Advanced PLGA hybrid scaffold with a bioactive PDRN/BMP2 nanocomplex for angiogenesis and bone regeneration using human fetal MSCs.

Biodegradable polymers have been used with various systems for tissue engineering. Among them, poly(lactic-co-glycolic) acid (PLGA) has been widely used as a biomaterial for bone regeneration because of its great biocompatibility and biodegradability properties. However, there remain substantial cruxes that the by-products of PLGA result in an acidic environment at the implanting site, and the polymer has a weak mechanical property. In our previous study, magnesium hydroxide (MH) and bone extracellular matrix are combined with a PLGA scaffold (PME) to improve anti-inflammation and mechanical properties and osteoconductivity. In the present study, the development of a bioactive nanocomplex (NC) formed along with polydeoxyribonucleotide and bone morphogenetic protein 2 (BMP2) provides synergistic abilities in angiogenesis and bone regeneration. This PME hybrid scaffold immobilized with NC (PME/NC) achieves outstanding performance in anti-inflammation, angiogenesis, and osteogenesis. Such an advanced PME/NC scaffold suggests an integrated bone graft substitute for bone regeneration.

INTRODUCTION

Although acute or small bone fractures can generally be cured without any surgical treatment, massive bone defects require orthopedic surgery, which presents huge challenges for patients and orthopedic surgeons (). Currently, there are three major methods of bone grafting: autologous, allogenous, and artificial scaffold grafts. Numerous studies have attempted to develop suitable synthetic materials for bone repair, including polymers (, ), hydrogels (–), ceramics (), and metal scaffolds (). The ideal bone graft biomaterial requires various capacities: osteoinductivity, osteoconductivity, proper mechanical property, antibacterial activity, and biocompatibility ().

Poly(lactic-co-glycolic) acid (PLGA) is one of the common biodegradable polymers for surgical implants. Although PLGA was approved by the U.S. Food and Drug Administration for clinical use (), there still remain substantial shortcomings, such as low mechanical property, low biocompatibility, and acidic by-products. PLGA has been reported that its degradation byproducts, lactic acid and glycolic acid during hydrolysis after implantation, cause an acidic microenvironment (). Amini et al. () and Washington et al. () reported that degraded lactic acid and glycolic acid cause inflammatory response at the implanting site and weaken the regenerative efficacy of the biomaterial. In our previous studies, various designs for biomedical devices containing magnesium hydroxide (MH) were proposed to overcome these drawbacks (–). Furthermore, as an extension of the study, we applied ricinoleic acid (RA), a biocompatible fatty acid, to create a modified MH (mMH) because it has two advantages, that of improving the organic solvent dispersibility of MH and synergistic effect in anti-inflammation (). Thus, we hypothesized that mMH would be appropriate for reducing inflammation during bone regeneration. In addition, the repair of critical-sized bone fractures is often disrupted by bacterial infections because of the large surgical area and long-term recovery period. Thus, recent approaches for developing bone implanting devices are attempted to give an antibacterial property (). The addition of mMH would enhance scaffolds with an antibacterial property (, ). Furthermore, to improve the osteoconductivity of the scaffold, a bovine inorganic extracellular matrix (bECM) consisting of mostly calcium and phosphate was opted.

With the design of a functionalized polymer-based scaffold for effective bone repair, many studies have emerged using prospective techniques such as bioactive molecule immobilization (–). However, it was said that bioactive molecules immobilized on the polymeric scaffold should have proper delivery systems because of several limitations, including rapid clearance, poor physiological stability, nonspecific targeting, and low permeability into the target tissue (). To solve these problems, we proposed a self-assembled nanocomplex (NC) immobilization system. The novel bioactive NC, which contains two different effective molecules, positively charged bone morphogenetic protein 2 (BMP2) and negatively charged polydeoxyribonucleotide (PDRN), can exert synergistic effects on bone regeneration. Here, the NC spontaneously sinks on the surface of the scaffold in calcium- and phosphate-rich environment by ionic interaction. By introducing natural physical phenomenon, these bioactive molecules can be effectively immobilized on the surface and induced controlled release.

Recent studies have suggested that angiogenesis performs an important role in bone regeneration. To address this issue, PDRN would be a considerable molecule, which enhances angiogenic effect during bone formation. Baek et al. () reported that PDRN has a brilliant ability for angiogenesis and plays as an agonist of adenosine A2A receptor. Consequently, it accelerates the down-regulation of acute inflammation and up-regulates angiogenic factors, such as vascular endothelial growth factor (VEGF) (–). PDRN is considered as an appropriate bioactive molecule for successfully regenerating a vascularized bone tissue. Meanwhile, bone morphogenetic proteins are predominant molecules that play an important role in many bone regeneration studies. Among them, BMP2 is estimated to be the most promising molecule for bone regeneration (). Although it has a great ability for bone repair, treatment with high dose of BMP2 has led to numerous side effects, such as cancer development, cytotoxicity, ectopic bone formation, and renal problems (, ). In many other studies, the BMP2 was incorporated in various scaffolds for bone regeneration. Walsh et al. () and Liu et al. () used 5 μg per scaffold and 400 ng per scaffold, respectively, whereas our PME/NC [PME hybrid scaffold immobilized with NC] scaffold has only 200 ng of recombinant human BMP2 per scaffold. In other words, BMP2 has side effects when used a lot; so, it is good to use as little as to achieve its effect, as in our scaffolds. The efficiency of NC immobilization to the scaffold can increase in supersaturated calcium-phosphate solution (Ca/P = 1.67). We hypothesize that this novel bioactive NC has a dual effect on angiogenesis and osteogenesis.

Moreover, mesenchymal stem cells (MSCs) are of great therapeutic potential for a variety of diseases because of their ability to self-renew and differentiate into multiple tissues. Among them, human fetal-derived MSCs (hfMSCs) correspond to the time of birth of human bone and show superior potential to osteogenic differentiation compared to adult MSCs (e.g., human bone marrow MSCs, human umbilical cord MSCs). It also displays great osteogenesis and lower immunogenicity, as compared to MSC from perinatal and postnatal sources (–). hfMSCs were seeded on an NC-immobilized hybrid scaffold to enhance bone regenerative efficacy. In this study, an advanced PLGA/mMH/bECM hybrid scaffold with NC (PME/NC) was prepared, and the great capacity for biocompatibility, inhibition of osteoclast differentiation, and osteogenic ability of hfMSCs was investigated in vitro. In addition, to evaluate the reduced inflammation response and angiogenic effect as well as bone regenerative efficacies of the PME/NC in vivo, 8 weeks of implantation into critical-size rat calvarial defect was performed and compared with other scaffolds (Fig. 1).

Fig. 1.

Schematic illustration of the advanced PME/NC.

First, the PLGA/mMH/bECM scaffold and PDRN/BMP2 NC were prepared, and subsequently, the NC was immobilized on PLGA/mMH/bECM scaffold in a saturated Ca/P solution to obtain a PME/NC scaffold. After implantation, the PME/NC scaffold was seeded with hfMSCs at the rat calvarial defect site, and it exhibited vascularization and bone repair.

RESULTS AND DISCUSSION

Synthesis and characterization of an advanced hybrid scaffold

In our previous work, an MH-incorporated PLGA scaffold was prepared and shown to enhance bone regeneration compared to untreated PLGA scaffold and the control group in vivo (). To design a functional hybrid scaffold for critical-sized bone fractures, numerous approaches adopt bioactive molecules such as BMP2, which was also applied in this study (–). BMP2 was known as an osteoinductive material for inducing the differentiation of mesenchymal stems or progenitor cells. However, there is a practical issue in that the growth factor would be damaged to the target tissue and has several side effects without an appropriate delivery system (). Here, a positively charged BMP2 was prepared using a self-assembled approach with negatively charged PDRN through a charge interaction to induce the sustained release of BMP2 as a form of the NC on the scaffold from the implantation site. Dynamic light scattering (DLS) analysis was subjected to characterize the size and zeta potential of NC with varying amounts of PDRN (Fig. 2A). BMP2 with a positively charged surface and zeta potential of +10.6 mV at pH 7.0 changed to a negative potential of −33.1 mV after its formation into NC. Moreover, the NC has a better size distribution (108.8 nm) at a ratio of 1:1. In particular, PDRN was selected for the development of a complex with BMP2 for the following reasons: (i) PDRN is an angiogenic polynucleotide biomolecule, which has a synergistic effect with BMP2 for bone regeneration (, ); (ii) it has enough charge to reverse positively charged BMP2 to a negative charge in NC; and (iii) it also displays considerable angiogenesis and a wound healing effect (–). After complexation, NC is still capable of staining by using a nucleic acid detection solution because of the unique property from PDRN, namely, DNA extracted from salmon (Fig. 2B). In addition, in Fig. 2C, to confirm the stable complexation of two molecules, circular dichroism (CD) analysis demonstrated a stable secondary structure. The spectrum of PDRN indicated a basic B-form conformation, which has a major positive band at 280 nm, a negative band at 248 nm, and a minor positive peak at 210 to 220 nm (). The PDRN and NC have the same peak. It was notable that PDRN kept the original B-form after binding at the surface of BMP2. In addition, the attenuated total reflection (ATR)–Fourier transform infrared (FTIR) result was analogous with the CD spectrum (fig. S1). The BMP2 has specific N-H stretching vibration peaks at 3200 cm−1, whereas the characteristic vibration peak of PDRN appeared at 1653 cm−1 (amide I band) and 1050 to 1220 cm−1 (phosphate stretching bands). As shown by the field emission–scanning electron microscopy (FE-SEM) image (Fig. 2D), the NC has a diameter from 100 to 120 nm, which is consistent with the results of the DLS, and the displayed star-shaped nanostructures were the same as those illustrated for the NC from transmission electron microscopy (TEM) (Fig. 2E). Figure 2F reveals that PDRN was stained using Hoechst because it kept its original B-form DNA structure after NC formation. On the basis of these imaging results, positively charged BMP2 was completely capped by negatively charged PDRN.

Fig. 2.

Optimization of NC fabrication.

(A) Mean size (a) and zeta potential (b) of NC in varying ratios. (B) Agarose gel electrophoresis retardation. (C) CD analysis. (D) FE-SEM image (a) and transmission electron microscopy (TEM) image (b). Scale bars, 200 and 100 μm, respectively. (E) Confocal microscopy image of NC stained by Hoechst (scale bar, 10 μm). Scaffold characterization. (F) Representative SEM images of the morphological structure of each scaffold (scale bars, 300 μm). (G) Changes in the pH of scaffolds during in vitro degradation in PBS at 37°C. (H) Mechanical properties of the scaffolds, representative stress-strain curve (a), and compressive modulus denoting 5 to 10% compressive stress/strain curves (b) (n = 3). (I) Bactericidal activity of the scaffolds using a colony-forming assay of E. coli and S. aureus on agar plates after 1 ml of bacteria solution incubation. #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate statistically significant differences, respectively. PM, PLGA scaffold containing MH.

Meanwhile, the biodegradable porous PLGA scaffold was opted to the main body for effective bone regeneration. In our previous studies, the porous PLGA scaffold containing MH exhibited an effective repair ability for the regeneration of various tissues (–, , , ). In particular, because mMH has a greater anti-inflammatory response and excellent dispersibility in an organic solvent, we opted to apply it to our new scaffold. Moreover, to enhance the NC immobilization capacity and osteoconductivity of the scaffold, bECM was introduced. SEM was conducted to confirm the cross-sectional morphology of the scaffolds (Fig. 2G). The image exhibited a well-distributed pore, high porosity, and interconnected pore structure that could induce cell migration, proliferation, and differentiation. To assess degradation-dependent pH changes (Fig. 2H) and mass loss (fig. S3), the scaffolds were immersed for 60 days in a phosphate-buffered saline (PBS) solution at 37°C. During the degradation process, the pH of the PLGA scaffold decreased to only 3.5 and fully degraded by day 60. However, in groups containing surface mMH, the pH increased to 8.4 and subsequently dropped to 6.0, as the scaffolds were decomposed for 60 days. These results imply that the mMH was effective for neutralizing the acidic environment of the implanting site caused by the acidic by-product of PLGA in vivo. The proportion of MH inorganic particles in the scaffolds was determined using thermogravimetric analysis (TGA; fig. S4) and inductively coupled plasma–optical emission spectroscopy (ICP-OES; table S1). The elemental distribution of the scaffold was verified by energy dispersive X-ray spectroscopy mapping (fig. S5). The Mg element of the MH-containing scaffold was uniformly distributed. In the PME/NC scaffold, the Ca and P elements were more abundant than in the PME scaffold, because the PDRN backbone contains phosphate groups, and the Ca2+ and PO42− ions were deposited onto the scaffold by a supersaturated calcium solution. The water contact angle was measured to evaluate the change in hydrophilicity. The angle decreased from 108° to 89° with the addition of mMH and bECM (fig. S6). The scaffold with NC was completely wet (0°), which indicated that NC was successfully immobilized onto the scaffolds. Hydrophilic scaffolds have a better capacity for initial cell adhesion in vitro and effectively penetrate the host cell in vivo. To investigate the mechanical properties of the scaffolds containing mMH and bECM, the compressive stress and modulus were measured using a universal testing machine (UTM) with 1 N cell. Scaffolds with different portions of bECM [30 and 50 weight % (wt %) compared to that of PLGA] showed an improved compressive modulus from 10.13 to 16.50 kPa in Fig. 2I. The robustness of the scaffold can prevent early degradation and ensure the sustainable release of mMH, PDRN, and BMP2.

It is known that bacterial infection occurs frequently when biomedical implants are used for critical-sized bone fractions (). Therefore, if the bone graft material has antibacterial effects, its regeneration effect would be enhanced, as the MH could supply a large number of hydroxide ions in an acidic environment (). On the basis of our previous study, positively charged Mg2+ ions are easily adsorbed to negatively charged bacterial surfaces by electrostatic interactions, which can destroy cell walls and induce bacterial death (). In a previous study, our group confirmed the antimicrobial activity of PLGA composite films containing mMH (). Figure 2I and fig. S7 show the bactericidal activity of each scaffold using the same amount of initial bacterial culture solution. The scaffolds containing mMH demonstrated almost 100% antimicrobial activity against both Gram-negative Escherichia coli and Gram-positive Staphylococcus aureus. These results suggest the possibility of not only bone repair but also superior antibacterial activity by using the multifunctional scaffold containing mMH.

As mentioned above, to intensively deposit the NC onto the scaffold, a negatively charged NC surface has the advantage of binding with the Ca2+ ion in a supersaturated calcium-phosphate (Ca/P) solution; thus, an increased amount of NC would be deposited onto the scaffold because of the calcium fertile environment (Fig. 3A). Figure 3B exemplified the morphology when the NC was immobilized on the surface of the scaffold by our immobilization system. As was confirmed previously, PDRN, which forms the outer layer of NC, is a DNA-derived molecule. Figure 3C represents an image of the NC stained with Hoechst using its DNA property, and it depicts NC as well distributed on the surface of the scaffold. It is well known that the importance of BMP2 retention in an artificial implant is considered the biggest obstacle because a high dose induces an abnormal bone structure and inflammation (, ). To make up for the longer retention time of BMP2, various delivery systems have been studied. Figure 3 (D and E) shows the release profile of BMP2 and PDRN from the hybrid scaffold after immobilization. Appreciably, while BMP2 was sustainedly released from the scaffold within 60 days in vitro, the release profile of PDRN from the scaffold showed that it was released for 28 days. Because PDRN was 100% released in 28 days, it was observed that BMP2 was only slightly released at 28 days. Accordingly, the retention time of BMP2, which has been pointed out as a limitation in existing studies (, , ), was successfully extended up to 60 days in vitro by using the bioactive molecule immobilization system on the scaffold. Moreover, it was followed by a sustained-release profile without initial bursts. These results indicated that the NC and its immobilization system provided an adequate strategy for the sustained release of the bioactive molecule, BMP2.

Fig. 3.

NC immobilization on the scaffold.

(A) Illustration of the process of NC immobilization by ionic interaction. (B) Surface morphology of immobilized NC on the scaffold using FE-SEM (scale bars, 2 and 1 μm). (C) Hoechst-stained images of NC on scaffold using confocal microscopy (scale bars, 100 and 50 μm). (D and E) In vitro BMP2 and PDRN release profiles from the NC-immobilized scaffold.

Angiogenic efficacy of the NC-immobilized hybrid scaffold using hfMSCs

To regenerate new functional bone, recent studies suggested the vascularization of repaired bone tissue as a further inspiring treatment (, , ). It was demonstrated that the NC has excellent angiogenic efficiency because of the synergistic effect of bioactive BMP2 and PDRN. Figure 4A shows the functional impact of the NC on cell migration using hfMSCs. The healed area increased for BMP2 and PDRN in 24 and 48 hours, compared to that of the control. By 24 hours, 75% of the scratched wound was healed for the NC-treated group, and it was approximately two times higher than the control group. In addition, the covered area of the NC-treated group was 1.8 times higher than that of the control group in 48 hours (41.6 and 74.9%, respectively). In the same way, in Fig. 4B, the angiogenesis ability of NC was investigated by tubule-forming assay with human umbilical vein endothelial cells (HUVECs). The total length and number of branching points substantially increased for the NC-treated group. To evaluate increased VEGF production due to PDRN, immunocytochemistry (ICC) was conducted with an anti–VEGF A (VEGFA) antibody using HUVECs (Fig. 4C). The PDRN- and NC-treated groups demonstrated an increase in the number of VEGF-positive cell observation compared to the control and BMP2-treated groups in 3 days. On the basis of these results, it was hypothesized that NC has the greatest angiogenic ability that is made by synergistic effect from BMP2 and PDRN. To confirm the angiogenic effect on NC, we executed quantitative reverse transcription polymerase chain reaction (qRT-PCR) on angiogenesis (Fig. 4D). As expected, the gene expression levels of VEGF in the PDRN- and NC-treated groups significantly increased (P < 0.01 and P < 0.001, respectively). The expression levels of one of the most common angiogenic factors, angiopoietin-2 (ANG2), statistically increased in the NC-treated group. As illustrated in Fig. 4E, PDRN acts as an adenosine A2 receptor agonist () and performs as the main angiogenic factor for up-regulating VEGF expression. BMP2 is a well-known molecule, which enhances angiogenesis, but PDRN shows greater angiogenic ability than BMP2.

Fig. 4.

Biological ability conformation of each component of the NC.

(A) Cell migration assay: optical images (scale bars:,200 μm) and quantification for 24 and 48 hours. (B) Tubule-forming assay: calcein AM–stained images (scale bars, 200 μm) and quantification of HUVECs. (C) Immunofluorescence staining labeled with VEGF (scale bars, 50 μm). (D) Gene expression onto the scaffolds related to angiogenesis, VEGF, and ANG2 using hfMSCs. ns, not significant. (E) Schematic illustration of angiogenesis through A2a receptor stimulation induced by PDRN and NC. #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate statistically significant differences, respectively.

Besides, fig. S8 shows the biocompatibility of the NC-immobilized hybrid scaffold (PME/NC). On the basis of LIVE/DEAD [calcein AM (acetoxymethyl ester) and ethidium homodimer 1 (EthD-1)] staining, fluorescence images representing hfMSCs were viable 1 day after seeding, and the cells proliferated well on each scaffold, especially PME/NC.

In vitro osteogenic differentiation behaviors

To evaluate osteogenic efficacy, the osteogenic potential was investigated by alkaline phosphatase (ALP) activity and mineralization using hfMSCs. Guillot et al. () identified that hfMSCs have substantial potential for osteogenic differentiation due to a fast growth rate, longer telomeres, and expression pluripotency markers (i.e., oct-4, rex-1, and nanog) compared to those of adult stem cells. We also highlight the better osteogenic ability of hfMSCs compared to umbilical cord–derived MSCs (UC-MSCs) and bone marrow–derived MSCs (BM-MSCs) (fig. S9). To maximize the bone regeneration efficiency of the PME/NC, hfMSCs were selected for the hybrid scaffold.

To identify the osteogenic capacity of NC, staining and quantification were performed using hfMSCs for 7 days in osteogenic media. Figure 5A represents ALP-stained images and their quantification. In Fig. 5B, alizarin red S staining was analyzed at 7 days in osteogenic media. As expected, the BMP2-treated group showed a significant increase in ALP activity and mineralization. However, PDRN also affected osteogenesis compared to control. Consequently, the NC has a brilliant osteogenic ability, which is made by a combinational effect from BMP2 and PDRN. When designing a growth factor delivery system, the main goal is that the fractured bone is healed with a low amount of growth factor, not only to prevent abnormal bone formation but also to curtail the higher cost. The advanced hybrid scaffold contains merely 200 ng of BMP2 in the form of NC, which was a relatively lower concentration compared with that reported for other research results obtained using a BMP2 delivery system (, ).

Fig. 5.

In vitro hfMSC differentiation.

Osteogenic ability of the NC using ALP (A), Alizarin Red S (B) staining, and their quantifications for 7 days in an osteogenic medium (scale bars, 200 μm). (C and D) Gene expressions related to osteogenesis in the scaffolds: ALP, RUNX2, OPN, OCN, and ON at 7 and 21 days of differentiation. (E to G) Reducing osteoclastogenesis and anti-inflammatory response of NC and the NC-immobilized hybrid scaffold (scale bars, 200 μm). (E) Osteoclastogenesis of RAW264.7 cells. Proinflammatory marker expressions of hfMSCs (F) and proinflammatory cytokine expression analyzed by enzyme-linked immunosorbent assay (G) using hfMSCs in the scaffolds for 7 days. #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate a statistically significant difference, respectively. P, PLGA; PM, PLGA scaffold containing MH; PE, PLGA scaffold containing bECM.

The osteogenic capacity of the scaffold was assessed. Considerably, in the PME/NC scaffold with hfMSCs, the expressions of osteogenesis-related genes, such as ALP, runt-related transcription factor 2 (RUNX2), osteopontin (OPN), osteocalcin (OCN), and osteonectin (ON), were enhanced by 3.16-, 6.68-, 4.74-, 3.10-, and 5.37-fold higher than the PLGA group at 7 days. The expressions on 21 days were also up-regulated by 3.20-, 5.14-, 2.35-, 3.90-, and 4.46-fold higher in PME/NC scaffold, respectively (Fig. 5, C and D). At 7 days of differentiation, PME/NC scaffold displayed more than five times higher ALP gene expression compared to PLGA. In addition, RUNX2, OPN, OCN, and ON were expressed to higher levels in the PME/NC scaffold compared to any other scaffolds. Similarly, after 21 days of differentiation, gene expression levels related to osteogenesis notably increased in the PME/NC scaffold.

In many studies, hematopoietic progenitor cells were related to bone resorption via differentiation into osteoclasts in bone homeostasis (). Adequate reduction in osteoclastogenesis is necessary for bone regeneration. The receptor activator of the nuclear factor κB ligand (RANKL)–mediated signaling pathway could promote differentiation of RAW264.7 cells. The canonical differentiation pathway for osteoclast differentiation was known to be differentiated by stimulation of RANKL and macrophage colony-stimulating factor (M-CSF) (). In the presence of stimulators, the tartrate-resistant acid phosphatase (TRAP)–positive mononuclear cells are aggregated, and then, the mononuclear precursors were multinucleated by cell fusion with the expression of inflammatory cytokines. To assess the inhibiting ability of NC for osteoclastogenesis, RAW264.7 cells were induced to differentiate into osteoclasts with RANKL and M-CSF. TRAP-positive macrophages were investigated through TRAP staining and activity assay (Fig. 5E). In the groups containing PDRN, the differentiation of osteoclast was inhibited as compared with the control and BMP2. The groups with BMP2 demonstrated that the multinucleation occurred. It was examined whether the NC regulates inflammatory gene expression. Most tissue regenerative strategies including bone aim to reduce inflammation for successful targeted tissue repair (–). Likewise, the PDRN and NC groups decreased the expressions of interleukin-1β (IL-1β) and IL-6 (Fig. 5F). Analogous results were observed in the IL-6 and tumor necrosis factor–α protein levels (Fig. 5G). Therefore, NC likely has dominant roles as an activator of osteogenic differentiation and regulator of osteoclast formation and anti-inflammation.

In vivo angiogenesis: NC-immobilized hybrid scaffold promotes angiogenesis around the defect area

Harvestin et al. () investigated the importance of revascularization during bone regeneration using preconditioned scaffold. In the same vein, Subbiah et al. () introduced the dual growth factor (VEGF and BMP2) delivery system. Meanwhile, we aimed to estimate whether sustainedly released PDRN from scaffold could induce angiogenesis during bone regeneration. As mentioned previously, the interplay between PDRN and BMP2 (NC) was revealed to enhance angiogenesis and osteogenesis for augmented bone regeneration. In Fig. 4, the PME/NC scaffold exhibited superior ability for enhancing angiogenesis. After demonstrating the multifunctional effects of the PME/NC scaffold in vitro studies, it was implanted into critical-sized rat calvarial defect model in groups with hfMSCs (PLGA, PME, and PME/NC). To investigate angiogenic effect in vivo, vascularization in the defect area was evaluated. After 8 weeks of implantation, Microfil was injected after full perfusion and imaging using micro–computed tomography (micro-CT) after fixation and decalcification to distinguish the vessel area (Fig. 6A). On the basis of the three-dimensional (3D) constructed images, the vessel volume density and vessel number were quantified. To identify the accurate vascularization efficiency of PME/NC, it is designed to be compared with native rats in this experiment. As shown in the quantified results, PME/NC appeared with abundant newly thick vessels, and it exhibited a negligible difference with the native group (P > 0.05) in the originally defected area (Fig. 6A, b and c). In a comparison between PLGA and PME groups, the number of vessels was small and thinner, respectively. In other studies, to assess vascularization in vivo, additional immunohistochemistry (IHC) analysis was performed using α–smooth muscle actin (α-SMA), platelet endothelial cell adhesion molecule 1 (PECAM-1; CD31), etc. (–). The regenerated defect sites were examined by IHC staining with anti-VEGF antibody because PDRN is an agonist of A2a receptor, which can induce VEGF production. In particular, VEGF-positive cells were observed for the PME/NC scaffold implanted group (Fig. 6B). Besides, Fig. 6C reveals angiogenesis-related gene expression, VEGF, fibroblast growth factor 2 (FGF-2), platelet-derived growth factor (PDGF), matrix metalloproteinase 2 (MMP-2), and MMP-9. In the PME/NC group, the osteogenesis-related genes were expressed to be 4.47-, 14.79-, 10.25-, 2.23-, 2.55-, and 16.16-fold higher, respectively, which were each statistically different (P < 0.0001).

Fig. 6.

Angiogenic and anti-inflammatory effects of the hybrid scaffold.

(A) Microfil images (a) and quantification; vessel volume density (b) (VV/TV; VV, vessel volume; TV, tissue volume) and vessel number (c). (B) Immunohistochemistry (IHC) analysis stained with VEGF antibody (scale bars, 50 μm). DAPI, 4′,6-diamidino-2-phenylindole. Gene expression related to angiogenesis; VEGF, FGF-2, ANG-1, platelet-derived growth factor (PDGF), matrix metalloproteinase 2 (MMP-2), and MMP-9, and (C) proinflammation; IL-1β and IL-6 (D). #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate a statistically significant difference, respectively.

In addition, the anti-inflammatory efficacy of PME/NC by gene expression related to proinflammation in vivo was confirmed (Fig. 6D). As expected, IL-1β and IL-6 in the PLGA group were expressed to be 2.31- and 27.02-fold higher than those of the control, respectively, because its degradation products made acidic microenvironment at the implanting site. However, MH-containing groups (PME and PME/NC) were detected to have no statistically significant difference because the pH was neutralized by MH. In other words, these observations indicate that the PME/NC scaffold could enhance not only anti-inflammation but also angiogenesis during bone repair.

In vivo regeneration: NC-immobilized hybrid scaffold promotes calvarial bone regeneration

To investigate new bone formation ability of the PME/NC in vivo, we evaluated three scaffolds implanted in groups with hfMSCs (PLGA, PME, and PME/NC) and one group as a control (defect only). Figure 7A represents the micro-CT and quantification after 8 weeks of implantation. As shown in the micro-CT images, there is a negligible difference in new bone formation between the control and that in the PLGA groups. In the PME group, some mineralization was observed in the middle of defects. Newly regenerated bone in the defect area was detected for the PME/NC implanted group. The bone volume density [bone volume (BV)/tissue volume (TV)] for the PME/NC implanted group was significantly higher than the control (P < 0.0001). Likewise, the bone mineral density (BMD) of the control, PLGA, PME, and PME/NC was observed to be 132.23, 140.15, 190.98, and 520.64 g·cm3, respectively, in which the BMD of the PME/NC increased by approximately four times higher than control. The BMD is defined as the amount of mineral per unit of bone and is used as the main diagnosis indicator of osteoporosis. It means that low BMD value can highly increase the potential of bone fracture. Therefore, the high BMD value in PME/NC indicated that the newly formed tissue can completely perform the original bone. As shown in Fig. 7A (c), the BMD value of the PME/NC increased about 3.7-fold than control, compared with the results of other studies that usually increased about twofold (, , ). In particular, despite a comparatively low dose of BMP2 (200 ng per scaffold) compared with other studies (–, ), newly formed bone tissue in defects, which significantly enhanced in both bone volume density and mineral density (P < 0.0001), was effectively repaired in the PME/NC. These micro-CT results indicated that the advanced hybrid scaffold, PME/NC, could significantly facilitate bone regeneration and that this outstanding performance is in comparison to the other conditions investigated in this study.

Fig. 7.

New bone formation after 8 weeks of implantation in vivo.

(A) Representative 3D constructed images of a superficial view of the calvarial defect using micro-CT (a). Quantified parameters of repaired bone, including (b) bone volume density (BV/TV) and (c) bone mineral density. Histological analysis with (B) H&E, (C) MT, and (D) IHC analyses of antiosteocalcin after 8 weeks of implantation. Scale bars, 1000 and 500 μm. (E) Gene expressions related to osteogenesis: ALP, RUNX2, OCN, OPN, and COL1A1. #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate a statistically significant difference, respectively.

Moreover, histological evaluation was also analyzed to further support the radiographic results and to verify the internal bone tissue repair. The sagittal section of the entire bone defect area was stained with hematoxylin and eosin (H&E) and Masson’s trichrome (MT) (Fig. 7, B and C). On the basis of the H&E- and MT-stained images, considerably newly developed bone tissue was observed in the defected area for the PME/NC scaffold than negative control. Furthermore, IHC analysis was conducted for clearer histological evidence of bony development using the OCN antibody (Fig. 7D). There were notably increased OCN-positive cells for the PME/NC scaffold implanted group. For further examination, the gene expression was identified, and the defect area was collected using a 4-mm micro drill in the same way after 8 weeks of implantation. Gene expressions related with osteogenesis such as ALP, RUNX2, OCN, OPN, and COL1A1 were investigated (Fig. 7E). Some of the gene expressions were enhanced for the PME scaffold; the PME/NC scaffold demonstrated highly up-regulated expressions for all five genes. On the basis of in vivo results, the PME/NC scaffold promotes angiogenesis, thereby increasing the effect of bone regeneration.

In summary, the advanced bone regeneration system using the PME/NC scaffold has multifunctional ability that plays a crucial role as a substantial bone graft biomaterial. The existing PME scaffold showed significantly improved mechanical properties and a pH-neutralizing capacity. The NC was successfully immobilized on the scaffold for the controlled release of bioactive molecules, BMP2 and PDRN. As shown in the in vitro results, the PME/NC scaffold has many magnificent properties that increase its regenerative effects, which are biocompatibility, osteoclast inhibition, and osteogenesis. Moreover, the PME/NC with hfMSCs performed outstanding bone tissue repair in critical-sized rat calvarial defects because of its surpassing potential of osteogenic differentiation. In addition, the PME/NC exhibited interesting effects on angiogenesis as well as anti-inflammatory and antibacterial responses around the defected area. In particular, neovascularization with the PME/NC scaffold was improved as much as that of the native group in vivo. In the field of bone tissue engineering, many other approaches have been attempted for effective bone regeneration via several side abilities such as neovascularization () and antibacterial effect (). The advanced PME/NC hybrid scaffold showed great multifunctional effects: osteogenesis, anti-inflammation, angiogenesis, and antibacterial effect. On the basis of these results, we suggest that the advanced PME/NC hybrid scaffold be developed using differentiated approaches from previous studies to achieve various applications in bone regeneration. It would be a highly effective therapeutic strategy in bone tissue engineering and regenerative medicine.

MATERIALS AND METHODS

Materials

Poly(lactic-co-glycolic) acid [PLGA; lactide:glycolide = 50:50, Mw (molecular weight) = 110 kDa] was purchased from Evonik Ind. MH was purchased by Sigma-Aldrich. RA was purchased from Tokyo Chemical Industry product. The bovine bone-derived extracellular matrix powder (bECM; InduCera) was supplied by Oscotec Inc. BMP2 (Mw = 26 kDa) was supplied by CGBIO. PDRN was obtained from GoldBio. Amicon centrifugal filter units were purchased from Merck Millipore. Nuclease-free water was purchased from Thermo Fisher Scientific. D-Plus CCK-8 (Cell Counting Kit-8) cell viability assay kit was obtained from Dongin LS (Seoul, Korea).

Preparation and characterization of NC

For the complexation of BMP2 and PDRN, 10 μl of BMP2 solution (1 mg/ml; 1 mg of BMP2 dissolved in 1 ml of nuclease-free water) was added to 1 ml of 10 mM NaCl solution, and 10 μl of PDRN solution (1 mg/ml; 1 mg of PDRN dissolved in 1 ml of nuclease-free water) was added (pH 7.0) under constant stirring (400 rpm) at 4°C for 1 hour. The solution was then subjected to filtration of unreacted materials using centrifugal filter. Prepared NC solution was freeze dried for 2 days to obtain NC in powder. The mean size and zeta potential of NC were determined using Zetasizer Nano ZS (Malvern Instruments Ltd.). The measurements were conducted at 25°C and the wavelength of 633 nm. In addition, the change of mean size and pH was evaluated in accordance with the additive of PDRN. The surface morphology of NC was visualized using an SEM (FE-SEM; SIGMA, Carl Zeiss), and TEM (H-7600, Hitachi) was conducted. The NC was characterized using an FTIR spectrometer (Spectrum Two, PerkinElmer). The retardation assay was performed with 1% agarose gel. Then, the gel was visualized by Bio-Rad Universal Hood II (Bio-Rad Laboratories). CD spectra were recorded using spectropolarimeter (J-815, JASCO) using a quartz cell of 10 mm, 163 to 320 nm, with a scan speed of 50 nm/min.

Fabrication of PLGA/mMH/bECM scaffold

Modified Mg(OH)2 with RA (mMH) was synthesized following the protocol of a previous study (). The PLGA/mMH/bECM scaffolds were fabricated using ice particles as a porogen by the freeze-drying method. Ice particles (100 to 200 μm) were prepared by spraying deionized water into liquid nitrogen. mMH (20 wt %; compared to PLGA) dissolved in 0.3 M dichloromethane solution was completely mixed with 0.5 g of PLGA. Then, 50 wt % bECM powder of the PLGA mass was added to the preceding solution. All equipment and materials for scaffolds should be stored at −20°C for 3 hours before production. Ice particles (6.5 g) were added in the mixing solution using a freezing spatula in a cold chamber (−20°C). The mixtures were stuffed into round PTFE (polytetrafluoroethylene) mold (ø5 × 2 mm2). The filled molds were freeze dried for 2 days to remove the ice particles and some organic solvents, and then the porous scaffolds were gained.

Preparation and characterization of PLGA/mMH/bECM/NC hybrid scaffold

To immobilize NC on PLGA/mMH/bECM scaffold, the scaffolds were hydrated by 70% ethanol and deionized water in order. Then, both NC and PLGA/mMH/bECM scaffolds were immersed together in a supersaturated calcium phosphate solution [NaCl (8.065 g), CaCl2 (0.554 g), and Na2HPO4 (0.284 g), buffered with tris (50 mM; pH 7.4)]. After immobilization, the products were washed with distilled water and then freeze dried for 1 day. The surface and cross-sectional morphology of the scaffolds were observed by SEM. The prepared scaffolds were analyzed by a UTM (Instron-4464) with 1 N load cell to measure mechanical property. To evaluate the hydrophilicity and hydrophobicity of the scaffolds, the water contact angle was measured using an optical bench-type contact angle goniometry (VCA Optima XE Video Contact Angle System, Crest Technology). Thermal property of the scaffolds was estimated using a thermogravimetric analyzer (TGA 4000, PerkinElmer). The elementary compositions of scaffold were measured by ICP-OES (Optima 8000, PerkinElmer). To evaluate neutralization effect of mMH and bECM on the by-product of PLGA, mass and pH changes were measured in 400 μl of PBS solution (pH 7.4) for 60 days. Each optical image for 60 days was cropped at a predetermined day and arranged to one image with the black background using Adobe Photoshop CC 2019.

Antibacterial test

Gram-negative bacteria, E. coli, and Gram-positive bacteria, S. aureus, were incubated in Luria-Bertani medium [tryptone (10 g/liter), yeast extract (5 g/liter), and NaCl (10 g/liter)] at 37°C with aeration. After 16 hours, the bacteria were centrifuged, and then the pellet was washed with sterilized 0.85% NaCl solution and resuspended in the sterilized 0.85% NaCl solution. The density of bacteria solution was controlled at ~104 CFU (colony-forming units)/ml. Various composite scaffolds were added to 1 ml of bacterial suspension for antibacterial testing. The composite scaffolds with bacteria suspension were incubated at 37°C for 1 day. Then, 100 μl of diluted suspension was spread on Luria-Bertani agar plates, and then the plates were incubated overnight at 37°C. The inhibition of bacterial growth was evaluated by the CFUs.

Isolation of hfMSCs

Human fetal MSCs were obtained from an 8-week-old fetus from an ectopic pregnancy under the donor’s consent. The internal review board of the Gangnam Cha Hospital approved obtaining the fetal MSCs. After the establishment of the research cell bank (Fetal MSC_#001), hfMSCs were expanded with Dulbecco’s modified Eagle’s medium (DMEM)/high glucose medium (HyClone, SH3024.01) supplemented with 10% fetal bovine serum (FBS) (Gibco, 16000-044, Thermo Fisher Scientific, Korea), basic fibroblast growth factor (bFGF) (10 ng/ml; Gibco, 17100-0170, Thermo Fisher Scientific), and 1X penicillin and streptomycin (Gibco, 15140-122, Thermo Fisher Scientific) until use. hfMSCs were grown at 37°C under a humidified 5% CO2 atmosphere. The medium was changed every 2 to 3 days. Cells at passages nos. 10 to 12 were used for experiments.

Cell proliferation on hybrid scaffold in vitro

To assess the capacity of osteogenic differentiation and biocompatibility of the scaffolds, hfMSCs were plated in 75-cm2 culture flasks at 37°C and cultured in DMEM high glucose supplemented with 10% FBS, bFGF (10 ng/ml), and 1X penicillin and streptomycin. Cells were seeded onto hydrated scaffold (5 × 105 cells per scaffold) in a dropwise manner with 10 μl of media. Cell proliferation on the scaffolds was evaluated using CCK-8 at days 1, 3, and 7. LIVE/DEAD staining was performed with LIVE/DEAD Viability/Cytotoxicity Kit (Invitrogen) for observation of seeded cells in each group of scaffolds using confocal microscopy.

Migration assay and angiogenesis

The scratch wound healing assay was subjected to analyze the migratory capacity of hfMSCs. The cells were seeded into six-well culture plate at the density of 5 × 105 cells per well and cultured for 1 day. The confluent monolayer of cells was scratched using a 1-ml pipette tip and then washed with PBS solution. Cells were cultured in DMEM high glucose containing 1% (w/v) FBS, and BMP2 (1 μg/ml), PDRN (1 μg/ml), and NC (1 μg/ml) were added. At the predetermined time point, the wells were photographed, and the wound area was quantified using ImageJ. To assess the angiogenic ability of NC, 250 μl of Matrigel matrix (Corning) was added to a precooled 24-well plate and then incubated in a 37°C incubator for 1 hour. HUVECs (1.2 × 105 cells per well) were seeded onto Matrigel matrix with 1 ml of EBM-2 media (Lonza) containing 1% FBS. Then, BMP2, PDRN, and NC were added. After 18 hours, HUVECs were stained with calcein AM (4 μmol; C1430, Thermo Fisher Scientific Inc.). The plates were incubated at 37°C for 15 min and then photographed by optical microscopy (DP-74, Olympus) with fluorescence lamp (U-RFL-T, Olympus). The tube formation was quantified using ImageJ for tube length and branch point at three random sites in each well. To visualize VEGF expression in hfMSCs, the ICC was performed with anti-VEGFA antibody as a primary antibody (4°C, overnight), and Alexa Fluor 488 secondary antibody was used for binding for 1 hour at 25°C. The gene expressions related with angiogenesis were analyzed by qRT-PCR, and the RNA was extracted from HUVECs at 48 hours.

ALP staining and activity

For 2D assays, hfMSCs (5 × 104 cells per well) were cultured in 24-well plates. After 1 day, the medium was replaced with an osteogenic induction medium containing 10% FBS, 1% antibiotic antimycotic (A/A), 50 μM l-ascorbic acid, 0.1 μM dexamethasone, and 10 mM β-glycerophosphate. After 7 days, cells were fixed with 10% formalin, rinsed with deionized water, and stained with Takara ALP stain kit (MK300). The stained samples were captured with an optical microscope. For the measurement of ALP activity, the cells were lysed using Takara ALP assay kit (MK301, Takara). The whole process was conducted according to the provided protocol.

Osteogenic differentiation ability: Staining and quantification of mineralization

After 14 days of osteogenic induction culture, mineralized extracellular matrix was stained with Alizarin Red S (A5533; Sigma-Aldrich). The cells were fixed with 10% formalin for 20 min and rinsed with deionized water. The plate was incubated in 2% Alizarin Red S solution for 20 min. The stained cells were observed with an optical microscope. To quantify the staining, the 10% cetylpyridinium chloride (CPC; C0732, Sigma-Aldrich) was added to each well and incubated for 15 min to elute the stain. The colorimetric measurement was read at a wavelength of 562 nm using a microplate reader (SpectraMax M2, Molecular Devices).

RNA extraction and qRT-PCR

RNA from 2D cultured cells and 3D scaffolds were extracted using Universal RNA Extraction Kit (K-3141, Bioneer) and TRIzol reagent (15596018, Thermo Fisher Scientific) following the manufacturer’s instructions. The RNA concentration was determined by a spectrophotometer (ND-1000; Thermo Fisher Scientific). RNA (100 ng) from each sample was reverse transcribed to complementary DNA using a PrimeScript RT Reagent Kit (Perfect Real Time). The qRT-PCR was performed using Power SYBR Green PCR Master Mix (Applied Biosystems) with a QuantStudio 3 real-time PCR instrument (Applied Biosystems). The relative osteogenic, angiogenic, and inflammatory gene levels were calculated with 18S rRNA as a reference gene (Table S2).

Inhibition of osteoclastogenesis: Staining and activity of TRAP

To differentiate immature precursor cells into osteoclast, the RAW264.7 cells and mouse macrophages were seeded into a 96-well plate (3 × 103 cells per well). After 1 day, RANKL (70 ng/ml) and M-CSF (30 ng/ml) were treated with the components of NC. At 5 days after differentiation, the cells were fixed with 4% paraformaldehyde, rinsed with deionized water, and stained with Takara TRAP stain kit (MK300). The stained samples were captured with an optical microscope. For the measurement of TRAP activity, the cells were lysed using Takara TRACP assay kit (MK301, Takara). The whole process was conducted according to the provided protocol.

Surgical procedure

The experimental protocols for the use of animals were approved by the Institutional Animal Care and Use Committee of CHA University (IACUC200120). At 8 weeks, Wistar rats, weighing between 190 and 230 g, were used in the study. The animals were anesthetized with an isoflurane (Terrell Isoflurane, Piramal Critical Care Inc., USA). The hair was shaved, and the exposed skin was sterilized with ethanol. Then, a sagittal incision was executed. A defect (ø5-mm and 1.5-mm thickness) was made on both sides of rat calvaria at a certain distance from the sagittal suture line using micro drill and trephine bur. The drilled calvarial disc was removed, and the scaffolds were implanted within the defect. The periosteum was closed with absorbable sutures (4-0 VICRYL, Ethicon Inc.), and the epidermis of implanted site was sutured with nonabsorbable sutures.

Microfil perfusion

To evaluate blood vessel formation after 8 weeks, the rats were perfused with Microfil (MV-122, Flow Tech). During anesthesia, the hair of the chest was shaved, and the ribs were opened using a scissors. The left ventricle was penetrated with a 21-gauge butterfly needle, and then 50 ml of heparinized saline and 50 ml of 4% paraformaldehyde were completely perfused at rate of 9 ml/min using Microfluidic Syringe pump (NE-1000-ES). Last, 20 ml of Microfil was perfused at a rate of 2 ml/min. The perfused samples were set overnight at 4°C to completely cure the contrast agent.

Micro-CT analysis for new bone formation and vascularization

The obtained calvarial samples were fixed in 4% paraformaldehyde at 25°C for 7 days for micro-CT scanning. Images were obtained at x-ray tube voltage, 90 kVp; x-ray tube current, 180 μA; and field of view (FOV), 20 mm using Quantum FX micro CT (PerkinElmer). The bone volume density (BV/TV; %) and BMD (%) were measured with the Analyze 12.0 (threshold, 5000~; region of interest, 100 × 100 × 50).

Histological evaluation

All fixed samples were incubated with RapidCal immune (BBC Biochemical) solution in gently shaking condition at room temperature. The solution was changed every 3 days. After decalcification, the samples were embedded into paraffin and sectioned at 10 μm. The sectioned slides were stained with H&E (Abcam) and MT (VitroVivo Biotech).

Statistical analysis

All experiments were repeated at least three times. The results are shown as means ± SEM. #P < 0.0001, ***P < 0.001, **P < 0.01, and *P < 0.05 indicate a statistically significant difference, respectively. Statistically significant differences were evaluated by one-way analysis of variance (ANOVA) using Tukey method in GraphPad Prism 7.0 software (GraphPad Software Inc., CA, USA).