Effective delivery of bone morphogenetic protein 2 gene using chitosan–polyethylenimine nanoparticle to promote bone formation

Abstract

Treating bone defects is still a challenge in clinical practice. Recently, researchers used human bone morphogenetic protein 2 gene (hBMP-2) to induce osteoblast differentiation and promote new bone formation. However, an efficient way to deliver hBMP-2 still needs to be created. In this study, we evaluated whether chitosan–polyethylenimine (CS–PEI) nanoparticle could effectively deliver hBMP-2 locally with lower or no toxicity and promote osteoblast differentiation and new bone formation in vitro and in vivo. Data demonstrated that the synthesized CS–PEI/hBMP-2 nanoparticle at a W/W ratio of 20 to 1, which was the smallest size (162 nm) and highest zeta potential (24 mV), effectively transfected MC3T3-E1 cells without cytotoxicity in vitro, and had the ability to promote cell proliferation. Interestingly, the CS–PEI/hBMP-2 nanoparticle eliminated disadvantages of lower transfection efficiency from chitosan and cytotoxicity from PEI. RT-QPCR data showed that MC3T3-E1 cells treated with CS–PEI/hBMP-2 nanoparticle dramatically expressed higher levels of BMP-2 and significantly increased gene expressions of Col1 on days 3 and 14, Sp7 on days 3, 7, and 14, and ALP on day 14. Alizarin red staining demonstrated that CS–PEI/hBMP-2 nanoparticle-treated MC3T3-E1 cells significantly increased cell mineralization. These in vitro data suggest that the CS–PEI/hBMP-2 nanoparticle can effectively induce osteogenic differentiation of MC3T3-E1 cells in vitro. Western blot analysis further demonstrated that transgene BMP-2 indeed phosphorylated Smad1/5/8, which indicates that CS–PEI/hBMP-2 nanoparticle affects cell differentiation through a BMP-2 signal pathway. Importantly, in vivo data showed that CS–PEI/hBMP-2 nanoparticle clearly promoted new bone formation at the bone defect area 12 weeks post-implantation. This indicates that synthesized CS–PEI/hBMP-2 nanoparticle has the potential to become a useful therapeutic vector for bone defect treatment with further modification.

---

1. Introduction

Bone defect is a very common medical situation. Many complicated conditions, such as pathological fractures and large bone defects, can cause the healing process to fail. Different therapeutic treatments, such as natural grafts, stem cells, tissue engineering, and growth factor stimulation, have their own therapeutic limitations. Therefore, each treatment of bone defect still need to improve in order to quickly and efficiently heal bone defect.

Human bone morphogenetic protein 2 (BMP-2) is an autocrine secretion protein, a member of the TGF-β superfamily, and plays an important role in the development of bone.^1,2 And BMP-2 is an essential osteoinductive growth factor for the bone regeneration process, which can induce osteogenic differentiation of mesenchymal cells, healing the critical size of bone defect.^1–4 BMP-2 forms a complex with type I and II serine/threonine kinase receptors, in which phosphorylate receptor-mediated Smad1, 5, and 8 proteins form complexes with Smad4 protein, and further translocate to nucleus in order to regulate gene expressions related to bone regeneration.^5–7 The FDA has approved use of human BMP-2 as a growth factor to stimulate bone regeneration clinically. In general, recombinant BMP-2 protein can be used directly at the bone defect area. Frequent application at the bone defect area may be required and is expensive and inconvenient. Many bone defects need a long time to heal. Therefore, a slow-release method could stimulate and improve the healing process of bone defect.

Gene transfer or gene therapy could deliver hBMP2-gene to the bone defect area and transduce local cells to continually produce hBMP-2 within a certain period of time. There are many viral or non-viral delivery systems that can be used in bone defect treatment.^8–10 Most systems, however, still do not efficiently transfect or transduce cells on the bone defect area.^11–13 Chitosan has the potential to act as a polycationic gene carrier with high biocompatibility and no toxicity profile.¹⁴ Chitosan also has the ability to bind and protect DNA from nuclease degradation.¹⁵ The disadvantage of chitosan is low transfection efficiency.¹⁶ Polyethylenimine (PEI), known for its “proton sponge effect”, has been demonstrated to be a useful non-viral carrier in vitro and in vivo for some applications.^17,18 PEI, however, has its own disadvantages, e.g., cytotoxicity and non-degradability,^18,19 the toxicity of PEI can decrease with molecular weight decrease but its transfection efficiency will also decrease.²⁰ Several previous studies tried to modify chitosan and PEI to overcome their disadvantages.^21–24 Further improvement is still required to effectively deliver therapeutic genes to a bone defect location.

In our current study, we created CS–PEI nanoparticles using chitosan and 1.8 kDa of PEI to carry the hBMP-2 gene. Our data demonstrated that CS–PEI could effectively carry hBMP-2 and transfect MC3T3-E1 cells resulting in MC3T3-E1 differentiation in vitro and local cells in the bone defect area, thus resulting in increasing new bone formation in vivo.

2. Materials and methods

2.1 Synthesis of CS–PEI nanoparticles

CS–PEI nanoparticles were synthesized according to a previous study.²⁵ Briefly, 0.1 M periodate-oxidized chitosan (Sigma-Aldrich, St. Louis, USA) and 0.01 M potassium periodate (Sigma-Aldrich) were dissolved in 1% sodium acetate (pH 4.5) separately and degassed with N₂ for 30 min. Then, an equal volume of each solution was mixed together at 4 °C for 48 h and the reaction stopped by adding ethylene glycol (10% v/v). This solution was dialyzed with 10 liters of NaCl (0.2 M, pH 4.5) for 5 days, then dialyzed in deionized water (pH 4.5) for 2 days. Twenty mmol 1.8 kDa of PEI (PEI, pH 4.5, Aladdin, Shanghai, China) was added into this solution, degassed with N₂ for 30 min, and continuously mixed by magnetic stirring at 4 °C for 2 days. Thereafter, the solution was neutralized by adding 0.1 M sodium borohydride, magnetically stirred for 30 min, and then dialyzed in the same way as mentioned above (Fig. 1a).

Fig. 1 Synthesis and characteristics of CS–PEI nanoparticles. (a) Reaction scheme of CS–PEI copolymer; (b) agarose gel electrophoresis of CS–PEI/hBMP-2 nanoparticles at various W/W ratios; (c) zeta potential of CS–PEI/hBMP-2 at various W/W ratios; (d) particle size of CS–PEI/hBMP-2 at various W/W ratios. Data are represented as means ± SD from three experiments. pDNA, plasmid pACCMV–hBMP-2.

2.2 Characteristics of CS–PEI/hBMP-2 nanoparticles

The CS–PEI/hBMP-2 nanoparticles were prepared by mixing hBMP-2 plasmid DNA (pACCMV–hBMP-2 with CS–PEI) and then gently mixing the solution by vortex. The mixture was then kept at room temperature for 30 min.

Particle size and zeta potential of CS–PEI/hBMP-2 nanoparticles were evaluated by dynamic light scattering (Malvern Zetasizer Nano Z, UK) in triplicate. To do this, samples were placed into an analyzer chamber and measured in water. Particle size was assessed with three cycles, and zeta potential was performed by three repeated cycles with 100 runs each. CS–PEI and hBMP-2 at various weight/weight (W/W) ratios of 1 : 1, 5 : 1, 10 : 1, 15 : 1, 20 : 1, and 25 : 1 were instantly prepared by mixing equal volumes of CS–PEI and DNA diluted with deionized water.

Agarose gel retardation assay was performed using CS–PEI/hBMP-2 nanoparticles at different W/W ratios of 0.1 : 1, 0.2 : 1, 0.5 : 1, 1 : 1, and 2 : 1. Each sample aliquot was 10 μL plus 2 μL of loading buffer (6×); gel electrophoresis was run at 100 V for 20 min.

2.3 Cell culture

293T (human kidney cell line, ATCC, Manassas, VA, USA), Hela (human cervical carcinoma cell line, ATCC), and MC3T3-E1 (mouse pre-osteoblast cell line, Cell Bank of the Chinese Academy of Sciences, Shanghai, China) were cultured in Dulbecco's modified Eagle medium (DMEM, Gibco BRL, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS, Gibco BRL), 100 U mL⁻¹ penicillin and 100 μg mL⁻¹ streptomycin (Gibco BRL) at 37 °C in 5% CO₂ humidified atmosphere.

2.4 Cell cytotoxicity assays

Cell viability, cell cycle, and cell apoptosis were measured to evaluate in vitro cytotoxicity of the CS–PEI nanoparticles. Cell viability was performed with an MTT assay (AMRESCO, Solon, OH, USA) using 293T and MC3T3-E1 cell lines. Cell apoptosis and cell cycle were performed using an Annexin V-FITC/PI Double Staining Assay Kit (KeyGen, Nanjing, China). Cells were seeded in a 96-well plate at 2 × 10⁴ cells per well and cultured for 24 h, then 5, 10, 20, 50, and 100 μg mL⁻¹ of EGFP (EGFP plasmid, pACCMV–EGFP), CS/EGFP, PEI/EGFP, or CS–PEI/EGFP was added into the corresponding well to culture for 24 hours, after which time point effects of cytotoxicity were clearly indicated. Cells could float and cause inaccurate data if allowed to continue to culture beyond 24 h post-treatment. Then, we followed instructions from the MTT kit, cell cycle, or the apoptosis kit to carry out assays.

2.5 Transfection efficiency assays

Next, we needed to know if the CS–PEI nanoparticle plus plasmid DNA could transfect any cells. 293T cell is a gold standard cell line to compare or test any transfection/transduction reagent and evaluate transfection/transduction efficiency. MC3T3-E1 cell is a special cell line to study osteoblast differentiation in vitro. Therefore, to evaluate CS–PEI nanoparticle transfection efficiency, we counted the EGFP positive 293T or MC3T3-E1 cells using confocal laser scanning microscopy and FACS. Cells were seeded in 6-well plates at 20 × 10⁴ cells per well for 24 h, the medium was replaced with serum-free medium containing EGFP only, CS/EGFP, PEI/EGFP or CS–PEI/EGFP nanoparticles at W/W ratio of 20 to 1 to culture for 6 h, then the medium was replaced with normal growth medium to culture for 36 h. Transfection efficiency was then evaluated under confocal laser scanning microscopy or FACS.

2.6 Osteogenic differentiation

In this study, our purpose was to improve hBMP-2 local delivery resulting in promotion of bone defect healing. Therefore, it was important to know if the CS–PEI/BMP-2 nanoparticles could induce osteogenetic differentiation in vitro and in vivo. Osteogenetic differentiation induced by CS–PEI/BMP-2 nanoparticles was performed using MC3T3-E1 cells. MC3T3-E1 cells were seeded in a 6-well plate at 20 × 10⁴ cells per well, then transfected as described in Section 2.5. After transfection, cells were cultured in H-DMEM medium (Gibco BRL) containing 20 mM β-glycerol phosphate and 0.5 μM ascorbic acid. Cells were transfected by CS–PEI/EGFP as a control group. Cells were stimulated by recombinant hBMP-2 (20 ng mL⁻¹) (R&D Systems, MN, USA) as a positive control. After 3, 7, and 14 days post-transfection, cells were collected for total RNA extraction using a Qiagen RNeasy Mini Purification Kit (Qiagen, Valencia, CA, USA). Reverse transcription was performed using an iScript™ cDNA Synthesis Kit (Takara Bio, Tokyo, Japan). Real-time PCR (qPCR) was run using a MxPro Mx3005P Real-Time PCR Detection System (Agilent Technologies, Santa Clara, CA, USA) with SYBR-Green Premix Ex Taq (Takara Bio). Primers used for qPCR were as follows:

β-actin, 5′-CATCCGTAAAGACCTCTATGCCAAC-3′ and 5′-ATGGAGCCACCGATCCACA-3′;

ALP, 5′-CTCAACACCAATGTAGCCAAGAATG-3′ and 5′-GGCAGCGGTTACTGTGGAGA-3′;

Col1, 5′-GACATGTTCAGCTTTGTGGACCTC-3′ and 5′-GGGACCCTTAGGCCATTGTGTA-3′;

Sp7, 5′AAGTTATGATGACGGGTCAGGTACA-3′ and 5′-AGAAATCTACGAGCAAGGTCTCCAC-3′.

To assess mineralization, cells were stained by Alizarin red at 21 days post-transfection, at which time point MC3T3-E1 cells commonly form visible calcium deposition in osteogenetic differentiation conditions in our lab. Briefly, cells were fixed in 95% ethanol at room temperature for 10 min, washed with PBS three times, then stained with 0.1% Alizarin red solution (pH 8.3) for 20 min, washed with PBS three times and soaked in PBS for 4 h to remove non-specific Alizarin red. Quantification of mineralization was completed by adding cetylpyridinium chloride CPC (Sigma-Aldrich), extraction, and measuring at 562 nm.

2.7 Western blot analysis

It is known that normally there is a trace of activation of Smad1/5/8 in osteoblast and, that Hela cells absolutely do not have any activation of Smad1/5/8 without BMP2 stimulation. Therefore, Hela cells were used to prove if the hBMP-2 we delivered could biologically induce phosphorylation of downstream Smad1/5/8.^26,27 For this assay, Hela cells were used. The transfection procedure was the same as described in Section 2.5. Cells were stimulated by recombinant hBMP-2 (20 ng mL⁻¹) (R&D Systems) as a positive control. Cells were lyzed with RIPA Lysis Buffer (Beyotime, China). Thirty μg of protein was loaded onto SDS-PAGE gels for western blots using rabbit polyclonal anti-rat phospho-Smad1 (Abcam, Cambridge, MA, USA) and mouse monoclonal anti-β-actin antibody (Cell Signaling Technology, Boston, MA, USA) for detection.

2.8 In vivo animal experiment

Our animal protocol was approved by the Animal Care and Use Committee of Jilin University, Changchun, People's Republic of China. Twenty male Wistar rats (∼200 g) were used in this study. Rat is a commonly used animal model for studying critical-size defects. Advantages of using rat as compared to mouse for cranial bone defect are the size and ease for conducting craniotomies. All animal models have their limitations, but this was the best option currently available. Therefore, we used the rat model for the current study. To create the bone defect animal model,²⁸ rats were anesthetized using ketamine (60 mg kg⁻¹) and xylazine (8 mg kg⁻¹), the local area was shaved and sterilized with iodophor solution, a 20 mm length incision was made along the midline in the cranial parietal, and the calvarias area was exposed. Two critical bone defects (diameter 6 mm) (Fig. 5a) were created on the midline with a bone trephine (3i Implant Innovation, Palm Beach Gardens, FL, USA). For vivo experiments, CS–PEI/hBMP-2 nanoparticles were created with 200 μg of CS–PEI and 10 μg of hBMP-2 at W/W ratio of 20 : 1, which were absorbed into a 5 mm diameter, 1.5 mm thick gelatin sponge. Gelatin sponge loaded with 10 μg recombinant hBMP-2 protein or a gelatin sponge loaded with CS–PEI/hBMP-2 nanoparticles was placed in the bone defect area (Fig. 5a). Treated rats were sacrificed at 6 and 12 weeks after implantation and exsanguinated via auricular dextra and perfused via ventriculus dexter with 4% paraformaldehyde solution. The cranial parietal, heart, liver, spleen, and kidney were removed and fixed in 4% paraformaldehyde solution for further examinations. Micro computed tomography (micro-CT) (micro-CT 35; Scanco Medical AG, Bassersdorf, Switzerland) was used to evaluate new bone formation. Quantification of new bone formation was calculated using Image-Pro Plus software. After imaging detection, decalcification of these samples was performed by fixing in 10% EDTA solution, replacing the decalcification solution every week for x months, and then the samples were examined using additional histological methods.

2.9 Statistical analyses

The data are presented as mean ± standard deviation (SD). One-way ANOVAs, followed by a Turkey's test were used. P value <0.05 was considered statistically significant.

3. Results

3.1 Characteristics of CS–PEI/hBMP-2 nanoparticles

Results of electrophoresis shows that migration of CS–PEI/hBMP-2 nanoparticles was retarded completely when the W/W ratio reached 2 (Fig. 1b), which suggests that a W/W ratio of CS–PEI/hBMP-2 beyond 2 was sufficient for our study. Next, data in Fig. 1c shows that the size of CS–PEI/BMP2 nanoparticles quickly decreased between W/W ratios of 1 and 5, then gradually decreased with an increase of the W/W ratio (Fig. 1c). The size of CS–PEI/hBMP-2 nanoparticles was smallest at W/W ratios of 20 or 25, 162 ± 7.9 nm (Fig. 1c). Lastly, data from the zeta potential assay shows that the zeta potential of CS–PEI/BMP2 nanoparticles quickly increased between W/W ratios of 1 and 5, then basically reached a plateau with an increase of the W/W ratio (Fig. 1d). Zeta potential of CS–PEI/hBMP-2 nanoparticles was the highest at W/W ratios of 15, 20, or 25, 24 ± 2.9 mV (Fig. 1d). These data indicate that CS–PEI/BMP2 nanoparticles at a W/W ratio of 20 possess the smallest size and highest zeta potential, which indicates that the CS–PEI/BMP2 nanoparticle is electrically stabilized and resists aggregation. Therefore, we decided to use a W/W ratio of 20 in all further in vitro and in vivo experiments.

3.2 Cytotoxicity of CS–PEI in vitro

Data from MTT assays showed that there was no cytotoxicity of CS and CS–PEI nanoparticles even at 100 μg mL⁻¹ in either MC3T3-E1 or 293T cells (Fig. 2a and b). Interestingly, all amounts of CS–PEI nanoparticles tested herein slightly stimulated both MC3T3-E1 (∼120%) and 293T cell (∼120–140%) proliferation (Fig. 2a and b). Importantly, data also showed that PEI only caused severe cytotoxicity for MC3TC-E1 and 293T cells (Fig. 2a and b). Cell viability was dramatically decreased even using 5 μg mL⁻¹ of PEI, 79% for MC3T3-E1 cells, and 85% for 293T cells (Fig. 2a and b). When using 20 μg mL⁻¹ of PEI, cell viability was only ∼30% for both MC3T3-E1 and 293T cells (Fig. 2a and b). Further apoptosis and cell cycle assays showed that the CS–PEI nanoparticles did not influence apoptosis and cell cycle of MC3T3-E1 cells (Fig. 2c and d). Interestingly, these data suggest that CS–PEI nanoparticles overcome toxicity effects of PEI on cells and possibly retain the ability of CS to slightly stimulate cell proliferation.

Fig. 2 Cytotoxicity of CS–PEI nanoparticles in vitro. (a) MC3T3-E1 cell viability time after 24 h post-treatment with CS, PEI, or CS–PEI nanoparticles; (b) 293T cell viability after 24 h post-treatment with CS, PEI, or CS–PEI nanoparticles; (c) MC3T3-E1 cell apoptosis after 24 h post-treatment with CS–PEI nanoparticles; (d) MC3T3-E1 cell cycle assay after 24 h post-treatment with CS–PEI nanoparticles. Data are represented as means ± SD from three experiments.

3.3 Transfection efficiency of CS–PEI/EGFP nanoparticle

Data from transfection assays demonstrated that CS/EGFP had very low transfection efficiency: ∼5% in 293T cells and ∼0.5% in MC3T3-E1 cells (Fig. 3a, e, d and h). Transfection efficiency of PEI/EGFP was slightly, but significantly higher, than that of CS/EGFP (Fig. 3b, e, d and h). Interestingly, transfection efficiency of CS–PEI/EGFP nanoparticles was dramatically increased compared to the CS/EGFP or PEI/EGFP group in both ME3T3-E1 and 293T cells (Fig. 3c, f, d and h). These data indicate that the transfection efficiency increased when we used a W/W ratio of 20 to 1 to create CS–PEI/plasmid DNA nanoparticles.

Fig. 3 Transfection efficiency of CS/EGFP, PEI/EGFP, and CS–PEI/EGFP nanoparticles. (a) CS/EGFP in 293T cells; (b) PEI/EGFP in 293T cells; (c) CS–PEI/EGFP in 293T cells; (d) FACS data of 293T cells; (e) CS/EGFP in MC3T3-E1 cells; (f) PEI/EGFP in MC3T3-E1 cells; (g) CS–PEI/EGFP in MC3T3-E1 cells; and (h) FACS data of MC3T3-E1 cells. Data are represented as means ± SD from three experiments. pDNA, plasmid pACCMV–EGFP.

3.4 Osteogenic differentiation in vitro

Data from alizarin red staining showed that MC3T3-E1 cells treated with CS–PEI/hBMP-2 nanoparticles and protein BMP-2 had significantly more dark red staining than that of the control group (Fig. 4a and b) (P < 0.01). These results indicate that CS–PEI/hBMP-2 nanoparticle-treated MC3T3-E1 cells differentiate into osteoblast with calcium deposition.

Fig. 4 Effects of CS–PEI/hBMP2 nanoparticles on osteogenic differentiation. (a) Alizarin red staining of MC3T3-E1 cells after 21 days post-transfection; (c) semi quantificational analysis with CPC extraction; (b) relative gene expression of hBMP-2, Sp7, Col1 and ALP in MC3T3-E1 cells after 3, 7, or 14 days post-transfection with CS–PEI/EGFP as control group and CS–PEI/hBMP2; (d) phosphorylation of Smad1/5/8 in Hela cells after 48 h post-transfection with CS–PEI/EGFP, CS–PEI/hBMP-2 or recombinant hBMP-2 from four experiments. Data are represented as means ± SD from two experiments. *: P < 0.05, **: P < 0.01.

To further understand osteogenic differentiation, gene expression of BMP-2, Sp7, Col1, and ALP were evaluated. Fig. 4 shows MC3T3-E1 cells treated with CS–PEI/hBMP-2 nanoparticles significantly expressed higher levels of hBMP-2, Sp7, Col1, and ALP at almost all three time points, days 3, 7, and 14 except Col1 on day 7 and ALP on day 3 (Fig. 4b). Interestingly, ALP expression was dramatically higher than that of the control group on day 14 (Fig. 4b). The recombinant hBMP-2 treated group didn't affect hBMP-2 expression, could increase gene expressions of Sp7, Col1, and ALP at an early time point, and had less effect on day 14 (Fig. 4b), which indicates that CS–PEI/hBMP-2 nanoparticles can retain gene expression of hBMP-2 for longer time, and induce MC3T3-E1 cell differentiation to become osteoblast in vitro.

To understand if delivered hBMP-2 biologically affects cells through the BMP-2 signal pathway, phosphorylation of Smad1/5/8 was checked in Hela cells. Phosphorylation of Smad1/5/8 is a downstream signaling pathway of hBMP-2.²⁹ Phosphorylation of Smad1/5/8 is the symbolic signal which indicates that BMP is involved in specific biologic effects.³⁰ Indeed, western blotting analysis showed that Hela cells treated with CS–PEI/hBMP-2 nanoparticles or recombinant protein or from CS–PEI/hBMP-2 had a clear positive band for phosphorylation of Smad1/5/8 (Fig. 4d). This data indicates that both kinds of hBMP-2 can play a biological effect through the BMP-2 signal pathway.

3.5 Bone formation in vivo

Fig. 5a is a representative image of cranial bone defect used in this study. In vivo data showed that new bone formation (∼40% of new bone formation in the bone defect area) was not significantly different between either the gelatin sponge or the gelatin sponge loaded with CS–PEI/hBMP-2 nanoparticles or recombinant hBMP-2 6 weeks post-implantation (Fig. 5c, e and g). New bone formation, however, was significantly increased in the CS–PEI/hBMP-2 nanoparticle-treated rats or recombinant hBPM-2 (61.2% or 57.3% of new bone formation in the bone defect area) 12 weeks post-implantation than that of only gelatin sponge-treated rats (40.9% of new bone formation in the bone defect area, see yellow arrow) (Fig. 5, P < 0.05).

Fig. 5 Data from in vivo animal experiments. (a) A representative image for bone defect (diameter of 5 mm) on skull of rats; (b) quantitative analysis of micro CT new bone formation data in the bone defect site using Image-Pro Plus software; (c) image of micro CT analysis 6 weeks post-implantation with gelatin sponge; (d) image of micro CT analysis 12 weeks post-implantation with gelatin sponge; (e) image of micro CT analysis 6 weeks post-implantation with gelatin sponge loaded with recombinant hBMP-2; (f) image of micro CT analysis 12 weeks post-implantation with gelatin sponge loaded with recombinant hBMP-2; (g) image of micro CT analysis 6 weeks post-implantation with gelatin sponge loaded with CS–PEI/hBMP-2; (h) image of micro CT analysis 12 weeks post-implantation with gelatin sponge loaded with CS–PEI/hBMP-2; (i) H&E staining of rat cranial defect area 12 weeks post-implantation with gelatin sponge; (j) H&E staining of rat cranial defect area 12 weeks post-implantation with gelatin sponge loaded recombinant hBMP-2; and (k) H&E staining of rat cranial defect area 12 weeks post-implantation with gelatin sponge loaded with CS–PEI/hBMP2. There were 5 rats per group for in vivo experiments. Black arrows indicate the broken ends of fractured bone, yellow arrows indicate new bone, and red arrows indicate osteoblast.

Data from H&E staining further demonstrated that new bone formation was dramatically higher in the CS–PEI/hBMP-2 nanoparticle-treated bone defect area than in the gelatin sponges-treated bone defect area 12 weeks post-implantation (Fig. 5i, j and k). Cuboidal shaped osteoblasts were observed at the new bone area in the CS–PEI/BMP-2 nanoparticle-treated bone defect. Furthermore, there were no significant histological changes in the liver, spleen, kidney, or heart, which were obtained from the same rats 12 weeks post-implantation that received both gelatin sponge and gelatin sponge loaded with CS–PEI/hBMP-2 (Fig. 6). These results indicate that CS–PEI/hBMP-2 nanoparticles can also promote new bone formation at the local bone defect area in vivo without cytotoxicity in the liver, spleen, kidney, or heart.

Fig. 6 Histological examination of liver, spleen, kidney and heart 6 or 12 weeks post-implantation with gelatin sponge (a, e, i, m, c, g, k, o) or gelatin sponge loaded with CS-PEI/hBMP-2 (b, f, j, n, d, h, l, p) by H&E staining. There were 5 animals per group for these in vivo experiments.

4. Discussion

There are several ways to treat bone defect in current clinical practice. Most treatments, however, still need to be improved; development of efficient treatment is necessary. A delivery system of low toxicity, high efficiency, and good biodegradability is required to deliver BMP-2 locally. In this study, we created a novel nanoparticle, CS–PEI nanoparticle with CS and PEI to locally deliver a therapeutic gene, hBMP-2, in order to effectively treat bone defect.

DNA condensation is one of the necessary factors needed as a qualified gene delivery vector.³¹ Synthesized CS–PEI/hBMP-2 nanoparticle was retarded completely when the W/W ratio of CS–PEI/hBMP-2 reached 2 (Fig. 1b). This indicates that the W/W ratio of CS–PEI/hBMP-2 beyond 2 will meet requirements as a qualified gene vector. Surface properties of nanoparticles, such as particle size and zeta potential, are also important factors for a gene delivery vector because they can influence cell uptake, stability, and aggregation of nanoparticles.³⁰ In general, smaller nanoparticles can be more efficient to cross cell membrane, and higher zeta potential with stronger positive charges can increase cell binding and stability of nanoparticles and resists nanoparticle aggregation.³² The size of CS–PEI nanoparticles synthesized in this study clearly decreased with the increase of the W/W ratio of CS–PEI/hBMP-2 (Fig. 1c). At a W/W ratio of 20 to 1, the CS–PEI/hBMP-2 nanoparticles possessed the smallest particles (162 nm ± 9.7 nm) (Fig. 1c) and the highest zeta potential (24 mV ± 2.9 mV) (Fig. 1d), which provided effective ionic interactions and stronger positive charge to efficiently bind to anionic cell surfaces leading to active uptake by the cells.

293T cells are a gold standard cell line to compare different transfection reagents in the gene transfer field. Fig. 3 shows that CS–PEI/hBMP-2 nanoparticles are highly efficient to transfect 293T cells. The transfected positive cells are ∼90% if counted by direct observation (data not shown). Therefore, the transfection efficiency of CS–PEI/hBMP-2 nanoparticles can be compared with many commercial transfection reagents. MC3T3-E1 cells are a cell/tissue specific cell line, normally with expected relative lower transfection efficiency.

Chitosan and chitosan derivatives have the ability to effectively condense plasmid DNA, leading to protection DNA from degradation. Previous studies have used chitosan as a vector to deliver genes.^14,33,34 The transfection efficiency of chitosan, however, was lower in order to limit its application. Our study herein also demonstrated that transfection efficiency of chitosan was much lower than that of PEI in vitro (Fig. 3). On the other hand, chitosan caused no cytotoxicity whatever in both tested cell lines, 293T and MC3T3-E1 cells, and even slightly promoted both cell proliferations (Fig. 2).

PEI is one of the most common non-viral vectors because it is a cationic polymer. Its cationic property allows PEI to bind nucleic acids, like plasmid DNA, and attach to cell membrane resulting in efficiently delivering plasmid DNA into the cells. Branched PEI exhibits high transfection efficiency.^18,35,36 In general, low molecular weight PEI has less toxicity side effects.³¹ Therefore, we selected relatively low molecular weight PEI 1.8 kDa. Data in Fig. 3 show that PEI had significantly higher transfection efficiency than that of CS (Fig. 3). The disadvantage of PEI, however, is that it can cause severe cytotoxicity in both 293T and MC3T3-E1 cells (Fig. 2). Indeed, previous studies also found PEI's side effect of cytotoxicity through disturbing functions of cell membrane as well as interfering with intracellular processes of cells.^22,36–38

Our data clearly showed that the W/W ratio of 20 to 1 for the CS–PEI nanoparticle not only retained no cytotoxicity from chitosan (Fig. 2) and higher transfection efficiency from PEI, but also increased transfection efficiency compared to CS/EGFP or PEI/EGFP only (Fig. 3). These data suggest that we created a good biocompatibility nanoparticle which overcomes the lower transfection efficiency from chitosan and cytotoxicity from PEI.

Importantly, CS–PEI/hBMP-2 nanoparticles transfected MC3T3-E1 cells expressed with much higher hBMP-2 resulting in increased gene expressions of Col1 on days 3 and 14, Sp7 on days 3, 7, and 14, and ALP on days 7 and 14 through a BMP-2 signal pathway in vitro (Fig. 4). Interestingly, recombinant hBMP-2 increased gene expressions of Sp7, Col1, and ALP on day 3; these effects gradually decreased, and had much less effects on day 14 (Fig. 4), which indicated that the CS–PEI/hBMP-2 can mediate sustained hBMP-2 synthesis. BMP-2 protein is a secreted protein which can induce osteogenic differentiation though direct effects and paracrine way to influence adjacent neighbor cells.^39,40 It is known that Col1 is a major protein for bone formation and repair.⁴¹ Sp7 is a transcription factor and an important indicator for osteogenic differentiation.⁴² ALP activity is another powerful indicator for osteogenic differentiation.⁴³ Alizarin red staining is a useful biochemical assay to quantitatively determine calcific deposition, an important step towards formation of calcified extracellular matrix to convert real bone. The increased hBMP-2 herein clearly induced osteogenic differentiation and calcium deposition increase in the MC3T3-E1 cells (Fig. 4). This indicates that CS–PEI/hBMP-2 nanoparticles effectively deliver hBMP-2 into MC3T3-E1 cells and induce MC3T3-E1 cells to differentiate into osteoblast in vitro.

It is well-known that BMP-2 can induce stem cell homing and differentiation of osteoblasts, and even induce ectopic ossification.^1,30 In our current study, new bone formation was significantly increased at the bone defect area of CS–PEI/hBMP-2 nanoparticle-treated rats 12 weeks post-implantation without any observable histological change in liver, spleen, kidney, or heart (Fig. 5). Our in vitro data indicate that chitosan has the potential to stimulate cell proliferation (Fig. 2). As seen in Fig. 5, we also noticed that there were some cuboidal shaped osteoblasts around the new bone area 12 weeks post-implantation with CS–PEI/BMP-2 nanoparticles (Fig. 5). Taken together, our data suggest that with CS–PEI/BMP-2 nanoparticles it is possible to directly promote osteoblast proliferation and also transfect local stem cells/osteoblast cells to promote new bone formation although more experiments need to be performed in order to understand in vivo mechanisms directly.

Our in vivo data (Fig. 5) showed that that new bone formation (∼40% of new bone formation in the bone defect area) was not significantly different between gelatin sponge and gelatin sponge loaded with CS–PEI/hBMP-2 nanoparticles or recombinant hBMP-2 6 weeks post-implantation, but was significantly different 12 weeks post-implantation. The reason for this delayed effect is not known. One possible reason may be due to two bone defects in each rat, which could cause a much higher burden to repair leading to delay healing. Second is that the size of the gelatin sponge may be not appropriate. The bigger gelatin sponge could cause early inflammatory responses and longer degradation time. On the other hand, the longer degradation time could meet a slow release requirement. Therefore, we will consider both factors to further modify our rat model and gelatin sponge to search for the optimum condition in future studies.

We understand that our current version of CS–PEI/hBMP-2 still needs further modification because it cannot fully repair bone defect on week 12. In this study, we used PEI 1.8 kDa, which has a much smaller molecular weight, to create CS–PEI/hBMP-2 nanoparticles with no cytotoxicity. Therefore, we will use different higher molecular weights of PEI to create CS–PEI/hBMP-2 in future experiments to evaluate if it still has no or lower cytotoxicity and significantly increases transfection efficiency.

5. Conclusion

Our newly synthesized CS–PEI/hBMP-2 nanoparticle, 162 nm at size with 24 mV of zeta potential, effectively transfects MC3T3-E1 cells in vitro without any cytotoxicity, resulting in induced MC3T3-E1 cell differentiation in vitro. Interestingly, CS–PEI/hBMP-2 nanoparticle does not retain disadvantages of lower transfection efficiency from chitosan and cytotoxicity from PEI. More importantly, the bone defect area is significantly decreased after 12 weeks post-implantation with CS–PEI/hBMP-2 nanoparticles. Our data suggest that CS–PEI/hBMP-2 nanoparticle has potential application in future bone defect treatment.

Acknowledgements

The authors would like to thank Cindy Clark (NIH Library Editing Service) for reviewing the manuscript. This study was supported by the National Natural Science Foundation of China (81320108011, 81271111, 30830108), the Graduate Innovation Fund of Jilin University (2015050), the Research Fund for the Doctoral Program of Higher Education of China (233200801830063, 20120061130010), and the Science Technology Program of Jilin Province (201201064).

References

1. D. Chen, M. Zhao and G. R. Mundy, Growth Factors, 2004, 22, 233–241.
2. W. Yang, M. A. Harris, Y. Cui, Y. Mishina, S. E. Harris and J. Gluhak-Heinrich, J. Dent. Res., 2012, 91, 58–64.
3. S. N. Lissenberg-Thunnissen, D. J. de Gorter, C. F. Sier and I. B. Schipper, Int. Orthop., 2011, 35, 1271–1280.
4. G. Luther, E. R. Wagner, G. Zhu, Q. Kang, Q. Luo, J. Lamplot, Y. Bi, X. Luo, J. Luo and C. Teven, Curr. Gene Ther., 2011, 11, 229–240.
5. Y. K. Wang, X. Yu, D. M. Cohen, M. A. Wozniak, M. T. Yang, L. Gao, J. Eyckmans and C. S. Chen, Stem Cells Dev., 2012, 21, 1176–1186.
6. X. Zhang, J. Guo, Y. Zhou and G. Wu, Tissue Eng., Part B, 2014, 20, 84–92.
7. H. M. Ryoo, M. H. Lee and Y. J. Kim, Gene, 2006, 366, 51–57.
8. C. Zhang, K.-z. Wang, H. Qiang, Y.-l. Tang, Q. Li, M. Li and X.-q. Dang, Acta Pharmacol. Sin., 2010, 31, 821–830.
9. V. Uskoković and D. P. Uskoković, J. Biomed. Mater. Res., Part B, 2011, 96, 152–191.
10. F. Wegman, R. E. Geuze, Y. J. Helm, F. Cumhur Öner, W. J. Dhert and J. Alblas, J. Tissue Eng. Regener. Med., 2014, 8, 763–770.
11. C. Wilson, F. Martín-Saavedra, N. Vilaboa and R. Franceschi, J. Dent. Res., 2013, 92, 409–417.
12. N. K. Bleich, I. Kallai, J. R. Lieberman, E. M. Schwarz, G. Pelled and D. Gazit, Adv. Drug Delivery Rev., 2012, 64, 1320–1330.
13. G. Pelled, A. Ben-Arav, C. Hock, D. G. Reynolds, C. Yazici, Y. Zilberman, Z. Gazit, H. Awad, D. Gazit and E. M. Schwarz, Tissue Eng., Part B, 2009, 16, 13–20.
14. S. Mao, W. Sun and T. Kissel, Adv. Drug Delivery Rev., 2010, 62, 12–27.
15. N. Duceppe and M. Tabrizian, Expert Opin. Drug Delivery, 2010, 7, 1191–1207.
16. R. Riva, H. Ragelle, A. des Rieux, N. Duhem, C. Jérôme and V. Préat, in Chitosan for Biomaterials II, Springer, 2011, pp. 19–44.
17. O. Boussif, F. Lezoualc'h, M. A. Zanta, M. D. Mergny, D. Scherman, B. Demeneix and J.-P. Behr, Proc. Natl. Acad. Sci. U. S. A., 1995, 92, 7297–7301.
18. S. Patnaik and K. C. Gupta, Expert Opin. Drug Delivery, 2013, 10, 215–228.
19. J.-L. Lee, C.-W. Lo, S.-M. Ka, A. Chen and W.-S. Chen, J. Controlled Release, 2012, 160, 64–71.
20. A. Beyerle, M. Irmler, J. Beckers, T. Kissel and T. Stoeger, Mol. Pharm., 2010, 7, 727–737.
21. D. Pezzoli, F. Olimpieri, C. Malloggi, S. Bertini, A. Volonterio and G. Candiani, PLoS One, 2012, 7, e34711.
22. F.-W. Huang, H.-Y. Wang, C. Li, H.-F. Wang, Y.-X. Sun, J. Feng, X.-Z. Zhang and R.-X. Zhuo, Acta Biomater., 2010, 6, 4285–4295.
23. W. He, Z. Guo, Y. Wen, Q. Wang, B. Xie, S. Zhu and Q. Wang, J. Biomater. Sci., Polym. Ed., 2012, 23, 315–331.
24. W. Wang, W. Nan, L. Sun and W. Liu, React. Funct. Polym., 2013, 73, 993–1000.
25. H.-L. Jiang, Y.-K. Kim, R. Arote, J.-W. Nah, M.-H. Cho, Y.-J. Choi, T. Akaike and C.-S. Cho, J. Controlled Release, 2007, 117, 273–280.
26. Y. B. Niu, X. H. Kong, Y. H. Li, L. Fan, Y. L. Pan, C. R. Li, X. L. Wu, T. L. Lu and Q. B. Mei, Mol. Med. Rep., 2015, 11, 4468–4472.
27. X. L. Cheng, H. Alborzinia, K. H. Merz, H. Steinbeisser, R. Mrowka, C. Scholl, I. Kitanovic, G. Eisenbrand and S. Wolfl, Chem. Biol., 2012, 19, 1423–1436.
28. H. Jin, K. Zhang, C. Qiao, A. Yuan, D. Li, L. Zhao, C. Shi, X. Xu, S. Ni and C. Zheng, Int. J. Nanomed., 2014, 9, 2179.
29. M. Okamoto, J. Murai, H. Yoshikawa and N. Tsumaki, J. Bone Miner. Res., 2006, 21, 1022–1033.
30. G. Chen, C. Deng and Y.-P. Li, Int. J. Biol. Sci., 2012, 8, 272.
31. A. Mann, G. Thakur, V. Shukla, A. K. Singh, R. Khanduri, R. Naik, Y. Jiang, N. Kalra, B. Dwarakanath and U. Langel, Mol. Pharm., 2011, 8, 1729–1741.
32. X. Lin, N. Zhao, P. Yan, H. Hu and F.-J. Xu, Acta Biomater., 2015, 11, 381–392.
33. T. Ishii, Y. Okahata and T. Sato, Biochim. Biophys. Acta, 2001, 1514, 51–64.
34. D. Lee, W. Zhang, S. A. Shirley, X. Kong, G. R. Hellermann, R. F. Lockey and S. S. Mohapatra, Pharm. Res., 2007, 24, 157–167.
35. B. Chen, M. Liu, L. Zhang, J. Huang, J. Yao and Z. Zhang, J. Mater. Chem., 2011, 21, 7736–7741.
36. K. Wong, G. Sun, X. Zhang, H. Dai, Y. Liu, C. He and K. W. Leong, Bioconjugate Chem., 2006, 17, 152–158.
37. Y. J. Choi, S. J. Kang, Y. J. Kim, Y.-b. Lim and H. W. Chung, Drug Chem. Toxicol., 2010, 33, 357–366.
38. A. Petri-Fink, B. Steitz, A. Finka, J. Salaklang and H. Hofmann, Eur. J. Pharm. Biopharm., 2008, 68, 129–137.
39. J. Park, J. Ries, K. Gelse, F. Kloss, K. Von Der Mark, J. Wiltfang, F. Neukam and H. Schneider, Gene Ther., 2003, 10, 1089–1098.
40. K. Riew, N. Wright, S.-L. Cheng, L. Avioli and J. Lou, Calcif. Tissue Int., 1998, 63, 357–360.
41. F. H. Zhou, B. K. Foster, G. Sander and C. J. Xian, Bone, 2004, 35, 1307–1315.
42. L. A. Kaback, D. Y. Soung, A. Naik, N. Smith, E. M. Schwarz, R. J. O'Keefe and H. Drissi, J. Cell. Physiol., 2008, 214, 173–182.
43. I. Avbersek-Luznik, T. Gmeiner Stopar and J. Marc, Clin. Chem. Lab. Med., 2007, 45, 1014–1018.

---