Biomechanical properties of 3D-printed bone scaffolds are improved by treatment with CRFP

A reverse engineering method was used to design 3D-printed ABS scaffolds (Fig. 1). An L5 vertebrae from a skeletally mature, male Sprague Dawley rat was isolated and scanned by micro-computed tomography (μCT40, Scanco Medical) at an isotropic voxel resolution of 18 um. The output was extracted to a series of DICOM files, each file describing a 2D cross-section representing of the site being imaged. Using InVesalius (Renato Archer Information Technology Center, Brazil), a medical imaging reconstruction software that takes the DICOM files from the CT scan and creates a three-dimensional (3D) surface representation, 3D image was exported into a stereolitography (STL) format, a standard among CAD software and 3D printing. To reduce the noise from the DICOM files, Geomagic (3D-systems, NC) was used as a STL image processing software. Once the noise of the STL file was completely reduced, a Geomagic algorithm was used to export the output as a CAD file. This file was uploaded to Solidworks (Dassault Systèmes, MA) (solid modeling CAD engineering software which is used for feature extraction to replicate the bone’s trabeculae pattern over scaffold’s surface) to export it as STL file which was used as an output to the 3D printer. The scaffolds were 3D printed using fused deposition modeling (FDM) process in a CubePro (3D Systems, USA) 3D printer. In order to successfully extract the trabecular pattern of the bone so that it matched our 3D printer resolution, several reiterations were made so as to resemble the actual bone, and the best structurally biomimetic scaffold was selected for our studies.

MC3T3 cells were used to seed on scaffolds as they are commonly used to study in vitro osteoblast differentiation [8]. The printed scaffolds were washed for 24 h with deionized water and sterilized in ethanol also, for 24 h. After that, scaffolds were air-dried in a cell culture hood. Each scaffold was attached to the bottom of a cell culture plate well using sterilized vacuum grease. The scaffolds were pretreated with CRFP (15 uM) for 24 h and air-dried under sterile conditions. MC3T3 cells at a density of 1 × 10³ cells per chamber were seeded onto the scaffolds and cultured in media (DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin). Media was changed every 3 days till the cells reached 70% confluency.

Osteogenesis was induced and assessed as previously described. Glycerol phosphate (G6P) and ascorbic acid are well known osteogenic agents [9, 10]. It is also of interest to evaluate CRFP in these scaffolds as an osteogenic agent before seeding cells. Osteogenic reagents were added to each well as described earlier [8]. Confluent cells treated with osteogenic inducers were cultured for 21 days. Media were changed accordingly every 5 days.

To evaluate the growth of cells on ABS scaffolds, scanning electron microscopy, fluorescence microscopy, and bright field microscopy were used. Cell-seeded and cultured scaffolds were fixed in 4% formaldehyde (Electron Microscopy Services) for 30 mins at room temperature. Two distinct staining protocols were used to assess cellular growth and osteogenesis. For cellular growth, scaffolds were double stained with Alexa 568-Phalloidin (Sigma-Aldrich) and DAPI (Sigma-Aldrich) to visualize actin and nuclei, respectively. To evaluate bone matrix deposition, 2% Alizarin red (Lifeline Cell Technologies) was used to stain calcium deposits for 20 mins at room temperature to evaluate mineralization process and bone matrix formation. Scaffolds were imaged with confocal microscopy (Zeiss LSM-510) and with a scanning electron microscope (ISI-SX30 SEM).

A mechanical model was used to obtain the stiffness, or modulus, of the specimens (Fig. 2). Briefly, a non-treated scaffold has only the stiffness of its material (in this case ABS), and it is denoted as k₁. When the scaffolds are treated and mineralization is induced, the scaffolds now carry not only the ABS stiffness but also the stiffness of the formed cellular matrix k₂. Mechanical testing cannot distinguish between these two stiffnesses, as it will only allow us to measure the equivalent stiffness of the hybrid, i.e., k_e. A straightforward mechanical model will allow us to calculate k_e as follows:

In this study, the anisotropy of the scaffolds was tested Eq. 1. Three different directions of compression were tested and compared (Fig. 2).

Also, not only the manufacturing process of the scaffolds contributes to their anisotropy, but also trabecular pattern itself adds this characteristic to the specimen as bone structure is indeed anisotropic. Bone structures perform better under compression rather than tension. However, the direction of the compression is also important. In previous statistical studies [11], the following inequality was noted, according to axis described in Fig. 3:

According to Eq. 2, the human bone performs better under compression but only if the axis of that compression is parallel to the longitudinal axis of the bone (Fig. 3). In our case, the trabecular feature of the bone was extracted from the natural anatomical position.

To carry out all these comparisons, the international standard, ISO 604 [12], was followed, which is a standard for determination of compressive properties of plastics. We chose this standard based on the core material of our scaffolds. As stated before, two main comparison tests were performed in the experiments using this standard on scaffolds seeded with cells with or without osteogenic agent treatment:

Two different dimensions of specimens were considered (Table 1). Type A models were used to calculate the compressive modulus of scaffolds, and type B models allowed us to obtain the compression strength. Table 2 and Fig. 2 detail the list of the number of specimens tested according to ISO 604 specifications and the specified directions of compressions.

An MTEST Quattro (Admet, USA) compression testing system was used in concordance with ISO 604 to carry out scaffolds testing. Specimens were tested in unconstrained compression by positioning them in the center of two stainless steel plates. Surface finishing of the specimens was enhanced prior to experiments using sandpaper in order to eliminate sharp edges and loose filaments formed due to 3D printing process. All environmental conditions established by ISO 604 were accomplished, stainless steel plate parallelism was inspected and corrected before any run, and every specimen was placed in the same position for any test to avoid discrepancies in results.

The speed for the tests varies depending on the outcome parameter and the dimension of the specimens. For our scaffolds, the following speeds were used as stated in ISO 604:

To evaluate biomechanical properties of cell-seeded scaffolds, compression strength (σ_M) and compressive modulus (E_C) were measured:

Average values for each set of five tests were taken to construct stress vs. strain curves based on Eq. 3.

Incubating the seeded scaffolds under osteogenic conditions showed osteoblasts promoting matrix formation and mineralization, as evidenced by Alizarin red staining in 3D reconstructed confocal z-sections images of cells grown on scaffolds (Fig. 4a). In addition, SEM imaging shows cells attached to the surfaces of the scaffolds (Fig. 4b).

Values for σ_M and E_C were shown in Table 3. For axis 1 and axis 2, flat scaffolds reported greater compressive strength and modulus; however, for axis 3, trabecular scaffolds exceeded flat specimens in the order of 20% for σ_M and 60% for E_C.

Another way to compare the results is from the point of view of axis of compression, grouping only flat scaffolds or trabecular scaffolds (Fig. 5a). Equation 2 shows that the compressive strength in axis 3 is greater than axis 2, which is also greater than axis 1.

The contribution of cellular bone matrix for compression strength and modulus was also investigated (Fig. 5b). Only trabecular scaffolds were compressed in direction 2. Four samples were tested: (1) unseeded, or untreated, scaffolds, which are the original specimens without any biological treatment or cell seeding;(2) seeded scaffolds, specimens with cells where mineralization was not induced; (3) seeded with osteogenic, specimens with cells induced for mineralization by using osteogenic reagents (G6P and ascorbic acid); and (4) seeded with CRFP, which are specimens with cells induced for mineralization by using CRFP.

Upon mineralization, the trabecular scaffolds showed an increase in the compressive modulus. Compressive strength also increased, but not as profoundly as E_C. Table 4 shows average values for each set of five compression tests recommended by ISO 604. Treating seeded scaffolds with CRFP significantly increased the compressive modulus compared to unseeded controls.