Functionalized 3D-Printed ST2/Gelatin Methacryloyl/Polcaprolactone Scaffolds for Enhancing Bone Regeneration with Vascularization

Growth factors were often used to improve the bioactivity of biomaterials in order to fabricate biofunctionalized bone grafts for bone defect repair. However, supraphysiological concentrations of growth factors for improving bioactivity could lead to serious side effects, such as ectopic bone formation, radiculitis, swelling of soft tissue in the neck, etc. Therefore, safely and effectively applying growth factors in bone repair biomaterials comes to be an urgent problem that needs to be addressed. In this study, an appropriate concentration (50 ng/mL) of Wnt3a was used to pretreat the 3D-bioprinting gelatin methacryloyl(GelMA)/polycaprolactone(PCL) scaffold loaded with bone marrow stromal cell line ST2 for 24 h. This pretreatment promoted the cell proliferation, osteogenic differentiation, and mineralization of ST2 in the scaffold in vitro, and enhanced angiogenesis and osteogenesis after being implanted in critical-sized mouse calvarial defects. On the contrary, the inhibition of Wnt/β-catenin signaling in ST2 cells reduced the bone repair effect of this scaffold. These results suggested that ST2/GelMA/PCL scaffolds pretreated with an appropriate concentration of Wnt3a in culture medium could effectively enhance the osteogenic and angiogenic activity of bone repair biomaterials both in vitro and in vivo. Moreover, it would avoid the side effects caused by the supraphysiological concentrations of growth factors. This functionalized scaffold with osteogenic and angiogenic activity might be used as an outstanding bone substitute for bone regeneration and repair.


Introduction
There are more than 4 million patients with bone defects caused by severe trauma, cancer, and infection every year in the world [1]. Especially, bone defects exceeding the critical size cannot be repaired or regenerated naturally, and therefore, transplantation of bone or bone repair graft is required. Both autologous and allogeneic transplantation is limited by resource shortage and donor site complications, while allogeneic and xenotransplantation also have issues of cost and risk of disease transmission [2]. These problems have prompted the development of new bone repair biomaterials for bone defect repair. However, the insufficient bioactivity of implanted biomaterials is prone to cause repair failure and require revision of these implants [3], resulting in higher costs and shorter survival [4]. Therefore, improving the bioactivity of biomaterials emerged as a key problem that needs to be solved urgently in bone tissue engineering.
In recent years, bone repair scaffolds fabricated by implant materials such as metal (stainless steel, titanium, alloy, etc.), ceramic, or polymer are functionalized and modified by loading inorganic salts, growth factors, or stem cells to create an osteogenic environment to improve bioactivity [5]. Bioactive inorganic salts were used to simulate the osteogenic differentiation environment of stem cells, such as borate [6], bioglass [7], and calcium silicate [8], and combined with 3D-bioprinting loaded stem cells or vascular endothelial cells to construct controllable shape and structure to simulate the internal structure of bone [9,10]. Moreover, two growth factors of bone morphogenetic proteins (BMPs), BMP-2 and BMP-7, have been approved by the FDA for clinical use for more than 20 years. Grafts containing an absorbable collagen sponge and BMP-2 or BMP-7 are referred to as the Infuse device and the OP-1 device, respectively. There is little debate about their osteogenic potential [11]. However, clinical success is limited by their low stability, short half-life, and rapid diffusion from the delivery site. Supraphysiological concentration of the growth factor is often required to demonstrate efficacy [12,13]. It was reported that BMP-2 exhibits a half-life time of~7 min in the bloodstream [14]. Effective concentrations are typically in the mg/mL (∼10 −5 M) range, which is much more than native BMP-2 concentrations of 18.8-22 pg/mL (7.2 × 10 −13 to 8.5 × 10 −13 M) [12,15]. This can cause serious adverse side effects, such as ectopic bone formation, radiculitis, cervical and soft tissue swelling, etc. [12]. Consequently, BMP-7 was withdrawn from the market, and restrictions were imposed on the clinical use of BMP-2, generating the need for the development of a more efficacious treatment strategy [16].
Bone marrow stromal cells (BMSCs) are often used as seed cells in bone tissue engineering [17]. The Wnt/β-catenin signaling pathway is a key regulator of bone homeostasis. Activation of the canonical Wnt/β-catenin signaling pathway leads to osteogenic differentiation, while β-catenin inactivation leads to adipogenic or chondrogenic differentiation [18,19]. Osteogenic gene expression and mineral deposition of bone marrow stromal cell line ST2 grown on the surface of titanium/titanium alloy were dramatically increased by activating the Wnt signaling with a chimeric peptide [20]. Studies based on the synergistic induction of angiogenesis and bone formation by multiple cell cultures have been practiced for many years. Vascularization provides blood supply to the bone defect, which carries nutrients and oxygen to improve the bioactivity of the bone repair scaffold, resulting in vascularized functional bone formation [21,22]. In the co-culture system of BMSCs and human umbilical vein endothelial cells (HUVECs) on 3D bone scaffolds, HUVECs induced osteogenic differentiation of BMSCs by activating its Wnt/β-catenin signaling [23]. However, as the number of components in a scaffold system increases, so does the complexity of construction and uncertainty in treatment outcomes [24].
It was reported that the efficacy of autografts can be traced back to BMSCs residing within the bone graft. Blocking the endogenous Wnt signaling using Dkk1 abrogates autograft efficacy whereas a brief incubation in Wnt3a reliably improves autologous bone grafting efficacy [25]. It suggests that functional modification of the implant material by activating the canonical Wnt/β-catenin signaling pathway of bone marrow stromal cells by Wnt3a in vitro can improve its bioactivity and increase the efficacy of bone regeneration while avoiding the use of supraphysiological concentrations in vivo and the side effects caused thereby. Moreover, after incubation with Wnt3a for 24 h, macrophages induced in vitro tubular pattern structures in endothelial cells resembling capillary-like vasculature via releasing angiogenic growth factor [26]. Intranasally delivered Wnt3a improves vascularized recovery after traumatic brain injury by up-regulation of the expression of VEGF [27]. Wnt3a is involved in the mechanical loading on the improvement of bone formation and angiogenesis [28]. These results provide strong evidence that Wnt3a promotes bone formation and vascularization.
As autografts are limited by resource shortage, our work aims to utilize the properties of Wnt3a and autografts bone to construct an osteogenic and angiogenic bifunctional scaffold, which could achieve stable batch production by 3D printing to promote bone regeneration. In our previous study, we found that Wnt3a can promote osteogenic differentiation of bone marrow stromal cell line ST2 and up-regulation of the expression of VEGF-A, while inhibition of Wnt signaling decreased the expression of osteogenic differentiation marker and VEGF-A ( Figure 1). Therefore, by combining gelatin methacryloyl (GelMA) and polycaprolactone (PCL), we envisaged the batch fabrication of ST2/GelMA/PCL scaffolds by 3D bioprinting to promote bone repair after Wnt3a pretreatment in vitro. To verify (GelMA) and polycaprolactone (PCL), we envisaged the batch fabricatio ST2/GelMA/PCL scaffolds by 3D bioprinting to promote bone repair after Wnt3a pr ment in vitro. To verify the potential of these ST2/GelMA/PCL scaffolds as bone engineering biomaterials, the ability of cell survival, proliferation, osteogenic differ tion, and mineralization in ST2/GelMA/PCL scaffolds were detected. Then, the sca were implanted into mice parietal critical-size bone defects to detect bone regene and vascular formation. These specific cellular responses are key steps in the develo of functional 3D-printed bone scaffolds to promote vascularized functional bone re ation.

Wnt3a Promotes Osteogenic Differentiation and VEGF-A Expression of ST2 Cells in Two-Dimensional Culture
Alkaline phosphatase is an important marker of osteoblast differentiation and can be measured by alkaline phosphatase staining and quantitative real-time polymerase chain reaction (qRT-PCR) [29]. According to the results of different concentrations, the alkaline phosphatase staining in the Wnt3a group had the darkest positive staining at the concentration of 50 ng/mL ( Figure 1A), and the ICG-001 group had the lightest staining at 20 µM ( Figure 1B). The expression of Alp and angiogenic factor VEGF-A was consistent with the trend of alkaline phosphatase staining ( Figure 1C,D). Therefore, in the subsequent experiments, the concentration of Wnt3a was selected as 50 ng/mL, and the concentration of ICG-001 was selected as 20 µM. The above results suggest that activation of Wnt signaling promotes osteogenic differentiation and vascular marker expression in ST2 cells, whereas inhibition of Wnt signaling reduces this function.

Preparation of ST2/GelMA/PCL Scaffolds by 3D Bioprinting and Their Biocompatibility Testing
The biocompatibility of bone repair scaffolds is the basis for their biological functions and is usually characterized by cell survival and proliferation [30]. Three-dimensional bone repair scaffolds of ST2/GelMA/PCL were prepared by 3D bioprinting (Figure 2A). At 1, 4, and 7 days after scaffold printing, live cells were above 92% of total cells in Wnt3a pretreated scaffolds ( Figure 2B,C). The proliferation number of cells in the 3D-printed functional modules increased with time, and the proliferation activity of cells in the control group increased by 4.12 times in 7 days. Among the three groups, the Wnt3a group had the greatest proliferative activity. That was 1.57 times and 2.62 times more than the control group and ICG-001 group, respectively, on the seventh day ( Figure 3A,B).

Wnt3a Pretreating ST2/GelMA/PCL Bone Repair Scaffold Enhanced Osteogenic Differentiation
Alkaline phosphate staining, quantification of alkaline phosphate biochemical activity, and osteogenic marker mRNA were used to detect the level of osteogenic differentiation. On the seventh day of scaffold culture, compared with the control group, the positive staining of alkaline phosphatase in the Wnt3a group was the deepest, followed by the control group, and the weakest in the ICG-001 group ( Figure 4A). Compared with the control group, the alkaline phosphatase activity of the Wnt3a group (Wnt3a group vs. ctrl group: (0.57 ± 0.07) nmol/µg/min vs. (0.23 ± 0.04) nmol/µg/min, p < 0.05) was significantly increased, while it was significantly decreased in the ICG-001 group ((0.08 ± 0.02) nmol/µg/min, p < 0.05) ( Figure 4B). qRT-PCR detection showed that the expression of β-catenin (the key factor of Wnt signaling) and osteogenic differentiation genes (Alp, Osx, Runx2, Col1a1, and Ocn) were up-regulated in the Wnt3a group, but decreased in the ICG-001 group, compared with the control group ( Figure 4C,D). These suggest that activating Wnt signaling enhanced the osteogenic differentiation of ST2 cells in the scaffold while inhibiting Wnt signaling decreased the osteogenic differentiation.

Wnt3a Pretreating Enhances the Mineralization of ST2 Cells in Bone Repair Scaffolds
Alizarin red S staining for calcium salt nodules usually reflects mineralization and osteogenic activity [31]. After 21 days of osteogenic induction culture, compared with the control group, the alizarin red S staining was enhanced in the Wnt3a group and decreased in the ICG-001 group ( Figure 5A). Quantitative analysis showed that calcium deposition in the Wnt3a group was 3.26 times higher than that in the control group (p < 0.05) and 8.63 times of the ICG-001 group (p < 0.05) ( Figure 5B). These results suggest that activation of Wnt signaling promoted ST2 cells' mineralization, whereas inhibition of Wnt signaling decreased ST2 cells' mineralization.

Wnt3a Pretreated Bone Repair Scaffold Accelerates Critical-Size Bone Defect Repair
The critical-size bone defect of mice parietal bone was used to verify the repair effect of bone tissue engineering scaffolds in vivo [32]. Eight weeks after surgery, compared with the control group, µCT results showed that the Wnt3a group had the best bone repair effect, and the ICG-001 group had the weakest effect ( Figure 6A,C). No ectopic bone formation was seen on the HE staining of the tissue section at the bone defect position ( Figure 6B). Compared with the control group, the bone tissue volume fraction (BV/TV) in the Wnt3a group (Wnt3a group vs. ctrl group: (22.43 ± 3.07)% vs. (8.48 ± 0.91)%, p < 0.05), and osteoblast number per tissue area (N.Ob/T.Ar) in the Wnt3a group (Wnt3a group vs. ctrl group: (7.13 ± 0.93) 10 −2 /mm 2 vs. (3.78 ± 0.95) 10 −2 /mm 2 , p < 0.05) increased significantly. However, BV/TV (3.35 ± 0.52)% and N.Ob/T.Ar (1.73 ± 0.59) 10 −2 /mm 2 in the ICG-001 group were significantly lower than those in the control group (p < 0.05) ( Figure 6D). These suggest that the functionalized ST2/GelMA/PCL bone repair scaffold promoted the repair of parietal critical-size bone defects in mice.

Wnt3a Pretreating ST2/GelMA/PCL Bone Repair Scaffold Enhanced Osteogenic Differentiation
Alkaline phosphate staining, quantification of alkaline phosphate biochemical ity, and osteogenic marker mRNA were used to detect the level of osteogenic differe tion. On the seventh day of scaffold culture, compared with the control group, the po staining of alkaline phosphatase in the Wnt3a group was the deepest, followed b control group, and the weakest in the ICG-001 group ( Figure 4A). Compared wit control group, the alkaline phosphatase activity of the Wnt3a group (Wnt3a group v group: (0.57 ± 0.07) nmol/μg/min vs. (0.23 ± 0.04) nmol/μg/min, p < 0.05) was signific increased, while it was significantly decreased in the ICG-001 group ((0.08 ± nmol/μg/min, p < 0.05) ( Figure 4B). qRT-PCR detection showed that the expression catenin (the key factor of Wnt signaling) and osteogenic differentiation genes (Alp, Runx2, Col1a1, and Ocn) were up-regulated in the Wnt3a group, but decreased i ICG-001 group, compared with the control group ( Figure 4C,D). These suggest

Wnt3a Pretreated Bone Repair Scaffolds Promote Vascularization
Eight weeks after surgery, there was a large amount of neovascularization in the Wnt3a group compared with the control group in bone tissue sections ( Figure 7A). Quantitative analysis of blood vessels showed that the Wnt3a group was 2.39 times (p < 0.05) higher than that of the control group and 4.78 times higher than that of the ICG-001 group (p < 0.05) ( Figure 7B). Furthermore, Wnt3a increased the amount of VEGF-A in cells ( Figure 7E) and conditioned medium ( Figure 7F) of the ST2/GelMA/PCL scaffolds, which promoted tubule formation of HUVEC cells in vitro ( Figure 7C,D). It is shown that the functionalized ST2/GelMA/PCL bone repair scaffold promotes vascularized functional bone formation.
activating Wnt signaling enhanced the osteogenic differentiation of ST2 cells in th fold while inhibiting Wnt signaling decreased the osteogenic differentiation.

Wnt3a Pretreating Enhances the Mineralization of ST2 Cells in Bone Repair Scaffold
Alizarin red S staining for calcium salt nodules usually reflects mineralizati osteogenic activity [31]. After 21 days of osteogenic induction culture, compared w control group, the alizarin red S staining was enhanced in the Wnt3a group and dec in the ICG-001 group ( Figure 5A). Quantitative analysis showed that calcium dep in the Wnt3a group was 3.26 times higher than that in the control group (p < 0.05) a times of the ICG-001 group (p < 0.05) ( Figure 5B). These results suggest that activa Wnt signaling promoted ST2 cells' mineralization, whereas inhibition of Wnt sig decreased ST2 cells' mineralization.

Wnt3a Pretreated Bone Repair Scaffold Accelerates Critical-Size Bone Defect Repair
The critical-size bone defect of mice parietal bone was used to verify the repair effect of bone tissue engineering scaffolds in vivo [32]. Eight weeks after surgery, compared with the control group, μCT results showed that the Wnt3a group had the best bone repair effect, and the ICG-001 group had the weakest effect ( Figure 6A,C). No ectopic bone formation was seen on the HE staining of the tissue section at the bone defect position ( Figure  6B). Compared with the control group, the bone tissue volume fraction (BV/TV) in the Wnt3a group (Wnt3a group vs. ctrl group: (22.43 ± 3.07)% vs. (8.48 ± 0.91)%, p < 0.05), and osteoblast number per tissue area (N.Ob/T.Ar) in the Wnt3a group (Wnt3a group vs. ctrl group: (7.13 ± 0.93) 10 −2 /mm 2 vs. (3.78 ± 0.95) 10 −2 /mm 2 , p < 0.05) increased significantly. However, BV/TV (3.35 ± 0.52)% and N.Ob/T.Ar (1.73 ± 0.59) 10 −2 /mm 2 in the ICG-001 group were significantly lower than those in the control group (p < 0.05) ( Figure 6D). These suggest that the functionalized ST2/GelMA/PCL bone repair scaffold promoted the repair of parietal critical-size bone defects in mice.

Discussion
For standard bone tissue engineering, three elements are required, i.e., a scaffo growth factors, and seed cells. BMSCs are the most used seed cells, and the regenerati strategy should effectively promote their osteogenic differentiation and angiogenesis. this study, ST2/GelMA/PCL scaffolds were fabricated by 3D printing and pretreated w Wnt3a (50 ng/mL) for 24 h in vitro. Based on the data of cell proliferation, osteogenic ge expression, mineralization, μCT, and histology analysis, we demonstrated that Wnt pretreatment increased the proliferation, osteogenic differentiation, and mineralization ST2 cells in the scaffolds in vitro, and promoted bone regeneration and angiogenesis

Discussion
For standard bone tissue engineering, three elements are required, i.e., a scaffold, growth factors, and seed cells. BMSCs are the most used seed cells, and the regeneration strategy should effectively promote their osteogenic differentiation and angiogenesis. In this study, ST2/GelMA/PCL scaffolds were fabricated by 3D printing and pretreated with Wnt3a (50 ng/mL) for 24 h in vitro. Based on the data of cell proliferation, osteogenic gene expression, mineralization, µCT, and histology analysis, we demonstrated that Wnt3a pretreatment increased the proliferation, osteogenic differentiation, and mineralization of ST2 cells in the scaffolds in vitro, and promoted bone regeneration and angiogenesis in vivo. Simply implanting materials or inoculating cells on the surface of materials before transplantation, it is difficult to make cells grow into materials. Unlike that, 3D bioprinting generates customized macroporous implants for biomaterials and cells to simulate a 3D growth environment and make cells orientated and evenly distributed in scaffolds [33]. In our scaffolds, PCL and GelMA have high biocompatibility, and PCL has been approved for clinical use by FDA [34,35]. ST2 cells survival and proliferation were evenly distributed in the structure, with a high survival rate (above 92%) and proliferation rate.
When analyzed for their osteoinductive potential in vitro, ST2/GelMA/PCL bone repair scaffolds were pretreated with the growth factor Wnt3a in vitro, as previously reported in the Wnt-mediated BMSCs stimulation model [25,36]. We hypothesized that ST2/GelMA/PCL scaffolds pretreated with Wnt3a in vitro were sufficient to trigger osteogenic responses. First, the canonical Wnt/β-catenin signaling pathway is activated in bone marrow stromal cells leading to bone regeneration after trauma [37]. Second, Wnt3a pretreatment of ST2 cells for 24 h was sufficient to enhance the expression of osteoblast transcription factors (Runx2 and Osx) and the formation of mineralization. Similarly, Aquino-Martinez et al. also found that Wnt3a could enhance the expression of Runx2 and Osx in bone marrow stem cells [36]. Finally, ICG-001 reduced the expression of osteogenic differentiation transcription factors in ST2 cells. Zhou et al. reported that the inhibition of Wnt/β-catenin by ICG-001 significantly decreased osteogenic differentiation of bone marrow stromal cells [38]. These results suggest that Wnt/β-catenin is an essential factor in maintaining the osteogenic differentiation of ST2 cells.
Here, we have employed the well-defined mice parietal critical-size bone defect model to evaluate Wnt3a-pretreated ST2/GelMA/PCL scaffolds functionality in vivo. Histological analysis showed significant increases in bone volume per tissue volume (BV/TV), and osteoblast number per tissue area (N.Ob/T.Ar) in the Wnt3a group than in the control group. In addition, according to literature reports, the concentration of Wnt3a in serum is about 1-2 ng/mL [39,40], while the Wnt3a concentration of 25-100 ng/mL is often used in the experiments of inducing osteogenic differentiation in vitro [41][42][43]. However, 40,000 ng/mL of Wnt3a was required for promoting the repair of parietal bone defects in vivo [44]. Different from the direct use of Wnt3a protein for bone repair, ST2/GelMA/PCL scaffolds were pretreated with 50 ng/mL Wnt3a for 24 h in vitro and washed with phosphatebuffered saline before transplantation in this study. This means facilitating the bone formation of critical-size bone defects and avoiding supraphysiological concentration of growth factors used in vivo.
Vascularization is directly related to the final success of newly regenerated bone [45]. McBride et al. reported that bone marrow stromal cells promote angiogenesis through Wnt3a [46]. Activation of Wnt/β-catenin signaling in mesenchymal stem cells is critical for angiogenesis, while the Wntβ-catenin inhibitor ICG-001 can reverse this process [47]. VEGF effectively promotes the coupling of angiogenesis and bone formation in bone repair. Activation of the Wnt signaling pathway leads to up-regulation of VEGF expression, which induces neovascularization, while inhibition of the VEGF signaling pathway inhibits angiogenesis [48]. In this study, VEGF-A expression was up-regulated in Wnt3a pretreated ST2 cell-loaded scaffolds, whose conditioned medium promoted vascular endothelial tubule formation in vitro, and scaffold implantation in bone defects promoted vascularized functional bone formation in vivo, while Wnt/β-catenin signaling inhibitor ICG-001 reduced tubule formation in vitro and angiogenesis in vivo. Dual cell implantation-for example, mesenchymal stromal cells and endothelial progenitor cells-promote bone formation and angiogenesis synergistically [49]. However, as the number of components in the scaffold increases, so does the complexity of the procedure and uncertainty about the treatment outcome [24]. The method reported in this study showed that the single-cell type (ST2) scaffolds induced by Wnt3a in vitro promoted bone formation and angiogenesis, making clinical transformation easier.

Cell Culture
To determine the appropriate concentration of Wnt3a-induced osteogenic differentiation of ST2 cells, ST2 cells (ATCC, Manassas, VA, USA) were expanded in the basal medium (α-MEM medium containing 10% fetal bovine serum (BI, Herzliya, Israel), 100 u/mL penicillin G, and 100 mg/mL streptomycin (Gibco, Gaithersburg, MD, USA)). ST2 cells were planted with 1 × 10 4 /well in 24-well plates and cultured at 37 • C with 5% CO 2 for 4 h until the cells adhered to the wall. Then, the cells were incubated with different concentration of Wnt3a (0, 25, 50, 100 ng/mL) (R & D Systems, Minneapolis, MN, USA), or ICG-001 (0, 5, 10, 20 µM) (MCE, Dallas, TX, USA) for 24 h. After being rinsed with phosphate-buffered saline (PBS) 3 times, the cells were cultured with the basal medium, and half of the medium was changed every 2 days. On the seventh day, total RNA was extracted and the expression of osteogenic marker Alp and angiogenic factor VEGF-A was detected by qRT-PCR. Methods and primer sequences are described later. The osteogenic differentiation and expression of VEGF-A in ST2 cells were significantly promoted by Wnt3a at 50 ng/mL, while significantly inhibited by ICG-001 at 20 µM. Thus, these concentrations were selected for subsequent experiments.

3D-Bioprinting and Culture of ST2/GelMA/PCL Scaffolds
GelMA (Sunpbiotech, Beijing, China) was fully dissolved in α-MEM medium to create a solution at the concentration of 20% (w/v). Then, ST2 cells were trypsinized and resuspended in the basal medium. After mixing them, lithium phenyl-2, 4, 6-trimethylbenzoylphosphinate (LAP) (Sunpbiotech, Beijing, China) was added. The final concentration of ST2 cells, GelMA, LAP was 10 6 /mL, 10% (w/v) and 0.25% (w/v), respectively. PCL (Sigma, Hesse, Germany) was sterilized by UV irradiation for 1 h in advance. Hard biomaterials and a cell-integrated 3D system were designed by the Laboratory of Bone Development and Regeneration, Chongqing Medical University, referring to previously reported work [21]. The spiral nozzle loaded with PCL material and the bio-cartridge loaded with ST2 were controlled by the computer to move in the X, Y, and Z axis directions, respectively. Before printing, PCL was put in a molten screw extrusion nozzle and melted at 95 • C for 10 min. At the same time, the ST2-loaded GelMA solution was placed into a syringe and pre-cooled at 4 • C for 5 min, and then the syringe was put in the pneumatic extrusion nozzle for the following printing. Then, the melted PCL was plotted at 300 µm diameter at a printing speed of 3 mm/s. As shown in Figure 2A, the PCL was printed as the frame structure of the scaffold, and then the ST2-loaded GelMA solution was plotted between the PCL strips with a 400 µm interval between cell bundles at a printing speed of 5 mm/s. Each printed layer was irradiated with blue light at 405 nm for 5 s to stabilize and solidify. After printing, the scaffolds were soaked in sterile PBS for 1 min and repeated 3 times to remove excess LAP photoinitiator. The scaffolds were cultured in the basal medium of 10% fetal bovine serum, 100 u/mL penicillin G, and 100 mg/mL streptomycin at 37 • C, 5% CO 2 for 4 h, and then incubated with Wnt3a (50 ng/mL, final concentration) or ICG-001 (20 µM, final concentration) for 24 h. Next, the scaffolds were rinsed 3 times with PBS before implantation in vivo or subsequent culture in vitro. Half of the medium was changed every two days.

Cell Viability
Cell viability in Wnt3a pretreated scaffolds was determined on days 1, 4, and 7 using the LIVE/DEAD kit (ThermoFisher Scientific, Waltham, MA, USA) as previously described [50]. Briefly, the supernatant was discarded by aspiration, and the scaffolds were washed twice with PBS. Each scaffold was completely soaked in 300 µL LIVE/DEAD staining solution and incubated at 37 • C with 5% CO 2 for 1 h. Images were taken with a fluorescence microscope (Leica, DMI8, Wetzlar, Germany). Live cells were stained green, while dead cells were stained red. The percentage of viable cells to total cells was calculated with ImageJ software v1.8.0.

Cell Proliferation
The proliferation of cells in the scaffolds was detected on days 1, 4, and 7 with a CCK-8 kit (MCE, Dallas, TX, USA) as previously described [51]. Before detection, ST2 cells were labeled with 5 µL/mL Dio dye and incubated at 37 • C with 5% CO 2 for 2 h. Next, they were washed 3 times with PBS to remove excess dye. After taking photos quickly by fluorescence microscope, 1 mL of CCK-8 dilution solution was added to each scaffold and incubated at 37 • C with 5% CO 2 for 2 h. An enzyme plate analyzer was used to detect the absorbance of the incubation solution at 450 nm.

Cell Osteogenic Differentiation Assays
On the seventh day of culture, the scaffolds were stained by BCIP/NBT alkaline phosphatase chromogenic kit (Beyotime Biotechnology, Shanghai, China) as previously described [52]. Cell lysates were collected, and alkaline phosphatase activity was detected with an alkaline phosphatase assay kit (Beyotime Biotechnology, Shanghai, China). The ultrasonic breaker was used to break the cells, centrifuge (12,000× g, 5 min), and collect the supernatant. The BCA protein method was used to measure the absorbance at a wavelength of 562 nm, and the protein content of the sample was calculated according to the standard curve. At the same time, an alkaline phosphatase kit was used to measure the absorbance at a wavelength of 405 nm, and the action time was recorded. The amount of AP substance was calculated based on the standard substrate concentration curve. Finally, the relative activity of AP was calculated through the concentration of protein and AP, and the duration of action.

Mineralization Assay
Matrix mineralization in ST2/GelMA/PCL scaffolds was analyzed by alizarin red S (ARS) (Sigma, Burlington, MA, USA) staining, referring to the methods reported in previous studies [53]. After the seventh day of culture, the scaffolds were incubated with osteogenic induction medium (basal medium containing 50 µg/mL ascorbic acid, 5 mM sodium β-glycerophosphate, 1 nM dexamethasone) for 21 days, and half of the medium was changed every two days. Scaffolds were then stained with 1% alizarin red S staining solution. Mineralization was quantified according to the reported method [54]. Briefly, the solution was collected after the scaffolds were incubated with 10% acetic acid for 12 h, and then neutralized with 10% ammonium hydroxide. The absorbance of the solution was measured at a wavelength of 405 nm.

Animal Experiments
The mice parietal critical-size bone defect model was used to evaluate the bone repair effect of the scaffold in vivo [55]. Six-week-old male C57BL/6 mice were used for animal experiments. Mice were fed with a regular diet, received water ad libitum, and were maintained on a 12-h light/dark cycle. Mice were randomly divided into 2 groups with 5 mice in each group: (1) ctrl (lift) and Wnt3a (right); (2) Wnt3a (lift) and ICG-001 (right). Briefly, mice were anesthetized with isoflurane. A trephine (MDJ090729, TUUME, China) was used to create a 5 mm diameter critical-size bone defect on both sides of the skull under low-speed drilling. A large saline flush was used to lower the temperature, and the surgery must be performed carefully to avoid damage to the dura and brain. The scaffolds were then implanted in the defect and the incision was sutured. After the operation, the mice were subcutaneously injected with tramadol (0.1 mg·g −1 ) to relieve pain. All animal experiments were approved by the Animal Care and Use Committee of Chongqing Medical University, and in accordance with institutional animal guidelines.

Analysis of Bone Regeneration In Vivo
Specimens from each group were taken 8 weeks after implantation. Mice in each group were sacrificed by cervical dislocation after inhaling CO 2 . The entire skull including the implants was surgically removed and fixed with 4% paraformaldehyde at 4 • C for 24 h. The parietal bone was scanned with nanoScan CT (Mediso, Budapest, Hungary) at 50 kV energy and 165 ms intensity. The PMOD (Zurich, Switzerland) software (version 4.3, threshold value, 500) [56] was used for the reconstruction of regenerated bone tissue. The area of interest was defined as a dotted circle with a diameter of 5 mm. The new bone formation was evaluated by the ratio of bone area and tissue area (B.A./T.A.). Then, specimens were decalcified in 10% EDTA (pH 7.4) for 21 days. After dehydration, the specimens were embedded in paraffin and cut into 5 µm thick tissue sections, which were stained with hematoxylin and eosin (H&E) to observe the healing and internal defects of different groups. The number of osteoblasts in the new bone and the area of the blood vessels containing red blood cells (RBCs) was counted by OsteoMeasure™ software (v4.1.0.2, OsteoMetrics, Knoxville, TN, USA). The terms and units used are those recommended by the Organizational Morphometric Nomenclature Committee of the American Society for Bone and Mineral Research [57].

Tubular Formation Experiment
According to a previously reported method [58], ST2/GelMA/PCL scaffold-induced tubule formation in HUVECs was performed in vitro. HUVECs (2 × 10 5 cells/cm 2 ) were mixed with 4-fold diluted cell supernatants and seeded on matrix (200 µL)-coated 24-well plates for 6 h. Capillary-like structures were counted with a microscope. The tube-forming network of HUVECs was quantified with Image-J.

ELISA
Angiogenic factor (VEGF-A) in scaffold culture supernatants was detected by ELISA. Sample supernatants were centrifuged at 1000× g for 5 min to remove cellular debris, and VEGF-A in scaffold culture supernatants was accurately measured using a mouse ELISA kit according to the manufacturer's instructions. The absorbance at 450 nm was measured by a Bio-Rad Microplate Analyzer (Hercules, CA, USA). A standard curve for the assay was established using cytokine standards provided by ELISA kits (Beyotime Biotechnology, Shanghai, China).

Statistical Analysis
Data were analyzed using GraphPad software (Chicago, IL, USA). One-way ANOVA was followed by Tukey's post hoc test. The final results are presented as the mean ± standard deviation of 3-5 experiments per group. For all tests, p-values less than 0.05 were considered statistically significant.

Conclusions
In conclusion, the 3D-printed ST2/GelMA/PCL bone repair scaffold pretreated with Wnt3a for 24 h in vitro was used to fabricate a single-cell scaffold with dual functions of osteogenesis and angiogenesis in this study, which could realize the mass and stable production of functional materials for bone regeneration and improve bone regeneration and angiogenesis in animal models with bone defects. It avoids the use of supraphysiological concentrations of growth factors in vivo and the resulting side effects, thus providing an outstanding bone substitute for bone regeneration and repair.