Biocompatible Customized 3D Bone Scaffolds Treated with CRFP, an Osteogenic Peptide

Background: Currently used synthetic bone graft substitutes (BGS) are either too weak to bear the principal load or if metallic, they can support loading, but can lead to stress shielding and are unable to integrate fully. In this study, we developed biocompatible, 3D printed scaffolds derived from µCT images of the bone that can overcome these issues and support the growth of osteoblasts. Methods: Cylindrical scaffolds were fabricated with acrylonitrile butadiene styrene (ABS) and Stratasys® MED 610 (MED610) materials. The 3D-printed scaffolds were seeded with Mus musculus calvaria cells (MC3T3). After the cells attained confluence, osteogenesis was induced with and without the addition of calcitonin receptor fragment peptide (CRFP) and the bone matrix production was analyzed. Mechanical compression testing was carried out to measure compressive strength, stiffness, and elastic modulus. Results: For the ABS scaffolds, there was a 9.8% increase in compressive strength (p < 0.05) in the scaffolds with no pre-coating and the treatment with CRFP, compared to non-treated scaffolds. Similarly, MED610 scaffolds treated with CRFP showed an 11.9% (polylysine pre-coating) and a 20% (no pre-coating) increase (p < 0.01) in compressive strength compared to non-treated scaffolds. Conclusions: MED610 scaffolds are excellent BGS as they support osteoblast growth and show enhanced bone growth with enhanced compressive strength when augmented with CRFP.


Introduction
The development of bone graft substitutes (BGS) is an essential field of orthopedic research [1]. BGS are usually used to fill the bone defects formed as a result of surgical procedures, fractures, or oncology [2][3][4]. In the majority of cases, the BGS used are either autografts or allografts. Historically, autografts have been considered the gold standard in bone grafting, as they have a lower probability of disease transfer, are osteoconductive, and allow for osteogenesis [5,6]. However, an autograft has many disadvantages, as the graft is extracted from another site in the body, resulting in excess pain in the donor site, surgical and wound complications, low graft availability, and increased hospital stay [7][8][9][10]. Allogenic bone grafts or allografts are the most commonly used BGS, as they are also osteoconductive and avoid donor site complications. However, they have their range of disadvantages, such as lack of uniformity, a higher chance of antigen response and disease transfer, and limited supply [11,12]. Additionally, sterilized allografts are usually weaker than healthy bone, increasing the graft site's risk of failure and reinjury [7,13].
Synthetic BGS are being developed to deal with the complications of autografts and allografts [14]. One example is demineralized bone matrix (DBM) which is derived by removing the mineral components of the allogenous bone with acid, leaving behind (<20 µm) in comparison to conventional FDM printing (200 µm). This material has great potential to enhance the flexibility and accuracy in developing a BGS with trabecular structural patterns as well as improve precision in the scaffold production.
In this study, we developed and evaluated the biocompatibility, osteoconductivity, and biomechanics of ABS and MED610 bone scaffolds inspired by natural bone's structural physiology, seeded with MC3T3 stem cells, and treated with CRFP.

Design and Fabrication of Cylindrical Scaffolds
A femur of a skeletally mature rat was computerized tomography (CT)-scanned (µCT40, Scanco medical) at an isotropic voxel resolution of 10 µm. The resulting CT imaging files (DICOM files) were converted into a 3D stereolithography model (STL file) by thresholding and combining the DICOM files using InVesalium software (Renato Archer Information Technology Center, Brazil). This 3D representation was processed in Geomagic ® DesignX software. Regions of interest were identified to extract the cortical bone structure (R1) from the diaphysis and the trabecular structure (R2) at the metaphysis of the femur assuming a segmental defect model, as illustrated in Figure 1b,d These structures (R1 and R2) were converted into a computer-aided design model (CAD file) using a non-uniform rational B-spline (NURBS)-based algorithm and uploaded into Solidworks ® software. We designed a cylindrical scaffold (5 mm height × 5 mm diameter) with internal structures created from the trabecular patterns extracted from R2 and four pillar structures with equal spacing on the outer rim of the cylinder, having dimensions (1 mm thickness × 1.5 mm length × 5 mm height) derived from the cortical bone thickness in R1. We carried out many reiterations by extracting different trabecular features and porosity and finalized our design based on the structural consistency of the 3D-printed result. Our final design had a porosity of 50% to facilitate penetration of the cells into the scaffold. Three-dimensional printing gives us the flexibility to change the number of pillars or the density of the trabecular structures based on the location and the strength requirements in the implantation site. Using the finalized design, we 3D-printed ABS scaffolds using FDM fabrication (Cube Pro Duo printer, 3D Systems-200 µm resolution, Rock Hill, SC, USA), and MED610 scaffolds were fabricated using polyjet 3D printing (Objet 30 Prime-16 µm resolution, Stratasys ® , Eden Prairie, MN, USA) [46]. We also developed planar (2D) scaffolds (10 mm × 10 mm) of MED610 material to test the osteoconductivity of the material surface.

Preparation of the Scaffolds
The ABS scaffolds were washed with deionized water for 24 h, followed by sterilization in an increasing concentration of ethanol solution (50%, 70%, 80%, 90%, and 100%) for 15 min and air-drying in a cell culture hood. The MED610 scaffolds were also washed in deionized water. However, sterilization was done by autoclaving at 132 • C for 4 min as per the manufacturer's recommendation.
We pre-coated the scaffolds by treating them with 0.01% polylysine solution for 2 h, rinsing them with sterile Dulbecco's phosphate buffer saline (DPBS), and air-drying for 1 h in a sterile cell culture hood. Polylysine, a known contributor to osteoconduction, has been shown to help the cells to attach to the scaffold surface [49,50].

Seeding MC3T3 Cells
After the sterilization and pre-coating (in some cases), the scaffolds were attached to the bottom surface of the cell culture plate using sterilized vacuum grease to prevent them from floating during the cell culture process. MC3T3 cells extracted from the C3 vertebra of a mouse with a density of 1 × 10 3 cells per chamber were seeded onto the scaffolds and cultured in media (MEMα supplement with 5% fetal bovine serum and 1% penicillin/streptomycin) [51]. The culture media was changed every three days.

Preparation of the Scaffolds
The ABS scaffolds were washed with deionized water for 24 h, followed by sterilization in an increasing concentration of ethanol solution (50%, 70%, 80%, 90%, and 100%) for 15 min and air-drying in a cell culture hood. The MED610 scaffolds were also washed in deionized water. However, sterilization was done by autoclaving at 132 °C for 4 min as per the manufacturer's recommendation.
We pre-coated the scaffolds by treating them with 0.01% polylysine solution for 2 h, rinsing them with sterile Dulbecco's phosphate buffer saline (DPBS), and air-drying for 1 h in a sterile cell culture hood. Polylysine, a known contributor to osteoconduction, has been shown to help the cells to attach to the scaffold surface [49,50].

Seeding MC3T3 Cells
After the sterilization and pre-coating (in some cases), the scaffolds were attached to the bottom surface of the cell culture plate using sterilized vacuum grease to prevent them from floating during the cell culture process. MC3T3 cells extracted from the C3 vertebra of a mouse with a density of 1 × 10 3 cells per chamber were seeded onto the scaffolds and cultured in media (MEM supplement with 5% fetal bovine serum and 1% penicillin/streptomycin) [51]. The culture media was changed every three days.

Bone Matrix Deposition on the 3D-Printed Scaffold
After the cells reached 80% confluency, we added 4mM β-glycerol phosphate (G6P) and 0.05 μg/μL ascorbic acid (AA) to induce osteogenesis in the cells [52,53]. The cells on the scaffolds were further cultured for 21 days for the bone matrix to proliferate on the surface. The culture media, along with osteogenic agents, was changed every three days. After 21 days, the hybrid scaffolds were incubated in 4% formaldehyde solution (Electron Microscopy Sciences, Hatfield, PA, USA) for 30 min at room temperature to fix the bone matrix on the scaffold. After that, the scaffolds were washed and stored in DPBS solution.

Bone Matrix Deposition on the 3D-Printed Scaffold
After the cells reached 80% confluency, we added 4 mM β-glycerol phosphate (G6P) and 0.05 µg/µL ascorbic acid (AA) to induce osteogenesis in the cells [52,53]. The cells on the scaffolds were further cultured for 21 days for the bone matrix to proliferate on the surface. The culture media, along with osteogenic agents, was changed every three days. After 21 days, the hybrid scaffolds were incubated in 4% formaldehyde solution (Electron Microscopy Sciences, Hatfield, PA, USA) for 30 min at room temperature to fix the bone matrix on the scaffold. After that, the scaffolds were washed and stored in DPBS solution.

Osteoconductivity of MED610 Scaffolds
The MED610 material is a USP Class VI standard biocompatible material and is certified not to cause an inflammatory response with implantation [54]. To investigate its feasibility as a suitable material for developing our BGS, we evaluated the osteoconductivity, i.e., cell growth and bone matrix deposition on the surface of MED610. We seeded the planar scaffolds with MC3T3 stem cells and induced osteogenesis, as explained in Sections 2.3 and 2.4. We evaluated the cell growth by carrying out scanning electronic microscopy imaging (10.0 Kx magnification at electron high tension (EHT) voltage of 2.5 kV). The von Kossa staining method was applied to assess the bone matrix deposition on the scaffold surface. Ten regions of interest (ROI) (1 mm × 1 mm) were identified on the planar scaffolds. The images of these ROI for seeded and unseeded scaffolds were taken after the von Kossa staining. These images were processed in Matlab ® and the pixel values were normalized with '0' being the brightest measure (100% white) and '1' being the darkest measure (100% black). We calculated the average of these normalized pixel intensities for each ROI as a measure to quantify dark spots on the scaffold (representing calcium salts from the bone matrix deposition). A von Kossa image with bone matrix deposition on the scaffold surface will have darker pixels in the ROI, and average dark pixel intensities would be higher in comparison with the scaffold surface without bone matrix deposition. Comparative t-tests at a significance level of 0.05, were carried out to determine the difference in average dark pixel intensity of ROIs of unseeded and seeded scaffold surfaces.

Mechanical Properties of Scaffolds
As the cylindrical scaffolds we designed were inspired by the cortical bone and trabecular bone (R1 and R2) in a segmental defect model (Figure 1), ideally, they should follow similar mechanical properties as demonstrated by a healthy bone. Mechanical compressive testing (MTEST Quattro, Admet, Norwood, MA, USA) was performed on ABS and MED610 scaffolds to study the anisotropy of the scaffolds. Ten (n = 10) compressive tests in each direction were carried out in the three orthogonal axes with the Z-axis as the longitudinal axis, as illustrated in Figure 2. The scaffolds were placed in between the two stainless steel plates, and unconstrained compression was carried out, as shown in Figure 2b. The speed of measurement was maintained at 5 mm/min based on ISO 604 convention for measurement of plastics [55]. Mechanical characteristics including stiffness (k), maximum compressive strength (σ M ), and compressive modulus (E C ) were calculated in each axis and compared to get a measure of anisotropy [56]. From previous studies, the relationship between the maximum compressive stress about the three axes in a femur bone is expressed as Equation (1)  generate more bone matrix deposition. Many studies suggested that pre-coating the scaffolds with polylysine helps in promoting the growth of bone matrix as it helps the cells to adhere to the surface [49]. CRFP, which is known to enhance the bone matrix production and contribute to osteoinduction, was actively used for inducing osteogenesis along with other osteogenic agents (ascorbic acid and glycerol phosphate). For the ABS and MED610 scaffolds, the different combinations for pre-coating and osteoinduction are listed in Tables 1 and 2, respectively. After inducing osteogenesis for 21 days and fixing the scaffolds with 4% formaldehyde, the scaffolds were evaluated for bone matrix deposition by compressive testing to evaluate the maximum compression strength ( ), stiffness (K), and the compressive modules ( ). Thereafter, we carried out the Shapiro-Wilk test for the compressive test results of each type of scaffold to assess the normality of the data before carrying out the statistical analysis.
For the ABS material, we carried out a one-way analysis of variance (ANOVA) at a significance level of 0.05 to compare the changes in mechanical properties of scaffolds pretreated with polylysine to the original scaffolds (Types A1, A2, and A3). In addition, we carried out a Student's t-test to compare the original scaffold to the scaffolds with no precoating (Types A1 and A4).
For MED610 material, we carried out two one-way ANOVA tests at a significance level of 0.05. The first one-way ANOVA test involved scaffolds pre-treated with polylysine being compared to the original strength of the scaffold (Types M1, M2, and M3). In the second test, we compared the scaffolds with no pre-coating (Types M1, M4, and M5). We followed up the one-way ANOVA tests with a pairwise comparison of the groups by carrying out Student's t-tests at a significance level of 0.017 (Bonferroni correction).

In Vitro Study and Strength Analysis of Cylindrical Scaffolds Treated with CRFP
As explained earlier, ideal BGS need to have similar strength characteristics to that of the natural bone, be osteoconductive (allow for bone growth), and be biocompatible (accepted by the host). To achieve this, we seeded MC3TC cells onto the cylindrical scaffolds and induced osteogenesis to generate bone matrix. This layer of bone matrix has been shown to improve the osteoconductivity of the BGS after implantation and helps in bone penetration into the implant [58]. It has also been shown that, in addition to helping the scaffold better integrate with the host bone, the bone matrix also enhances the strength of the scaffold [42]. We carried out studies with ABS and MED610 cylindrical scaffolds (n = 10), with varying protocol combinations of scaffold preparation and osteoinduction to generate more bone matrix deposition. Many studies suggested that pre-coating the scaffolds with polylysine helps in promoting the growth of bone matrix as it helps the cells to adhere to the surface [49]. CRFP, which is known to enhance the bone matrix production and contribute to osteoinduction, was actively used for inducing osteogenesis along with other osteogenic agents (ascorbic acid and glycerol phosphate). For the ABS and MED610 scaffolds, the different combinations for pre-coating and osteoinduction are listed in Tables 1 and 2, respectively. After inducing osteogenesis for 21 days and fixing the scaffolds with 4% formaldehyde, the scaffolds were evaluated for bone matrix deposition by compressive testing to evaluate the maximum compression strength (σ M ), stiffness (K), and the compressive modules (E c ). Thereafter, we carried out the Shapiro-Wilk test for the compressive test results of each type of scaffold to assess the normality of the data before carrying out the statistical analysis.  For the ABS material, we carried out a one-way analysis of variance (ANOVA) at a significance level of 0.05 to compare the changes in mechanical properties of scaffolds pre-treated with polylysine to the original scaffolds (Types A1, A2, and A3). In addition, we carried out a Student's t-test to compare the original scaffold to the scaffolds with no pre-coating (Types A1 and A4).
For MED610 material, we carried out two one-way ANOVA tests at a significance level of 0.05. The first one-way ANOVA test involved scaffolds pre-treated with polylysine being compared to the original strength of the scaffold (Types M1, M2, and M3). In the second test, we compared the scaffolds with no pre-coating (Types M1, M4, and M5). We followed up the one-way ANOVA tests with a pairwise comparison of the groups by carrying out Student's t-tests at a significance level of 0.017 (Bonferroni correction).
Even though the 3D structure and porosity of the ABS and MED 610 scaffolds are similar, the intrinsic mechanical properties of these materials differ, which leads to varying results in compression testing of both materials. To compare the performance of ABS with MED 610 scaffolds, we evaluated the structural contribution of bone matrix in enhancing the strength of the scaffolds in both materials by subtracting the strength of seeded and differentiated scaffolds with the mean strength of original unseeded scaffolds. We compared similar types in ABS and MED 610 by carrying out a Student's t-test.
Post-compression, ABS scaffolds were stained with 2% Alizarin red (Lifeline Cell Technologies) for validating calcium deposits (a mark of bone deposition) and imaged (Zeiss LSM-510). Imaging studies were carried out as explained in Section 2.5, and the residual red spots formation (post-staining) representing calcium deposits were evaluated by calculating the average red color intensity values. Student's t-tests at significance level of 0.05 were carried out to compare the calcium deposition in different types of ABS scaffolds.
Similarly, in MED610 scaffolds, the von Kossa staining method was applied to validate the bone matrix deposition on the scaffold surface and an imaging study was carried out as explained in Section 2.5.

Bone Matrix Deposition on MED610 Planar Scaffolds
From the scanning electron microscopy studies on planar scaffolds, we see that there is indeed bone growth on the surface of MED610, as the cells attached to the surface are shown in Figure 3a. The von Kossa staining study to evaluate the deposition of bone matrix shows a significant increase in the average dark pixel intensity (p = 0.001), demonstrating significant mineralization on the surface of MED610 scaffolds seeded with bone cells in comparison to unseeded scaffolds, showing that it is an osteoconductive material and allows for bone matrix deposition (Figure 3).

Mechanical Properties of the Cylindrical Scaffolds
The compression studies of ABS and MED610 scaffolds are expressed in the stress vs. strain plots in Figure 4a,b, respectively. These results follow a similar trend that we observe in the literature for healthy bone, where the scaffolds exhibit higher strength in the direction of the longitudinal axis, and the other two orthogonal axes performing at considerably lower strength.

Mechanical Properties of the Cylindrical Scaffolds
The compression studies of ABS and MED610 scaffolds are expressed in the stress vs. strain plots in Figure 4a,b, respectively. These results follow a similar trend that we observe in the literature for healthy bone, where the scaffolds exhibit higher strength in the direction of the longitudinal axis, and the other two orthogonal axes performing at considerably lower strength.
In the in vitro study of the ABS scaffolds, we find that the data collected on the mechanical properties follows normal distribution after undergoing the Shapiro-Wilk test. A one-way ANOVA test showed no significant difference in maximum compressive strength (σ M ), stiffness, and elasticity modulus when the scaffold groups with polylysine pre-coating were compared to the original scaffold (Types A1, A2, and A3). The Student's t-test (between A1 and A4) showed us that the σ M significantly increases by 9.8% (p = 0.009) when CRFP is used to induce osteogenesis (σ A4 = 7.71 MPa) along with G6P and AA, when compared to non-treated scaffolds (σ A1 = 7.02 MPa) (Figure 5a). There was also no significant difference in the stiffness and elasticity modulus of the A1 and A4 scaffolds. Additionally, we were able to visualize and validate the calcium deposits on the surface of the implant using Alizarin red staining. We found that there is a significant increase in the calcium deposits distributed on the surface of the cylindrical implant in the case where CRFP was used with and without polylysine pre-coat (p = 0.012 and p = 0.004, respectively), (illustrated in Figure 5b).
ROI image for unseeded planar scaffolds. (d) Results showing the increase of average dark pixel intensities on the seeded scaffolds in comparison to unseeded scaffolds after von Kossa staining (n = 10). Error bars represent the standard deviation.

Mechanical Properties of the Cylindrical Scaffolds
The compression studies of ABS and MED610 scaffolds are expressed in the stress vs. strain plots in Figure 4a,b, respectively. These results follow a similar trend that we observe in the literature for healthy bone, where the scaffolds exhibit higher strength in the direction of the longitudinal axis, and the other two orthogonal axes performing at considerably lower strength. In the in vitro study of the ABS scaffolds, we find that the data collected on the mechanical properties follows normal distribution after undergoing the Shapiro-Wilk test. A one-way ANOVA test showed no significant difference in maximum compressive strength ( ), stiffness, and elasticity modulus when the scaffold groups with polylysine pre-coating were compared to the original scaffold (Types A1, A2, and A3). The Student's t-test (between A1 and A4) showed us that the significantly increases by 9.8% (p = 0.009) when CRFP is used to induce osteogenesis ( = 7.71 MPa) along with G6P and AA, when compared to non-treated scaffolds ( = 7.02 MPa) (Figure 5a). There was also no significant difference in the stiffness and elasticity modulus of the A1 and A4 scaffolds. Additionally, we were able to visualize and validate the calcium deposits on the surface of the implant using Alizarin red staining. We found that there is a significant increase in the calcium deposits distributed on the surface of the cylindrical implant in the case where CRFP was used with and without polylysine pre-coat (p = 0.012 and p = 0.004, respectively), (illustrated in Figure 5b). In the case of MED 610 scaffolds, we found that the data collected on the compression testing follows normal distribution from the Shapiro-Wilk test. The first one-way ANOVA study showed significant differences (p = 0.0015) in the for the three types of scaffolds (M1, M2, and M3). In the pairwise comparison by the Student's t-test, we found that the strength increased significantly (p = 0.002) by 11.9% for scaffolds treated with CRFP ( = 6.85 MPa) and coated with polylysine when compared to the plain scaffold ( = 6.07 MPa). There was no significant improvement in the compressive strength when CRFP was not used. The second ANOVA study also showed a significant difference in for M1, M4 and M5 scaffolds. In the pairwise comparison, we found that there was a significant 19.6% increase (p < 0.001) in scaffolds with CRFP treatment ( = 7.26 MPa) when compared to plain scaffolds ( ). There was no significant improvement in the compressive strength when CRFP was not used (p = 0.05). There was no significant increase in the stiffness and elasticity modulus even though increasing trends were observed. This is illustrated in Figure 6a. Post-compression von Kossa staining on the cylindrical scaffolds to detect calcium salts showed a significant increase in bone matrix deposition when CRFP is used with and without polylysine, (p = 0.0011 and p < 0.001, respectively). The results of microscopy are illustrated in Figure 6b. The results of stiffness and In the case of MED 610 scaffolds, we found that the data collected on the compression testing follows normal distribution from the Shapiro-Wilk test. The first one-way ANOVA study showed significant differences (p = 0.0015) in the σ M for the three types of scaffolds (M1, M2, and M3). In the pairwise comparison by the Student's t-test, we found that the strength increased significantly (p = 0.002) by 11.9% for scaffolds treated with CRFP (σ M3 = 6.85 MPa) and coated with polylysine when compared to the plain scaffold (σ M1 = 6.07 MPa). There was no significant improvement in the compressive strength when CRFP was not used. The second ANOVA study also showed a significant difference in σ M for M1, M4 and M5 scaffolds. In the pairwise comparison, we found that there was a significant 19.6% increase (p < 0.001) in scaffolds with CRFP treatment (σ M5 = 7.26 MPa) when compared to plain scaffolds (σ M1 ). There was no significant improvement in the compressive strength when CRFP was not used (p = 0.05). There was no significant increase Bioengineering 2021, 8, 199 9 of 14 in the stiffness and elasticity modulus even though increasing trends were observed. This is illustrated in Figure 6a. Post-compression von Kossa staining on the cylindrical scaffolds to detect calcium salts showed a significant increase in bone matrix deposition when CRFP is used with and without polylysine, (p = 0.0011 and p < 0.001, respectively). The results of microscopy are illustrated in Figure 6b. The results of stiffness and elasticity modulus for both materials are plotted in Figure 7.   Table 1; (c,d) plots of Young's modulus (blue) and stiffness (orange) for MED610 scaffolds (n = 10) under different combinations of development as explained in Table 2. Error bars represent the standard deviation.
On comparing the performance of ABS and MED 610 materials, we find that ABS scaffolds pre-coated with polylisine and not treated with CRFP show a significantly (p = 0.0085) higher contribution of strength from the bone matrix deposition ( = 0.58 MPa)   Table 1; (c,d) plots of Young's modulus (blue) and stiffness (orange) for MED610 scaffolds (n = 10) under different combinations of development as explained in Table 2. Error bars represent the standard deviation.
On comparing the performance of ABS and MED 610 materials, we find that ABS scaffolds pre-coated with polylisine and not treated with CRFP show a significantly (p = 0.0085) higher contribution of strength from the bone matrix deposition ( = 0.58 MPa)  Table 1; (c,d) plots of Young's modulus (blue) and stiffness (orange) for MED610 scaffolds (n = 10) under different combinations of development as explained in Table 2. Error bars represent the standard deviation.
On comparing the performance of ABS and MED 610 materials, we find that ABS scaffolds pre-coated with polylisine and not treated with CRFP show a significantly (p = 0.0085) higher contribution of strength from the bone matrix deposition (σ A2−A1 = 0.58 MPa) in comparison to MED 610 scaffolds (σ M2−M1 = 0.21 MPa). There is no significant change in the strength contributed by the bone matrix when ABS and MED610 were compared for the scaffolds treated with CRFP in both cases of polylysine pre-coat and no pre-coat. The results are illustrated in Figure 8. There is no significant change in the strength contributed by the bone matrix when ABS and MED610 were compared for the scaffolds treated with CRFP in both cases of polylysine pre-coat and no pre-coat. The results are illustrated in Figure 8.

Discussion
In this study, we worked on developing a synthetic BGS which is designed based on the cortical structure and internal trabecular structure of the bone. We used two non-biodegradable materials, vis. ABS and MED610 to fabricate 3D cylindrical implants to study their biocompatibility and osteoconductivity. These materials are fabricated with high resolution 3D printing processes to design complex structures to match the structure and mechanical properties of the bone surrounding the implantation site [42,47]. This reduces stress shielding in the implantation site [59]. Previous studies suggest that highly porous structures (>90% porosity) with a large surface area are conductive to cell proliferation on the scaffold surface [60]. However, higher porosity leads to reduction of mechanical strength of the scaffold and reduces the accuracy of consistent structural fabrication. Gregor et al. showed that similar results for cell proliferation can be obtained with BGS of lower porosities (<50% porous) [41]. Our 3D-printed scaffolds have 50% porosity to account for higher structural integrity. Moreover, the non-biodegradability of these materials helps in resolving the problem of reduced mechanical strength experienced over time by other synthetic (PLA/PCL) and ceramic materials such as tri-calcium phosphate (TCP) due to resorption [22,36].
In our in vitro studies, we established that MED610 material is osteoconductive and promotes mineralization through the evaluation of scanning electron microscopy imaging and von Kossa staining imaging studies. Other groups have shown that ABS material is biocompatible and allows for bone growth on the material surface [43]. Osteoconductivity of BGS is essential to facilitate complete integration with the surrounding bone. Moreover,

Discussion
In this study, we worked on developing a synthetic BGS which is designed based on the cortical structure and internal trabecular structure of the bone. We used two nonbiodegradable materials, vis. ABS and MED610 to fabricate 3D cylindrical implants to study their biocompatibility and osteoconductivity. These materials are fabricated with high resolution 3D printing processes to design complex structures to match the structure and mechanical properties of the bone surrounding the implantation site [42,47]. This reduces stress shielding in the implantation site [59]. Previous studies suggest that highly porous structures (>90% porosity) with a large surface area are conductive to cell proliferation on the scaffold surface [60]. However, higher porosity leads to reduction of mechanical strength of the scaffold and reduces the accuracy of consistent structural fabrication. Gregor et al. showed that similar results for cell proliferation can be obtained with BGS of lower porosities (<50% porous) [41]. Our 3D-printed scaffolds have 50% porosity to account for higher structural integrity. Moreover, the non-biodegradability of these materials helps in resolving the problem of reduced mechanical strength experienced over time by other synthetic (PLA/PCL) and ceramic materials such as tri-calcium phosphate (TCP) due to resorption [22,36].
In our in vitro studies, we established that MED610 material is osteoconductive and promotes mineralization through the evaluation of scanning electron microscopy imaging and von Kossa staining imaging studies. Other groups have shown that ABS material is biocompatible and allows for bone growth on the material surface [43]. Osteoconductivity of BGS is essential to facilitate complete integration with the surrounding bone. Moreover, these synthetic BGS do not cause immunological reactions as compared to natural BGS which are derived from other hosts [34,61]. Using MED610 will be essential in our future studies owing to its high-resolution (<16 µm) fabrication process which grants flexibility in designing more complex trabecular features, mimicking shapes of the healthy bone [62]. The compression tests show that our 3D cylindrical scaffolds developed using ABS and MED610 materials exhibit similar trends of anisotropy as that of a femur. We found that the scaffolds show the highest strength along the longitudinal axis of the cylinder (Z-axis) which mimics the longitudinal axis of the bone; whereas, the strength in the sagittal and frontal axes (X and Y directions) were considerably lower, as shown in Figure 4a,b. This behavior is similar to that of anatomical stresses of a femur bone [56,57].
σ Longitudinal axis > σ Sagittal axis ≥ σ Frontal axis In our initial studies on cylindrical ABS scaffolds, we observed bone growth, osteogenesis, and osteointegration by evaluating the effect of the increase in the bone matrix on the mechanical strength of the implants. In some cases, we coated the scaffolds with polylysine before cell seeding, as polylysine is known to promote cell adherence to plastic surfaces [49]. In some cases, we added CRFP along with the osteogenic reagents as it is known to contribute to bone formation [42,44]. The analysis showed that the CRFP-treated scaffold with no pre-coating (Type A4) showed significant compressive strength increases (p < 0.05). We attribute this increase in strength to the formation of bone matrix deposition (shown using Alizarin red staining). The increases in the strength of other combinations of ABS scaffolds (Type A2 and Type A3) was not found to be significant.
In the studies with MED 610, we observed that the scaffolds showed a significant increase in compressive strength compared to the (Type M1) dry scaffolds (9.6% for Type M3 and 19.8% for Type M5) when they were treated with CRFP along with the osteogenic reagents. This shows that the treatment with CRFP itself plays a significant role in generating the bone matrix and improving the mineralization and biomechanical properties of the scaffolds. The stiffness and elasticity modulus for both ABS and MED610 scaffolds revealed that there was no significant increase in these properties in all of the scaffold combinations. This shows that the scaffolds treated with CRFP, which are not pre-coated, give quantitively better results.
When we compared the performance of ABS with MED610, we observed that in the case where the scaffold was pre-coated with polylysine, contribution of strength from the bone matrix decreased. However, we see that in the case where CRFP was used along with the osteoinduction reagents (G6P and AA) in both pre-coated and non-coated scaffolds, there was no significant difference between the performances of the two materials; This shows us that the bone matrix deposition on both ABS and MED610 scaffolds is nearly similar when CRFP is used during osteoinduction.
One of the limitations in the design of our scaffolds is that the cylindrical scaffold was designed by mimicking the cross-section of a femur and a cylindrical design pattern was selected for ease in our in vitro studies. Eventually, when biomimetic scaffolds are to be used for in vivo studies to replace bone stock, the scaffolds will be designed using the shape and contour of the bone surface. In this study, the pillars in the scaffolds were designed to be as thin as possible to increase the porosity of the scaffold. This pattern was selected to have a better field of view in the imaging studies and also to observe how far cells can penetrate within the grooves of the scaffolds. In the future, the thickness of the pillars can be changed to adjust for the axial stiffness of the scaffold so that it uniformly distributes stress when integrated with the healthy bone.

Conclusions
We successfully demonstrated that artificial 3D-printed scaffolds developed by seeding cells perform better biomechanically and produce more bone matrix when CRFP is used in combination with osteogenic reagents. We also found that scaffolds without pre-coating