Characterization of the Co (0, 1, 2.5, 5, 10) scaffolds
In the present study, Co-doped scaffolds were synthesized by doping CLP with various concentrations of Co (0, 0.1, 0.25, 0.5, 1 mol%) and SE-3D printing techniques, named Co0, Co1, Co2.5, Co5, and Co10 scaffolds. The effect of Co doping on the CLP structure was determined by characterizing the Co (0, 1, 2.5, 5, 10) scaffolds via X-ray diffraction (XRD), Raman spectrophotometry, and scanning electron microscope (SEM). Figure 2A shows the XRD patterns of the Co (0, 1, 2.5, 5, 10) scaffolds. The XRD patterns of the Co (1, 2.5, 5, 10) scaffolds were found to be similar to that of Co0 and were devoid of any additional peaks corresponding to Co. With increasing Co-doping content, the XRD profile moved toward a slight right shift as a whole. When the doping percentage of Co2+ ions increased up to 1 mol%, the crystal planes (110), (128), (4,0,10), and (2,0,20) obviously changed.
Figure 2B displays the Raman spectra of the Co (1, 2.5, 5, 10) scaffolds. Compared with the Co0 scaffold, the band positions of the Co (1, 2.5, 5, 10) scaffolds were altered, moving to the direction of a higher Raman shift with increasing Co-doping level.
In Fig. 3A, a, color difference was observed between the Co-doped and -undoped scaffolds. After adding Co2+ ions to CLP, the color of the scaffolds transformed from white to purple. With increasing Co content, the color of the scaffolds successively changed from white (Co0) to light purple (Co1) to purple (Co2.5) to dark purple (Co5) and to dark reddish purple (Co10). Figure 3A and B demonstrates the SEM surface morphology of the Co (0, 1, 2.5, 5, 10) scaffolds. Based on the micrographs, regular pores and cylindrical wires of scaffolds fabricated by SE-3DP were interlaced in each layer. The printing wire and pore size were approximately 500 μm and 400 μm × 400 μm, respectively. At high magnification (Fig. 3B), the crystal size of the Co (0, 1, 2.5, 5) scaffolds decreased gradually with increasing Co dopant. When the doping percentage was up to 1 mol%, some melted crystals and microcracks were observed on the Co10 scaffold surface. In Fig. 3C, energy-dispersive spectroscopy (EDS) results showed peaks of Co within the Co (1, 2.5, 5, 10) scaffolds. However, there was no characteristic peak of Co in the Co0 scaffold. In addition, the measured Co/ (Ca + Co) ratios enhanced with increasing Co-doping amount. The EDS results also revealed that Co2+ ions were successfully incorporated into Co (1, 2.5, 5, 10) scaffolds.
In vitro degradation and mineralization properties of the Co (0, 1, 2.5, 5, 10) scaffolds
To study the effect of Co doping on the physicochemical properties of CLP, degradation and mineralization tests were performed. Figure 4A and B shows the degradation properties of the Co (0, 1, 2.5, 5, 10) scaffolds in Tris–HCl solution. As shown in Fig. 4A, the weight loss of the Co (0, 1, 2.5, 5, 10) scaffolds tended to increase with the extension of soaking time. At days 21 and 28, the weight loss of the Co (0, 1, 2.5, 5, 10) scaffolds also raised with increasing Co-doping concentration. Meanwhile, the weight loss of the Co (1, 2.5, 5, 10) scaffolds was higher than that of Co0 scaffolds, and the differences were significant (p < 0.01). Figure 4B shows that the pH value of Tri-HCL solutions in the Co (0, 1, 2.5, 5, 10) groups revealed a synchronously ascending state over the prolonged soaking time.
Figure 4C–E shows the surface morphology of the Co (0, 1, 2.5, 5, 10) scaffolds after mineralization in vitro. As displayed in Fig. 4D and E, the Co (0, 1, 2.5, 5) scaffolds showed apparent mineralization effects in vitro. A particulate and loose microstructure was consistently dispersed on the surface of the Co (0, 1, 2.5, 5) scaffolds, and cracks were observed. The XRD data (Fig. 4F) demonstrated that the mineralization phases of the Co (0, 1, 2.5, 5) scaffolds are hydroxyapatite. However, the Co10 scaffolds showed none of the changes mentioned above.
Mechanical properties of the Co (0, 1, 2.5, 5, 10) scaffolds
The mechanical properties of the Co (0, 1, 2.5, 5, 10) scaffolds were obtained by compressive strength testing. As shown in Fig. 4G, the compressive strength of the Co (1, 2.5, 5) scaffolds slightly decreased with increasing Co2+ ion content compared with the Co0 scaffolds; remarkably, the compressive strength of the Co10 scaffold sharply declined, and the differences between the Co10 group and the other groups were significant (p < 0.01).
Biocompatibility of the Co (0, 1, 2.5, 5, 10) scaffolds
Preliminary evaluation of the biocompatibility on the Co (0, 1, 2.5, 5, 10) scaffolds was detected by means of a cytotoxicity assay for subsequent biological studies. As shown in Fig. 5A, the cell viability in the Co (0, 1, 2.5, 5) groups was 103.31%, 103.03%, 106.17%, and 79.25%, which were all greater than 70% and did not show obvious cytotoxicity, while the cell viability in the Co10 group was 69.94%, which was less than 70% and displayed obvious cytotoxicity.
To further assess the biocompatibility of the Co (0, 1, 2.5, 5) scaffolds, the proliferation of rat BMSCs (rBMSCs) on the Co (0, 1, 2.5, 5) scaffolds was measured. As shown in Fig. 5B, the OD values in the Co (0, 1, 2.5, 5) groups increased with the extension of culture time, which indicated that rBMSCs cultured on the Co (0, 1, 2.5, 5) scaffolds proliferated well. In particular, after 24 h of incubation, there were no significant differences in cell proliferation among the Co (0, 1, 2.5, 5) scaffolds. At day 3, the proliferation of rBMSCs on the Co1 scaffold showed a significant difference from that on the Co (0, 2.5, 5) scaffolds, whereas cell proliferation on the Co (2.5, 5) scaffolds was less than that on Co0 scaffold. Interestingly, as the culture time increased, the Co2.5 scaffold effectively promoted rBMSCs proliferation at days 5 and 7 compared with the Co (0, 1, 5) scaffolds. However, the Co5 scaffold showed a slight decrease in cell proliferation compared with the other Co-doped scaffolds at each time point.
The adhesion and morphology of rBMSCs on the Co (0, 1, 2.5, 5) scaffolds were also observed. Figure 5C and D shows SEM images of rBMSCs at day 3 and confocal laser scanning microscope (CLSM) images of rBMSCs at days 3 and 6 after seeding on the Co (0, 1, 2.5, 5) scaffolds, respectively. As revealed by the SEM images in Fig. 5C, the rBMSCs on Co the (0, 1, 2.5) scaffolds attached and spread well with more abundant filopodia than those on the Co5 scaffold. No obvious distinction was noted in cell morphology on the Co (0, 1, 2.5) scaffold surfaces. Intriguingly, the CLSM results presented a similar trend as the Cell Counting Kit-8 (CCK-8) assay data and SEM images. In the CLSM results, cell skeleton was stained green, and the nucleus was stained blue. As shown in Fig. 5D, there were few rBMSCs on the surface of the Co (0, 1, 2.5, 5) scaffolds at day 3, and there were more rBMSCs at day 6, as observed by CLSM images. However, compared with the Co (0, 1, 2.5) scaffolds, the Co5 scaffold had slightly fewer cells at each time point.
In vitro angiogenic properties of the Co (0, 1, 2.5, 5) scaffolds
The angiogenic properties of the Co (0, 1, 2.5, 5) scaffolds were investigated by conducting an in vitro tubule formation assay. As shown in Fig. 6A, human umbilical vein endothelial cells (HUVECs) in all groups sprouted and self-assembled to form branched nodes and circles after 4 h of culture, finally forming continuous tubular networks. It is worth noting that the Co2.5 group induced the most tube-like structures among all of the groups. After 6 h of culture, the Co2.5 group maintained its advantage in stimulating tubule formation. Although the HUVECs began to undergo apoptosis after 16 h of culture, which was the case in all of the groups, the cells in the Co2.5 group displayed a better network structure than the cells in Co the (0, 1, 5) groups. Quantitative analysis of the branches number (Fig. 6B) and total branching length (Fig. 6C) showed significantly higher values in Co (1, 2.5, 5) groups than that in the Co0 groups, especially in the Co2.5 group.
HIF-1α and VEGF expression in rBMSCs incubated with Co (0, 1, 2.5, 5) extracts was also analyzed. As exhibited in Fig. 6D and E, both HIF-1α and VEGF expression was significantly upregulated in the Co (1, 2.5, 5) groups compared with that of the Co0 group. In particular, the Co2.5 group showed greatly enhanced expression of HIF-1α and VEGF in comparison with that of the Co 1 and Co 5 groups.
In vitro osteogenic properties of the Co (0, 1, 2.5, 5) scaffolds
The osteogenic properties of the Co (0, 1, 2.5, 5) scaffolds were measured by alkaline phosphatase (ALP) staining and alizarin red staining. As shown in Fig. 7A, after culture for 7 days, the intensity of ALP staining in the Co (0, 1, 2.5, 5) groups was similar to that in the osteogenic medium (OM) group. With prolonged culture time, the staining intensity in all groups on day 14 was higher than that on day 7. Moreover, the Co2.5 group showed more intense staining than the other groups. With the exception of the initial osteogenic marker, the terminal state of osteogenic differentiation of rBMSCs was examined via alizarin red staining. As displayed in Fig. 7B, alizarin red staining showed that a large number of red-stained calcium nodules were formed in all groups, and the calcium nodule numbers in the Co (1, 2.5) groups and OM group were significantly higher than those in the Co (0, 5) groups. To quantitatively evaluate the osteogenic properties of the Co (0, 1, 2.5, 5) scaffolds, the ALP activity and calcium content were analyzed. As revealed in Fig. 7C, the ALP activity in the Co (0, 1, 2.5, 5) groups was lower than that in the OM group over 7 days of culture. Additionally, the Co1 group enhanced ALP activity among the Co (0, 1, 2.5, 5) groups. Interestingly, the Co2.5 group exhibited markedly increased ALP activity compared to with the other groups over 14 days of stimulation, the difference was significant (p < 0.05). As shown in Fig. 7D, the group exposed to the Co2.5 extract had higher calcium contents than the groups exposed to the Co (0, 1, 5) extracts. Although the OM group had a higher calcium content, the differences between the Co2.5 and OM groups were not significant (p > 0.05).
The expression levels of osteogenic markers, including ALP, bone morphogenetic protein-2 (BMP-2), runt-related transcription factor 2 (RUNX2), and osteocalcin (OCN), in rBMSCs cultured in Co (0, 1, 2.5, 5) extracts at day 14 were also examined. As shown in Fig. 7E and F, the Co2.5 group exhibited a more significant upregulation of the expression of ALP and BMP-2 than the Co (0, 1, 5) groups at day 14. Although OCN expression did not show a significant difference among the Co (1, 2.5, 5) groups, OCN expression in the Co2.5 group was higher than that in the Co0 group. RUNX2 expression in the Co2.5 group at day 14 was also higher than that in the Co (0, 5) groups but slightly lower than that in the Co1 group, which showed no noticeable difference (p > 0.05).
In vivo bone regeneration of the Co2.5 scaffold
A cranium bone defect model with 5 mm in diameter in rats was constructed to validate the in vivo osteogenic properties of Co-doped scaffolds. Since the in vitro results revealed that the osteogenic and angiogenic properties of the Co2.5 scaffold were superior to those of the Co (0, 1, 5) scaffolds, the Co2.5 scaffold was applied to investigate whether it resulted in enhanced bone regeneration in vivo.
At 8 weeks after implantation, CT-constructed images were obtained to evaluate the extent of new bone formation by micro-CT. As shown in Fig. 8A, the round bone defect in the blank group clearly existed, and a small amount of bone-like tissue could be observed on the edge of the defect. In the Co0 group, scaffolds in bone defects were visible in situ, and the edge of the scaffold was covered by bone-like tissue. Notably, the scaffold in the Co2.5 group was almost integrated with the surrounding normal bone tissue and was not easily visible, and the scaffold–host bone interface was completely fused. In addition, as displayed in planar views (Fig. 8B), a larger amount of new bone along the margin and in the center of the defect was formed in the Co2.5 group compared with the Co0 and blank groups. Quantitative evaluation of new bone formation was performed using a bone volume/tissue volume (BV/TV) assay. As shown in Fig. 8C, the BV/TV was the highest in the Co2.5 group, followed by the Co0 group and blank group. Compared with that in the blank and Co0 groups, the BV/TV in the Co2.5 group was significantly increased (p < 0.01).
Following micro-CT analysis, the samples were performed histologically to assess new bone formation. The hematoxylin and eosin (H&E) staining data indicated a bone regeneration tendency parallel to the above micro-CT data. New bone tissue was stained with different shades of red in H&E staining. As revealed in Fig. 9A and B, new bone tissue was found in the center and along with the margin of the defect in the Co2.5 group, but was relatively rare in the Co0 group, with the least new bone formation in the blank group. Furthermore, the new bone area was evaluated quantitatively. As displayed in Fig. 9C, the new bone area was significantly increased in Co2.5 group, followed by the Co0 and blank groups, and the differences among groups were significant (p < 0.01). Additionally, Masson’s trichrome staining was implemented to histologically evaluate the maturity of new bone tissue. Masson’s trichrome staining showed that collagen fibers and immature bone tissue were blue, and mature bone tissue was red. As shown by Masson’s trichrome staining in Fig. 9D and E, in the blank group, only a small amount of collagen fibers and immature bone tissue were formed at the defect edge. Although the new bone tissue formed in the Co2.5 group and Co0 group was mainly mature bone tissue, the amount of new bone tissue in the Co2.5 group was significantly greater than that in the Co0 group.