yang, L. Huang, Gene Ther. 5 (1998) 380–387.
[39] S. Li, W.C. Tseng, D.B. Stolz, S.P. Wu, S.C. Watkins, L. Huang, Gene Ther. 6 (1999) 585–594. [40] A. Oyane, Y. Yazaki, H. Araki, Y. Sogo, A. Ito, A. Yamazaki, et al., J. Mater. Sci. Mater. Med. 23 (2012) 1011–1019.
Contents lists available at SciVerse ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
DNA-lipid-apatite composite layers enhance gene expression of mesenchymal stem cells Xiupeng Wang a,⁎, Atsuo Ito a, Xia Li a, Yu Sogo a, Motohiro Hirose a, Ayako Oyane b, Hideo Tsurushima c a b c
Human Technology Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Central 6, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan Nanosystem Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Central 4, 1-1-1 Higashi, Tsukuba, Ibaraki 305-8562, Japan Department of Neurosurgery, Faculty of Medicine, University of Tsukuba, 1-1-1 Tennoudai, Tsukuba, Ibaraki 305-8575, Japan
a r t i c l e
i n f o
Article history: Received 19 April 2012 Received in revised form 27 July 2012 Accepted 25 September 2012 Available online 3 October 2012 Keywords: Apatite Gene transfer Fibroblast Mesenchymal stem cell (MSC) Bone morphogenetic protein-2 (BMP-2)
a b s t r a c t DNA-apatite composite layer (D-Ap layer) and DNA-lipid-apatite composite layer (DLp-Ap layer) were prepared on ceramic hydroxyapatite disk and scaffold using supersaturated calcium phosphate solutions supplemented with 0.5–5 μg/mL plasmid and 0–10 μL/mL lipid transfection reagent FuGENE®. Both in vitro and in vivo studies were carried out using mesenchymal stem cells (MSCs) and two kinds of gene (luciferase and bone morphogenetic protein (BMP)-2) for demonstrating potential application of the gene transfer system using the D-Ap and DLp-Ap layers in bone tissue engineering. In the in vitro study using luciferase gene, the DLp-Ap layers showed 1–2 orders of magnitudes higher gene transfer efficiency to MSCs than the D-Ap layer. In the in vivo study using BMP-2 gene, DLp-Ap layer slightly increased BMP-2 protein concentration than D-Ap layer, thereby enhancing their osteogenic differentiation than D-Ap layer. The present gene transfer system using the DLp-Ap layers, with the advantages of good biocompatibility, bone-bonding ability, and efficacy in in vitro and in vivo gene transfer to MSCs, would be useful in bone tissue engineering. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Tissue engineering has emerged as an important approach to grow new tissues and organs in vitro or in vivo by using combinations of cells, biomaterials, and biologically active molecules [1–4]. For the in vivo tissue engineering, a suitable delivery system of biologically active molecules such as protein and gene is required to control cell behavior. Unfortunately, direct protein delivery can suffer from protein instability. Moreover, inadequate post-translational modifications of recombinant protein may limit protein bioactivity [5,6]. Gene transfer is preferred over the protein delivery to the cell as, typically, (1) protein is expressed by endogenous cells, (2) in situ protein expression increases the effectiveness of recombinant protein, and (3) production of DNA is more cost-effective than recombinant protein [4,7–13]. Among the gene transfer systems developed hitherto, non-viral calcium phosphate-DNA composite layer gene transfer system shows advantages of good biocompatibility, bone-bonding ability, and areaspecificity, and hence great potential in tissue engineering especially for the bone tissue [14,15]. Moreover, a gene transfer system using a DNA-apatite composite layer (D-Ap layer) showed excellent gene transfer efficiency as efficient as an optimized commercial lipid transfection
⁎ Corresponding author. Tel.: +81 29 861 6528; fax: +81 29 861 6149. E-mail address: [email protected] (X. Wang). 0928-4931/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.msec.2012.09.023
reagent in vitro [14,15]. To further improve the gene transfer efficiency, cell adhesion molecules were immobilized in the D-Ap layer, which resulted in 1–2 orders of magnitude increase in gene transfer efficiency [16–19]. Recently, immobilization of lipid transfection reagents in the D-Ap layer [15,20] was shown to be even more effective than that of cell adhesion molecules in improving the gene transfer efficiency in vitro. Efficiency of the gene transfer system using the DNA-lipid-apatite composite layer (DLp-Ap layer) was higher than that mediated using a commercial lipid-based transfection reagent applied using the manufacture's recommended optimum condition [21]. However, these previous gene transfer studies were carried out in vitro using cell lines, and efficiency of gene transfer to mesenchymal stem cells (MSCs), which are useful in bone tissue engineering, has yet been evaluated neither in vitro nor in vivo. The gene transfer efficiency to stem/primary cells is generally low compared with that to cell lines. In this study, DLp-Ap layers were prepared on ceramic hydroxyapatite (HA) disk and scaffold using a lipid transfection reagent FuGENE® and two kinds of gene (luciferase and BMP-2). For the first time, the gene transfer efficiency of the DLp-Ap layer and the D-Ap layer as a control was tested in vitro using MSCs and fibroblastic NIH3T3 cell line for comparison. The HA scaffolds with Ap, D-Ap, and DLp-Ap layers and those seeded with MSCs were implanted in the dorsal subcutaneous tissue to test in vivo gene expression and osteogenic differentiation of cells under skin and MSCs, respectively.
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
2. Experimental 2.1. HA disk and HA scaffold preparation For HA disk preparation, stoichiometric and pure HA powders (Ca/P = 1.67, Advance Co., Ltd., Japan) supplemented with 3% polyvinyl alcohol and 1% polyethylene glycol were molded into disk shape. Then the molded powders were sintered at 1150 °C for 1 h to prepare HA disks (1 mm thickness and 13 mm diameter). The HA disk was used as a substrate for in vitro assay. For HA scaffold preparation, HA powders (3 wt.%) were dispersed in binder solution to prepare a HA suspension. Polyurethane sponges with cylindrical shape were then immersed in the HA suspension and stirred at room temperature for 2 h. Then the polyurethane sponges were picked up and dried at 60 °C for 1 day. The raw HA powders contained in the sponges were sintered at 1150 °C for 1 h to obtain cylindrical porous HA scaffolds (5 mm height and 5 mm diameter). The HA scaffold was used as a substrate for in vivo assay. 2.2. Ap, D-Ap and DLp-Ap layer preparation Ap, D-Ap and DLp-Ap layers were formed on HA disk and HA scaffold by a method described elsewhere [17,18]. Briefly, a coating solution for an apatite layer (Ap layer), i.e. a supersaturated calcium phosphate solution (CP solution) was prepared by dissolving NaCl (final concentration = 142 mM), K2HPO4·3H2O (1.50 mM), 1 M HCl (40 mM), and CaCl2 (3.75 mM) (Nacalai Tesque, Japan) in ultrapure water and then buffering at pH 7.40 at 25.0 °C with tris(hydroxymethyl)aminomethane (final concentration = 50 mM) and a necessary amount of 1 M HCl (Nacalai Tesque, Japan). Coating solutions for DLp-Ap layers were prepared by supplementing the CP solution with 0.5, 1, 2, and 5 μg/mL plasmid and 1, 2, 4, and 10 μL/mL FuGENE® (Roche, Germany). Plasmid pGL3 (Promega, USA) including luciferase cDNA was used for the in vitro study, since luciferase has been widely used as a reporter gene. Plasmid pCI-neo (Promega, USA) including BMP-2 cDNA was used for the in vivo study, since BMP-2 is effective for bone formation [22]. Before immersion in the coating solutions, the HA disk and HA scaffold were sterilized at 160 °C for 3 h in a dry sterilizer (Model SG600, Yamato Scientific, Co., Ltd., Japan). After being cooled to room temperature, the disk and scaffold were aseptically immersed in 3 mL of the coating solution at 25 °C for 24 h. 2.3. Analysis of Ap, D-Ap and DLp-Ap layers The HA disks and HA scaffolds taken out from the coating solutions were gently washed with ultrapure water and freeze-dried for scanning electron microscope (SEM) observation, analysis by X-ray diffractometer (XRD), and DNA leaching test. The surface morphology of the layers formed on the HA disks and scaffolds was observed by SEM (Model XL30, FEI Company, USA) at an accelerating voltage of 10 kV. The crystalline phase of the layers detached from the HA disks by sticking Scotch® double sided tape (Sumitomo 3 M, Japan) was analyzed by powder X-ray diffractometer (XRD, Rigaku, Japan) using Cu Kα X-ray at 40 kV and 300 mA. Because of the low DNA concentration in the coating solution and interference by FuGENE®, we failed to quantify the amount of DNA and FuGENE® in the D-Ap and DLp-Ap layers by the conventional UV spectroscopic technique [16–19]. The DNA leaching test followed by real-time polymerase chain reaction (rt-PCR) analysis was conducted for the HA scaffolds with Ap, D-Ap and DLp-Ap layers under the same condition to confirm that BMP-2 cDNA was successfully immobilized in the D-Ap and DLp-Ap layers. The HA scaffolds with Ap, D-Ap and DLp-Ap layers were respectively immersed in 0.5 mL of ultrapure water for 2 h. Then, the obtained solutions were used for rt-PCR analysis to detect any BMP-2 cDNA. The rt-PCR analysis was performed with 40 cycles
513
of denaturation at 95 °C for 10 s and annealing at 56 °C for 30 s, using an rt-PCR detection system (MiniOpticon, Bio-Rad) with the following primers: forward primer 5′-AAGGCACCCTTTGTATGTGG-3′ and reverse primer 5′-CATGCCTTAGGGATTTTGGA-3′. Amount of BMP-2 gene in the leaching solutions from the D-Ap and DLp-Ap layers was normalized to that from the Ap layer. 2.4. In vitro gene transfer to NIH3T3 cells The HA disks with D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water, and placed in each well of a cell culture plate. Then 1 mL of 5×104 cells/mL suspension of fibroblastic NIH3T3 cells (NIH3T3-3-4, Riken Bioresource Center, Japan) was poured onto the HA disk. Dulbecco's modified essential medium supplemented with 10% fetal bovine serum (FBS) was used as a culture medium. After 3 days of culture, the cells were washed three times with phosphate buffered saline (PBS(−)). After washing with PBS(−), HA disks with the cells were moved to wells of new cell culture plate one by one. Then an aliquot of 200 μL of cell culture lysis reagent (Promega, USA) was added to each well and the mixture was kept at −80 °C for 10 min. After vortexing, the supernatant was obtained by centrifuging the cell lysate at 12,000 g for 1 min. The luciferase activity was evaluated by measuring the luciferase luminescence of the supernatant using a luminometer (Gene Light 55, Microtec, Japan) employing a luciferase assay kit (Promega, USA). 2.5. In vitro gene transfer to MSCs MSCs were obtained from both femora from Fischer 344 syngeneic rats (SLC, Inc. Japan). Both ends of the rat femora were cut away from the epiphysis. Bone marrow was flushed out using 15 mL of culture medium (minimal essential medium with Earle's salts and L-glutamine) containing 10% FBS and 1% antibiotics solution (100 U/mL penicillin G, 100 μg/mL streptomycin sulfate, and 0.25 μg/mL Amphotericin B) and then dispersed through an 18-gauge needle. The marrow suspension was poured in a 75 cm2 tissue culture polystyrene flask and cultured until reaching about 90% confluence. Culture medium was replaced with fresh one every 3 days. The HA disks with D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water, and placed in each well of a cell culture plate. Then 1 mL of 5 × 104 cells/mL MSC suspension was poured onto the HA disk. After 3 days of culture, the luciferase activity was evaluated with the same procedure described in Section 2.4. 2.6. In vivo gene transfer to cells under skin The HA scaffolds with Ap, D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water and implanted into the dorsal subcutaneous tissue of Fischer 344/N rats (SLC, Inc. Japan) [23,24]. After an intraperitoneal injection of pentobarbital (60 mg/kg body weight, Somnopentil; Dainippon Sumitomo Pharma Co., Ltd., Japan), preoperative shaving, washing and sterilization of skin surface with 70% ethanol, three subcutaneous pockets were made by skin incision followed by blunt dissection of both the right and left flanks. The pockets were 15 to 20 mm apart from the midline. One pocket was filled with one scaffold. The skin incisions were then sutured using disposable skin staplers (Precise™, 3 M Health Care Ltd., Japan). One, two, three, and four weeks after the implantation, the rats were sacrificed with an overdose administration of pentobarbital. During the implantation, cells under skin (most of them were considered as fibroblasts) migrated into the macropores of the scaffolds. The scaffolds were harvested together with the cells within their macropores and used for protein assays. In more detail, the harvested scaffolds were immersed in 300 μL of 0.1% Triton X-100 and the cells within the scaffolds were
514
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
homogenized. The supernatant was obtained by centrifuging the cell suspension at 12,000 g for 1 min. The supernatant was analyzed for ALP activity, protein BMP-2 and protein OCN contents using a Laboassay™ ALP kit (Wako Pure Chemicals, Japan), a BMP-2 Quantikine® kit (R&D Systems, Inc., USA) and a rat OCN EIA kit (Biomedical Technologies, Inc., USA) in accordance with the manufacturers' instructions, respectively. 2.7. In vivo gene transfer to MSCs The HA scaffolds with Ap, D-Ap and DLp-Ap layers taken out from the coating solutions were gently washed with sterilized water and seeded with MSCs (5 × 10 5 cells/scaffold). The scaffolds were allowed to stand in a humidified atmosphere of 5% CO2 at 37 °C for 3 h for MSCs attachment. Then, the scaffolds with MSCs were implanted into the dorsal subcutaneous tissue of Fischer 344/N rats, harvested, and examined as described in Section 2.6. All the animal experiments and feeding were carried out in accordance with the guidelines of the Ethical Committee of the National Institute of Advanced Industrial Science and Technology (AIST), Japan, and the National Institutes of Health, USA, for the care and use of laboratory animals (NIH Pub. No. 85-23 Rev. 1985). 3. Results
about 100–300 μm (Fig. 1, A, inset), while the HA disk showed a macroscopically dense surface (data not shown). Before coating, round shape primary particles of HA with diameter of about 500–1000 nm were observed on the HA scaffold surface (Fig. 1, A). After coating with Ap layer, a continuous and homogeneous layer consisting of fine flakes of 100–200 nm thickness and 2–5 μm length was formed on the HA scaffold surface (Fig. 1, B). The D-Ap layer showed similar surface morphology with the Ap layer (Fig. 1, C and E). Both D-Ap layers on HA scaffold and HA disk showed similar surface morphology. The flakes consisting of the DLp-Ap layer (Fig. 1, D and F) greatly decreased in size as compared with those consisting of the Ap and D-Ap layers. The flakes on the DLp-Ap layer showed about 50–100 nm thickness and 1–2 μm length (Fig. 1, D and F). The DLp-Ap layers on HA scaffold and HA disk showed similar surface morphology. Representative XRD patterns of the Ap, D-Ap, and DLp-Ap layers are shown in Fig. 2. Note that, the Ap, D-Ap, and DLp-Ap layers were mildly detached from the HA disks before XRD analysis. No interference peaks from the HA disk were observed in the XRD patterns. All the coatings consisted of low-crystalline HA, as proved by the broad peaks at 31.8°, 32.2° and 32.9° (ICDD No. 09–432). The rt-PCR analysis was carried out to confirm whether cDNA was successfully immobilized in the D-Ap and DLp-Ap layers (Fig. 3). The BMP-2 gene in the leaching solutions from the D-Ap and DLp-Ap layers was much higher than from the Ap layer. The result means D-Ap and DLp-Ap layers contain plasmid including cDNA of BMP-2.
3.1. Formation of Ap, D-Ap, and DLp-Ap layers on HA disk and HA scaffold 3.2. In vitro luciferase gene expression of NIH3T3 cells and MSCs Representative SEM images of the HA scaffold/disk before and after coating with Ap, D-Ap, and DLp-Ap layers are shown in Fig. 1. The as-prepared HA scaffold showed macropores with diameter of
To check the efficiency of gene transfer to the NIH3T3 cells and MSCs on the D-Ap and DLp-Ap layers and to optimize the preparation
Fig. 1. Representative SEM images of as-prepared HA scaffold (A, inset: low magnification image of HA scaffold); HA scaffold with Ap layer (B); HA scaffold with D-Ap layer (C, 1 μg/mL DNA in coating solution); HA scaffold with DLp-Ap layer (D, 2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution); HA disk with D-Ap layer (E, 1 μg/mL DNA in coating solution); HA disk with DLp-Ap layer (F, 2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution).
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
apatite: * Intensity (a.u.)
**
*
DLp-Ap D-Ap Ap
Luciferase activity (Count)
1E7
28
30
32
34
36
A
DLp-Ap
1E6
1E5
1E4
1E3
100 26
515
D-Ap
10
2θ (o)
1
0.5
Fig. 2. Representative XRD patterns of Ap, D-Ap (1 μg/mL DNA in coating solution), and DLp-Ap (2 μg/mL DNA and 4 μL/mL FuGENE® in coating solution) layers scraped off from the HA disks.
Increasing Fold (Times)
10000
8000
6000
4000
2000
0 Ap
D-Ap
DLp-Ap
Fig. 3. Relative amount of BMP-2 gene in ultrapure water after immersion of HA scaffolds with the Ap, D-Ap and DLp-Ap layers (quantified by rt-PCR).
5
DLp-Ap
Luciferase activity (Count)
1000
100
D-Ap
10
1
3.3. In vivo BMP-2 protein concentration of cells under skin and MSCs The HA scaffolds with Ap, D-Ap, and DLp-Ap layers (Section 2.6) and those seeded with MSCs (Section 2.7) were implanted into the subcutaneous tissue of rats to evaluate in vivo BMP-2 expression to cells under skin (mostly fibroblasts) and MSCs. Both cells under skin (Fig. 5, A) and MSCs (Fig. 5, B) on the D-Ap and DLp-Ap layers showed improved BMP-2 protein concentration than those on the Ap layer. Whereas, only slightly improved BMP-2 protein concentration was observed on DLp-Ap layer than D-Ap layer. BMP-2 protein concentration on the DLp-Ap layer was 1.8 fold that on the D-Ap layer for MSCs at the implantation period of 1 week (Fig. 5, B).
2
B 10000
condition of the DLp-Ap layer, luciferase assay was performed. As shown in Fig. 4, the DLp-Ap layers showed 4 and 1–2 orders of magnitudes higher gene transfer efficiency as compared with the D-Ap layer for the NIH3T3 cells (Fig. 4, A) and MSCs (Fig. 4, B), respectively. Among the DLp-Ap layers, that prepared in the coating solution with 2 μg/mL DNA and 4 μL/mL FuGENE showed the highest gene transfer efficiency to MSCs, and this was used in the following in vivo experiments.
1
DNA concentration (μg/mL)
0.5
1
2
5
DNA concentration (μg/mL) Fig. 4. In vitro gene transfer efficiency to NIH3T3 cells (A) and MSCs (B) on the D-Ap (control) and DLp-Ap layers prepared in the coating solutions with different DNA concentrations.
3.4. In vivo ALP activity and OCN protein concentration of cells under skin and MSCs To confirm that BMP-2 enhances osteogenic differentiation of cells under skin and MSCs in vivo, the osteogenic differentiation markers (ALP and OCN) were investigated at the protein level. Both cells under skin (Fig. 6, A) and MSCs (Fig. 6, B) on the D-Ap and DLp-Ap layers showed an increase in ALP activity, an early marker of osteogenic differentiation as compared with those on the Ap layer, at the implantation period of 2 weeks. Improvement in ALP activity on the DLp-Ap layer in comparison with that on the D-Ap layer was found for the cells under skin and MSCs, at the implantation period of 1 and 2 weeks. Similar to the results on ALP activity (Fig. 6), both cells under skin (Fig. 7, A) and MSCs (Fig. 7, B) on the D-Ap and DLp-Ap layers showed slightly increased OCN protein concentration, a late marker of osteogenic differentiation, as compared with those on the Ap layer. Slight improvement in OCN protein concentration was found for either of the two cells on the DLp-Ap layer compared with those on the D-Ap layer. 4. Discussion The Ap, D-Ap and DLp-Ap layers were formed on HA disks and HA scaffolds in the coating solutions, i.e., the CP solution supplemented
516
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
A
1w 2w 3w 4w
40
BMP-2 (pg/sample)
35 30
600
ALP activity (nmol/min/sample)
45
25 20 15 10
A
1w 2w 3w 4w
500 400 300 200 100
5 0
0 Ap
B
1w 2w 3w 4w
40
BMP-2 (pg/sample)
35 30
Ap
DLp-Ap
600
ALP activity (nmol/min/sample)
45
D-Ap
25 20 15 10
B
D-Ap
DLp-Ap
D-Ap
DLp-Ap
1w 2w 3w 4w
500 400 300 200 100
5 0
0 Ap
D-Ap
DLp-Ap
Ap
Fig. 5. BMP-2 protein concentration of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
Fig. 6. ALP activity of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
with various concentrations of plasmid (0–5 μg/mL) and lipid transfection reagent FuGENE® (0–10 μL/mL) (Figs. 1–3). The DLp-Ap layer is likely to be formed via the following process. Sintered HA particles at the surface of the disk and scaffold induce secondary nucleation of HA on their surfaces in the coating solution, as the solution is highly supersaturated with respect to HA. The HA nuclei grow into flakes, merged, and finally formed a continuous and homogeneous layer. During the spontaneous growth of HA crystals on the HA disk and HA scaffold surfaces, plasmid and lipid (FuGENE®), in the form of either discrete molecule or complex, are adsorbed onto the growing crystal surface through electrostatic interaction, taking into account our previous reports on protein-apatite composite layers [25–33]. The obviously decreased HA crystal size of the DLp-Ap layer (Fig. 1, D) than that of the Ap layer (Fig. 1, B) is probably due to the surface adsorption of plasmid and lipid during the coating process, which inhibits crystal growth of HA. As a result of such a coprecipitation process, DNA and lipid are immobilized in the DLp-Ap layer, compared with a simple adsorption process on the Ap layer surface, although we failed to quantify the amount of DNA and lipid in the layers. The present gene transfer system using the D-Ap and DLp-Ap layers was effective in not only in vitro gene transfer to cell line as was reported previously [21] but also in vitro and in vivo gene transfer to MSCs (stem cells, primary cells). It should be noted that the gene transfer efficiency to stem/primary cells is generally low compared with that to cell lines. MSCs are multipotent stem cells that can differentiate into a variety of cell types, including chondrocytes, osteoblasts, adipocytes, fibroblasts, marrow stroma cells, and other cells
of mesenchymal origin. MSCs have enormous therapeutic potential for tissue repair owing to their wide source in bone marrow and high capacity for self-renewal while maintaining multipotency [7–10,34]. Although any therapeutic effect of the MSCs seeded on the scaffold was unapparent in this animal model, the present gene transfer system using the D-Ap and DLp-Ap layers, with potential to control behavior of MSCs, would be useful in tissue engineering. Potential of the present gene transfer system to control cell behavior was demonstrated for cells under skin (Figs. 5–7, A) and MSCs (Figs. 5–7, B) by the in vivo results. The in vivo results showed that immobilization of BMP-2 gene into the Ap layer (resulting in the D-Ap and DLp-Ap layers) increased BMP-2 protein concentration (Fig. 5), thereby increasing ALP activity and OCN protein concentration (Figs. 6 and 7). These findings indicate that BMP-2 expressed by gene transfer from the D-Ap and DLp-Ap layers retained its biological activity and enhanced osteogenic differentiation of these cells adhering to the layers in vivo. Generally, non-viral gene transfer systems are less efficient than viral systems, and have difficulty in attaining gene expression at therapeutically effective level in vivo. The present gene transfer system using the D-Ap and DLp-Ap layers is a non-viral system effective in enhancing cell differentiation in vivo. Therefore, the present gene transfer system has advantages of high efficiency in addition to good biocompatibility and bone-bonding ability owing to the HA matrix. The high efficiency should be due to the sustained release of gene from the D-Ap and DLp-Ap layers, which acts to increase the local concentration of gene and/or lipid within the vicinity of the cell [4,11,35]. It is hypothesized that the sustained gene release is mediated by the gradual dissolution of the HA matrix
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
3.0
A
OCN (ng/sample)
DLp-Ap layer and the D-Ap layer as a control was tested in vitro and in vivo using MSCs and fibroblastic NIH3T3/cells under skin. The DLp-Ap layers showed 1–2 orders of magnitudes higher gene transfer efficiency to MSCs than the D-Ap layer in vitro. The DLp-Ap layer slightly increased BMP-2 protein concentration than D-Ap layer, thereby enhancing their osteogenic differentiation than D-Ap layer in vivo. The present gene transfer system is a non-viral system with the advantages of good biocompatibility, bone-bonding ability, and the potential to control cell differentiation of MSCs, hence it would be useful in bone tissue engineering.
1w 2w 3w 4w
2.5
2.0
1.5
1.0
Acknowledgment
0.5
0.0 Ap
3.0
B
D-Ap
DLp-Ap
1w 2w 3w 4w
2.5
OCN (ng/sample)
517
2.0
References
1.5
1.0
0.5
0.0 Ap
We thank Dr. Hajime Ohgushi, Dr. Mika Tadokoro, Dr. Koji Hattori, Dr. Youichi Katou and Mr. Yushin Yazaki from the National Institute of Advanced Industrial Science and Technology (AIST) for technical assistance. This study was supported in part by KAKENHI (Grant-in-Aid for Young Scientists B; 22700499) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan.
D-Ap
DLp-Ap
Fig. 7. OCN protein concentration of cells under skin (A) and MSCs (B) on the Ap, D-Ap, and DLp-Ap layers after implantation for 1, 2, 3, and 4 weeks.
serving as a reservoir for gene. This hypothesis is supported by our previous reports that the gene release is accompanied by partial dissolution of the HA matrix and this plays a critical role in controlling gene transfer efficiency and timing on the layer [36]. Immobilization of lipid transfection reagent FuGENE® to the D-Ap layer (resulting in the DLp-Ap layer) remarkably improved efficiency of gene transfer in vitro, although this improvement was less noticeable in vivo. The cells grow in a much more complex environment in vivo than in vitro. 1) Many factors influence in vivo cellular response, e.g. a complex extracellular matrix, soluble bioactive factors, and products of homo- and hetero-typical cell–cell interactions. 2) The cells grow in a circulation environment in vivo, thus the DNA concentration in vivo is lower than the still case in vitro. 3) The serum in vivo generally reduces the gene transfer efficiency of conventional lipids due to the interaction of serum components with the DNA–lipid complexes, which causes disintegration of the complexes and degradation of DNA molecules [37–40]. 4) Inflammatory response occurs immediately to implantation in vivo, which is different with in vitro case. Therefore, the optimized gene transfection condition obtained based on the in vitro results may not apply to the in vivo case. The in vivo gene transfection condition should be further optimized to prepare more effective composite apatite layer for their in vivo applications. 5. Conclusions DLp-Ap and D-Ap layers were prepared using supersaturated calcium phosphate solutions supplemented with 0.5–5 μg/mL plasmid and 0–10 μL/mL FuGENE®. The gene transfer efficiency of the
[1] R. Langer, D.A. Tirrell, Nature 428 (2004) 487–492. [2] S. Mitragotri, J. Lahann, Nat. Mater. 8 (2009) 15–23. [3] D.S. Benoit, M.P. Schwartz, A.R. Durney, K.S. Anseth, Nat. Mater. 7 (2008) 816–823. [4] L.D. Shea, E. Smiley, J. Bonadio, D.J. Mooney, Nat. Biotechnol. 17 (1999) 551–554. [5] Y.C. Huang, C. Simmons, D. Kaigler, K.G. Rice, D.J. Mooney, Gene Ther. 12 (2005) 418–426. [6] J. Fang, Y.Y. Zhu, E. Smiley, J. Bonadio, J.P. Rouleau, S.A. Goldstein, et al., Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 5753–5758. [7] H. Hosseinkhani, T. Azzam, H. Kobayashi, Y. Hiraoka, H. Shimokawa, A.J. Domb, et al., Biomaterials 27 (2006) 4269–4278. [8] H. Hosseinkhani, M. Hosseinkhani, F. Tian, H. Kobayashi, Y. Tabata, Biomaterials 27 (2006) 5089–5098. [9] H. Hosseinkhani, M. Hosseinkhani, F. Tian, H. Kobayashi, Y. Tabata, Biomaterials 27 (2006) 4079–4086. [10] H. Hosseinkhani, M. Yamamoto, Y. Inatsugu, Y. Hiraoka, S. Inoue, H. Shimokawa, et al., Biomaterials 27 (2006) 1387–1398. [11] J. Bonadio, E. Smiley, P. Patil, S. Goldstein, Nat. Med. 5 (1999) 753–759. [12] D. Lew, S.E. Parker, T. Latimer, A.M. Abai, A. Kuwahararundell, S.G. Doh, et al., Hum. Gene Ther. 6 (1995) 553–564. [13] S.E. Parker, H.L. Vahlsing, L.M. Serfilippi, C.L. Franklin, S.G. Doh, S.H. Gromkowski, et al., Hum. Gene Ther. 6 (1995) 575–590. [14] H. Shen, J. Tan, W.M. Saltzman, Nat. Mater. 3 (2004) 569–574. [15] B.B. Sun, K.K. Tran, H. Shen, Biomaterials 30 (2009) 6386–6393. [16] A. Oyane, M. Murayama, A. Yamazaki, Y. Sogo, A. Ito, H. Tsurushima, J. Biomed. Mater. Res. A 92 (2010) 1038–1047. [17] A. Oyane, H. Tsurushima, A. Ito, Gene Ther. 14 (2007) 1750–1753. [18] A. Oyane, H. Tsurushima, A. Ito, J. Gene Med. 12 (2010) 194–206. [19] X. Wang, A. Oyane, H. Tsurushima, Y. Sogo, X. Li, A. Ito, Biomed. Mater. 6 (2011) 045004. [20] D.H. Kohn, L.N. Luong, K.M. McFalls, Biomaterials 30 (2009) 6996–7004. [21] A. Oyane, Y. Yazaki, H. Araki, Y. Sogo, A. Ito, A. Yamazaki, H. Tsurushima, J. Mater. Sci. Mater. Med. 23 (2012) 1011–1019. [22] E.A. Wang, V. Rosen, J.S. Dalessandro, M. Bauduy, P. Cordes, T. Harada, et al., Proc. Natl. Acad. Sci. U. S. A. 87 (1990) 2220–2224. [23] H. Ohgushi, M. Okumura, S. Tamai, E.C. Shors, A.I. Caplan, J. Biomed. Mater. Res. 24 (1990) 1563–1570. [24] Y. Dohi, H. Ohgushi, S. Tabata, T. Yoshikawa, K. Dohi, T. Moriyama, J. Bone Miner. Res. 7 (1992) 1173–1180. [25] X. Wang, A. Ito, Y. Sogo, X. Li, H. Tsurushima, A. Oyane, Acta Biomater. 5 (2009) 2647–2656. [26] X. Li, X.P. Wang, A. Ito, Y. Sogo, K. Cheng, A. Oyane, Mater. Sci. Eng. C 29 (2009) 216–221. [27] A. Oyane, M. Uchida, K. Onuma, A. Ito, Biomaterials 27 (2006) 167–175. [28] A. Oyane, Y. Yokoyama, M. Kasahara, N. Ichinose, M. Saito, A. Ito, Mater. Sci. Eng. C 29 (2009) 1681–1686. [29] X. Wang, A. Ito, X. Li, Y. Sogo, A. Oyane, Biofabrication 3 (2011) 022001. [30] X.P. Wang, A. Ito, Y. Sogo, X. Li, A. Oyane, J. Biomed. Mater. Res. A 92A (2010) 1181–1189. [31] X.P. Wang, A. Ito, Y. Sogo, X. Li, A. Oyane, Acta Biomater. 6 (2010) 962–968. [32] X.P. Wang, X. Li, K. Onuma, A. Ito, Y. Sogo, K. Kosuge, et al., J. Mater. Chem. 20 (2010) 6437–6445. [33] A. Oyane, T. Ootsuka, K. Hayama, Y. Sogo, A. Ito, Acta Biomater. 7 (2011) 2969–2976. [34] G. Ciapetti, L. Ambrosio, G. Marletta, N. Baldini, A. Giunti, Biomaterials 27 (2006) 6150–6160. [35] J.M. Fang, Y.Y. Zhu, E. Smiley, J. Bonadio, J.P. Rouleau, S.A. Goldstein, et al., Proc. Natl. Acad. Sci. U. S. A. 93 (1996) 5753–5758.
518
X. Wang et al. / Materials Science and Engineering C 33 (2013) 512–518
[36] Y. Yazaki, A. Oyane, Y. Sogo, A. Ito, A. Yamazaki, H. Tsurushima, Biomaterials 32 (2011) 4896–4902. [37] J.P. Yang, L. Huang, Gene Ther. 4 (1997) 950–960. [38] J.P. Yang, L. Huang, Gene Ther. 5 (1998) 380–387.
[39] S. Li, W.C. Tseng, D.B. Stolz, S.P. Wu, S.C. Watkins, L. Huang, Gene Ther. 6 (1999) 585–594. [40] A. Oyane, Y. Yazaki, H. Araki, Y. Sogo, A. Ito, A. Yamazaki, et al., J. Mater. Sci. Mater. Med. 23 (2012) 1011–1019.