The micrographs of PCL/HA scaffolds were examined by scanning electron microscopy (SEM, JEOL JSM-7001F, Japan) at 5–10 kV and the specimens were coated with gold-sputter using an auto fine coater (JEOL, Ltd., Japan) under argon atmosphere for 60 s. As shown in Fig. 1a, HA nanoparticles scatter uniformity in the scaffolds with some aggregated occasions. Uniform pore with different diameters is observed from the 3D printed PCL/HA scaffolds. Figure 1b shows the mapping distribution of C, O, Ca, and P elements in the HA powders, which appears uniformly distribute in the scaffolds. Herein, the elemental distribution in the scaffolds is investigated using an EDS (Phenom World BV, Netherlands). In this work, PCL/HA scaffolds with 400–800 nm of uniform pore sizes are chosen as the implanted samples as such pore size is expected to be beneficial for cell proliferation/differentiation in vivo.
Chemical characterization of pure PCL, HA and PCL/HA scaffolds with different content of HA were conducted by a Nicolet is10 Fourier Transform-Infrared (FT-IR) spectrophotometer (Thermo Fisher Scientific Inc., Madison, WI, United States of America) and their characteristic peaks are shown in Fig. 2a. For the composite samples, many of the absorption bands are overlapped but the properties of the functional groups remain unchanged. Whereas, the shape, location and intensity of the spectral peaks change significantly. The spectra of PCL/HA scaffolds show typical ester peaks (at 1724 cm− 1, 1640 cm− 1) and CH peaks (2923 cm− 1). The increase of HA content in the scaffolds (from 0 to 25%) decreases the peak intensities. Specifically, PCL/HA scaffolds show spectral features in the range of 1150 and 960 cm− 1, which are related to the P-O and P = O vibrations of HA. The peak intensities of the spectral range of 3500 − 3200 cm− 1 that represents the absorption peak of OH groups of HA, decrease evidently.
Typical X-ray diffraction (XRD) spectra obtain for PCL/HA scaffolds are presented in Fig. 2b. The scaffold specimens were fixed on a specimen holder by double side adhesive tape and the XRD diffractometer (German Bruker Co., Germany) with Cu Ka radiation (λ = 0.154056 nm, 40 kV, 40 mA) was utilized. The data were recorded in the 2θ range of 10° − 75° with a scanning speed of 8 min− 1. As illustrated in Fig. 3b, the spectra of PCL/HA scaffolds appear similar as that of pure PCL. In addition to the strong peaks associated with the crystalline PCL phase, relatively weak peaks at ∼25.32°, 32.78°, 41.43°, 45.58° and 46.58° are observed, which correlate to the crystalline HA phase. These observations imply the presences of both PCL and HA in the mixture powders. Combined with the above SEM image in Fig. 2a, these observations indicate that the fine HA nanocrystals are well dispersed in the PCL matrix.
Figure 2c compares the water contact angle of the PCL/HA scaffolds containing different content of HA. The water angle was measured using an automatic contact angle meter (SL200A/B/D Series, Solon Tech. Inc. Ltd., Shanghai, China). As it is seen, the water contact angle decreases from 91.6° for pure PCL to 80.70° for PCL with 15wt% HA, and then increases gradually to 81.4° when the HA content increases further to 25wt%. In all examined samples, the water contact angle on the PCL/HA scaffolds is much smaller than that on the pure PCL. Such result demonstrates that presence of the nano-HA enhances the surface hydrophilicity of the PCL scaffolds [21, 22].
The compressive modulus of PCL/HA scaffolds are measured and shown in Fig. 2d. Here, the scaffolds are fabricated in the form of cylinders (with a diameter of 4 mm and a height of 6 mm) and vertically placed between two parallel plates in a universal testing machine (MTS Industrial Systems Co., Ltd, Shenzhen, China,). A compression rate of 1.0 mm min− 1 was utilized. Each compressive test was repeated five times. As it is seen, the compressive modulus increases from 112 MPa to 330 MPa when the HA content increases from 0 to 25wt%. In literature, the compressive modulus for pure PCL is measured between 85 MPa and 224.9 MPa [23, 24], which agrees with our measurements. The enhancement is expected as originated from the higher intrinsic mechanical properties of HA (an average compressive strength of 174 MPa and a Young's modulus of 6 GPa) , and the uniform distribution of HA in the PCL matrix which show in the Fig. 1 play a very important role in the mechanical improvement of PCL/HA nanocomposite [24, 26].
The glass transition temperature (Tg) of the PCL/HA scaffold was measured by a differential scanning calorimeter (DSC) (NETZSCH DSC 200 F3, Erich NETZSCH GmbH & Co. Holding KG, Gebrüder-Netzsch-Strasse, Selb, Germany). For pure PCL, the glass transition temperature (Tg) is -66.4°C . Figure 3a shows the DSC curves of PCL and PCL/HA composites, from which the Tg varies from − 70oC to -60oC, suggesting a good compatibility between PCL and HA. According to Fig. 3b, the degradation temperature for weight loss at 10% decreases from 389.7oC to 363.3oC when HA content increases from 0% to 25wt%. Similar result is also observed the degradation temperature for weight loss at 50% (from 515.0oC to 414.0oC). In all, the results from thermogravimetric analysis (TGA) analysis indicate that the addition of HA in PCL decreased the thermal stability of scaffolds as the starting decomposition temperature decreases.
Additionally, the in vitro biocompatibility of the PCL/HA scaffolds was carried out in terms of the proliferation of MC3T3-E1 cells. The osteoblast cells MC3T3-E1 were provided by the China Center for Type Culture Collection of Tongji Medical College, Huazhong University of Science and Technology, and raised according to the method described in literature . In general, no significant differences of cell proliferation activity among all scaffolds are observed after 1 day (Fig. 4.). Whereas, evident differences of cell proliferation activity start to appear among all scaffolds after 3 days. Specifically, PCL/25wt% HA content exhibits the similar cell proliferation activity with the control cells. These results signify that PCL/25wt% HA possesses good biocompatibility and can stimulate cell proliferation well. In all, based on above analysis, PCL/25wt% HA shows the best comprehensive performances and thus being selected to assess their biomedical application potentials.
3.2 In vitro degradation property of PCL/HA scaffolds
Ideally, scaffold materials should be biodegradable during the growth of new bone tissue but maintain good mechanical properties before the bone tissue is completely regenerated. Thus, it is important to understand the biodegradable behaviors of the scaffolds during the degradation for bone regeneration. For such purpose, PCL and PCL/HA scaffolds were suspended in 10 mL of PBS in a dialysis bag and then slowly shaken in 90 mL of phosphate buffer saline (PBS) (pH 7.4) at 37°C. At predetermined time intervals, the samples were taken out of the degradation medium, rinsed with distilled water and then dried in vacuo for 48 h.
As illustrated in Fig. 5a, the number of large pores and cracks increases in the PCL/HA scaffolds when the degradation time increases, and the PCL/HA scaffolds gradually lost their regularity and uniformity. The degradation of the sample can be well reflected by the weight loss and molecular weight loss shown in Fig. 5b and 5c. Compared with pure PCL, the weight loss for PCL/25wt% HA sample is a much smaller during degradation. For instance, after 180 days degradation, the weight losses of pure PCL and PCL/25wt% HA scaffolds are about 10% and 1%, respectively. These results indicate that mass loss occurs at a much lower rate for samples with higher HA content. It is probably that the alkalinity of HA nanoparticles induces neutralize acidic substances during the degradation of PCL, which resulting in the inhibition of self-acceleration of acidolysis and decreasing of hydrolysis of ester bonds.
The Young’s modulus of PCL/25wt% HA content after degradation are measured from tensile experiments and summarized in Table S2 (see Supporting Information S2). It is found that Young’s modulus decreases gradually when the degradation time increases from 0 to 4 month, which aligns with the weight loss during the degradation process. It is expected that during the first three months, the degradation processes mainly occur at the end and pendant functional groups of the polymeric chain, and thus result in a minor influence on the molecular chain. After 90 days, the degradation is expected to induce remarkable reduction to the molecular weight of the polymeric main chain and thus results in significant reduction in Young's modulus.
DSC curves of pure PCL and PCL/25wt%HA composites after degradation are also measured and shown in Fig. S3 (see Supporting Information S3). Pure PCL and PCL/HA are found to keep the characteristic peaks of glass-transition temperature varying from − 70℃ to -60℃ after degradation in PBS for months. Specifically, Tg of both samples is found to decrease slightly with the increase of degradation time.
3.3 In vitro drug-release property of PCL/HA scaffolds
To probe the drug-release properties of the PCL/HA composites, Doxorubicin (DOX) were chosen as a model drug and mixed in the samples. DOX is a chemotherapy drug that is used to inhibit the growth of tumors cells [26–29]. In this work, 10 mg of DOX is mixed with each 100 mg of the pure PCL or PCL/HA composite sample. After mixture, the samples were evaporation and then shaped in a disc mold (with a diameter of 10 mm and a height of 2 mm) and suspended in 10 mL of phosphate buffer saline (PBS) (pH 7.4) in a dialysis bag with a continuous shaking for 250 h in the dark environment. The DOX loading capacity was determined using UV-Vis spectrophotometer (UV-2800 series, Unico, Shanghai, China) at 483 nm. The changes of the concentrations of DOX were obtained from curves of the absorption versus concentration of DOX in PBS following the basis of Lambert-Beer law .
Recent studies show that the release amount of DOX at first 100 h could reach over 80% [29, 30], such high release rate could inhabit the regeneration of new bone tissue in the tissue engineering. Thus, a controllable release rate is a necessity. Figure 6 compares the DOX-release properties of the pure PCL and PCL/HA scaffolds. After ~ 25 hours (or a day), the DOX-incorporated pure PCL and PCL/HA scaffolds display steady drug-release rates (as indicated by the linear profile of the curves in Fig. 6). Compared with the pure PCL scaffolds, PCL/HA scaffolds exhibit faster drug-release rates, which is suspected as resulted from the increased drug diffusion coefficient (due to the presence of HA). Moreover, the PCL/25wt% HA scaffold show higher release rate than that of the counterpart with 10wt% of HA. After 34 days, the cumulative DOX-release percentage is around 22.0% and 37.7% for the samples with 10wt% and 25wt% of HA, respectively, which is much higher than that of the pure PCL scaffolds (about 10.1%). It is expected that the high content of HA decreases the entanglement degree of PCL, which promotes the DOX release from the scaffolds . Above results suggest that the DOX release rate can be effectively controlled by the content of HA in the PCL/HA scaffolds.
3.4 Cell response of BMSCs of PCL/HA scaffolds
To further elucidate the effect of PCL/HA scaffolds on the differentiation of rat bone marrow-derived mesenchymal stem cells (BMSCs), cell proliferation was evaluated using 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay after culturing for 1 and 2 days, respectively. Cells adhesion was performed with 4, 6-diamidino-2-phenylindole (DAPI, Sigma) and FITC-Phalloidin (Sigma) as previously described . The expression of osteogenic differentiation related genes and proteins in rat BMSCs cell was also evaluated. The mRNA transcript levels of Actin,alkaline phosphatase (ALP), collagen (COL), runt-related transcription factor-2 (RUNX2), and osteocalcin (OCN) mRNA within rat BMSCs (cultured in different supplemented osteogenic-inducing medium) were assessed by real-time polymerase chain reaction (PCR). In both cases, cells were harvested on day 7 and day 14 then lysed in Trizol (Life Technologies, Carlsbad, CA, USA) and mRNA was extracted according to the manufacturer's protocol. Reverse transcription was carried out using the RNeasy Plus Micro Kit (Hilden, Germany)) and the PCR test was performed using S1000™ Thermal Cycler (Bio-rad, Hercules).
As shown in Fig. 7a, the cell proliferation activity on PCL/25wt% HA scaffolds is obviously higher than the PCL/HA/DOX scaffolds at either day 1 or day 2, whereas the gap between them decreases on day 2. Fluorescence images show that the rat BMSC cells display good adhesion on the surface of the PCL/25wt% HA scaffolds (Fig. 7b), which is beneficial for bone tissue regeneration. From Fig. 7c, PCL/25wt% HA scaffolds support the growth of BMSCs cells and evidently promote the expression of Actin, ALP, COL, RUNX2, OCN mRNA and proteins. The expressions of those genes and proteins are much higher at day 14 than that at day 7. These results demonstrate that the PCL/25wt% HA scaffolds possess good biocompatibility and can stimulate cell proliferation well. Compared with the PCL/25wt% HA scaffolds, the expressions of those genes and proteins are much smaller in the PCL/HA/DOX scaffolds, which is expected as resulted from the quickly released DOX.
3.5 In vivo bone regeneration with skull defect
In vivo bone regeneration ability of PCL/HA scaffolds was investigated by measuring new bone formation using the model of rats with calvarial bone defects. Twelve 6-week-old Sprague-Dawley rats were divided into two groups, with 6 rats in each group. A circular skull defect (with a diameter of 10 mm) was created using a trephine bur (3i Implant Innovation, Palm Beach Gardens, FL, USA) on the left parietal bone of each rat. This size of defect was chosen because it is a defect of this size does not heal by itself without intervention . Thereafter, a circular scaffold (with a diameter of 10 mm and a thickness of 2 mm) was sterilized and implanted into the bone defect location for each rat. To assess the new bone formation in the bone defective area after 4 weeks, micro-computed tomography (micro-CT 50, Scanco Medical AG, Bassersdorf, Switzerland) was utilized under the fixed conditions (24 kV, 2 mA, 90 seconds). The muscle tissues around the scaffolds were taken out and embedded in paraffin and sectioned at a thickness of 5 µm for Hematoxylin-eosin (H&E) staining. The H&E staining slides were visualized under an optical microscope.
According to the micro-CT data, there is no observable new bone formed in the control group with pure PCL scaffolds after 4 weeks post-surgery (see Fig. 8a). In comparison, the group with PCL/25wt% HA shows evident formation of new bone at the edge of the bone defect region (after 4 weeks). In the meantime, there is a certain amount of new bone deposition in the central area of the defects. Further histological analysis affirms the new bone formation after 4 weeks in the group with PCL/25wt% HA scaffolds. These observations signify that the PCL/25wt% HA scaffolds promotes the osteogenic activity, which can repair the bone defect. In the histological analysis,no obvious inflammation was observed in the tissue sections in the H&E staining micrographs after 4 weeks post-surgery (Fig. 8b). This observation further suggests that the PCL/25wt% HA scaffolds possess good biocompatibility and can provide a good microenvironment for osteoblast proliferation and differentiation.
3.6 In vivo bone regeneration with thigh-bone defect
The effects of the scaffolds on bone formation in vivo are also evaluated in rabbits with a thigh bone defect in their lower limbs. Eighteen 6-week-old New Zealand white rabbits were divided into three groups with 6 in each group, including the control group (with no scaffold), the group with PCL/25wt% HA scaffolds and the group with PCL/HA/DOX scaffolds. A circular thigh-bone defect (with a diameter of 4 mm) was introduced to the thigh bone of each rabbit. The PCL/25wt% HA and PCL/HA/DOX scaffolds (with a diameter of 4 mm and a height of 6 mm) were firstly coated with collagen I (see Supporting Information S4), and then sterilized and implanted into the bone defects for each rabbit following the method used in literatures [33–37]. The implanted position was determined by the magnetic resonance imaging using M7 Small animal MRI system (1.0 Tesla, Aspect Imaging Ltd, Israel) (see Supporting Information S5). Area of new bone formation and percentage were measured using the Image Processing and Analysis software taking the femoral implant as the reference baseline. The cylindrical area with a diameter of 4.1 mm and a thickness of 6 mm was set as the three-dimensional reconstruction area of interest. The three-dimensional image was reconstructed with N-Recon software and the three-dimensional analysis was performed with CT-AN software.
As shown in Fig. 9, there is a large amount of new bone formed in the rabbits either with PCL/25wt% HA scaffolds or PCL/HA/DOX scaffolds after 8 weeks. Bone tissues are found to gradually penetrate into the scaffold, adhere on the scaffold surface, and then grow to form a network. Along with the repair process, the PCL/HA scaffolds degrade gradually. These results indicate both PCL/25wt% HA and PCL/HA/DOX scaffolds promote the osteogenic activity and can repair the bone defect. To further assess the new formed bone, Fig. 9b compares bone tissue volume/total tissue volume (BV/TV, %), trabecular thickness (Tb.Th, mm), number of trabecula (Tb.N, mm− 1), and bone mineral density (BMD, g.cm− 3) after 8 weeks surgery. Interestingly, more new bone formation is observed from the rabbit with PCL/25wt% HA scaffolds than the counterpart with PCL/HA/DOX scaffolds after 4 weeks. Probably the large amount of DOX released from the PCL/HA/DOX scaffolds inhibit the cell proliferation activity due to its high cell cytotoxicity to the bone cell. The PCL/25wt% HA scaffolds are found to result in higher bone volume fraction, larger trabecular thickness, larger number of trabecula and higher bone density. It is expected that, the slowly released DOX at the earlier stage of implantation is conducive to cell proliferation and inhibits inflammation.