Three-Dimensional Porous Trabecular Scaffold Exhibits Osteoconductive Behaviors In Vitro

Abstract

In the USA, approximately 500,000 bone grafting procedures are performed annually to treat injured or diseased bone. Autografts and allografts are the most common treatment options but can lead to adverse outcomes such as donor site morbidity and mechanical failure within 10 years. Due to this, tissue engineered replacements have emerged as a promising alternative to the biological options. In this study, we characterize an electrospun porous composite scaffold as a potential bone substitute. Various mineralization techniques including electrodeposition were explored to determine the optimal method to integrate mineral content throughout the scaffold. In vitro studies were performed to determine the biocompatibility and osteogenic potential of the nanofibrous scaffolds. The presence of hydroxyapatite (HAp) and brushite throughout the scaffold was confirmed using energy dispersive X-ray fluorescence, scanning electron microscopy, and ash weight analysis. The active flow of ions via electrodeposition mineralization led to a threefold increase in mineral content throughout the scaffold in comparison to static and flow mineralization. Additionally, a ten-layer scaffold was successfully mineralized and confirmed with an alizarin red assay. In vitro studies confirmed the mineralized scaffold was biocompatible with human bone marrow derived stromal cells. Additionally, bone marrow derived stromal cells seeded on the mineralized scaffold with embedded HAp expressed 30% more osteocalcin, a primary bone protein, than these cells seeded on non-mineralized scaffolds and only 9% less osteocalcin than mature pre-osteoblasts on tissue culture polystyrene. This work aims to confirm the potential of a biomimetic mineralized scaffold for full-thickness trabecular bone replacement.

Lay Summary

Bioengineered options for trabecular bone replacement should be porous for cellular infiltration and bioactive to promote the formation of new bone tissue. This work features techniques to increase mineralization within full-thickness porous nanofibrous scaffolds. In vitro studies elucidated the biocompatibility and osteogenic potential of the mineralized trabecular scaffold by promoting human bone marrow–derived stromal cell proliferation and primary bone protein expression. The end goal of this work is to combine this porous trabecular scaffold with a pre-vascularized cortical bone scaffold to yield a full-dimensional biomimetic bone construct with the ability to promote simultaneous multidifferentiation of stem cells in vivo.

Introduction

Structurally, bone is organized into two distinct types based on its density and location, cortical bone and trabecular bone [1–3]. Cortical bone, or compact bone, has a highly dense organized structure composed of tightly packed osteons [4]. Trabecular bone, or spongy bone, is 90% porous with lower mechanical strength than cortical bone and is a meshwork of collagenous tissue filled with bone marrow [3,4]. The main components of bone, collagen type I, hydroxyapatite (HAp) and water, contribute to the toughness, stiffness and viscoelastic properties of bone [1, 2, 5]. Naturally, bone regenerates without scar tissue formation when injured or damaged, but surgical intervention is required when there is significant bone loss or disease [6]. The gold standard orthopaedic procedure to replace damaged bone is autografting with bone tissue from the iliac crest and a metal fixation. This procedure can result in donor site morbidity and longer recovery time due to multiple surgeries. Allografting is an alternative option with only one surgery required, but there is a 30–60% failure rate over 10 years in vivo associated with this procedure [7]. Similar limitations are seen with xenografts from a bovine or porcine source. Alternatives such as metallic implants (used in joint replacement) lack the ability to bond with surrounding bone, whereas bone cements cannot be used exclusively to repair load-bearing bones. Therefore, many researchers are investigating the use of bioactive materials to tissue engineer bone.

The ideal biocompatible scaffold for bone tissue engineering should promote cellular infiltration and have comparable strength to native bone [6–8]. Furthermore, the scaffold should be highly porous to allow for nutrient transport and promote osteogenic differentiation. Therefore, we developed a porous osteogenic scaffold that mimics the structure and composition of native trabecular bone. The scaffold is composed of FDA-approved poly (α-hydroxy acid) polyesters poly-l-lactide (PLLA) and poly-d-lactide (PDLA), synthetic HAp and bovine gelatin. PLLA and PDLA were selected based on their mechanical properties and hydrolytic degradation behavior [9, 10]. HAp serves as the osteogenic inducer and has been shown to promote osteogenesis in several in vitro studies [5, 11]. Gelatin derived from bovine or porcine collagen is a commonly used scaffolding material because it is biocompatible, biodegradable, commercially available, non-immunogenic, and promotes cell proliferation [12–14]. Additionally, mineral is accumulated at the carboxyl sites of gelatin when exposed to an ionic solution, such as simulated body fluid (SBF) [15, 16]. This process, known as mineralization, can be prepared under static, active or passive conditions. Static mineralization is the standard technique, but this time-consuming process has been shown to lead to non-uniform mineral deposition throughout three-dimensional structures [17, 18]. Previous work also concluded flow mineralization can lead to decreased mechanical properties due to scaffold degradation caused by active flow of ions (data not published). Therefore, we explored mineralization via electrodeposition. This rapid method has been used to generate apatite coatings on metallic substrates by promoting directional ion flow using electrodes [19–21]. We hypothesize that the incorporation of inorganic calcium-based minerals via gelatin nucleation and direct integration into a porous scaffold will promote differentiation of human bone marrow–derived stromal cells (hMSCs) along the osteoblastic lineage. The findings from this work will lead to an optimized bioengineered scaffold for trabecular bone replacement.

Materials and Methods

Nanofibrous Scaffold Preparation and Characterization

PLLA (MW = 152 kDa) was purchased from Sigma-Aldrich (St. Louis, MO, USA). PDLA (MW = 124 kDa) was purchased from Evonik Birmingham Laboratories (Birmingham, AL, USA). Solvents dichloromethane (DCM), dimethylformamide (DMF) and tetrahydrofuran (THF) were purchased from Fisher Scientific (Pittsburgh, PA, USA). Gelatin, from porcine skin, was purchased from Sigma-Aldrich (St. Louis, MO, USA). The solutes used in the simulated body fluid (SBF)—sodium chloride (NaCl), potassium chloride (KCl), calcium chloride dihydrate (CaCl·H₂O), magnesium dichloride heptahydrate (MgCl₂·7H₂O), sodium bicarbonate (NaHCO₃) and sodium phosphate monobasic (NaH₂PO₄)—were purchased from Fisher Scientific (Pittsburgh, PA, USA). Electrospinning solutions and scaffolds were prepared as thoroughly mentioned in previous work [22]. In brief, the base electrospinning solutions were prepared by dissolving PLLA (7% w/v) in DCM and DMF with 10% gelatin and PDLA (22 w/v%) in THF and DMF at a 3:1 ratio. Prior to electrospinning, 10% synthetic nanopowder HAp was suspended in a PDLA (22 w/v%) in dimethylformamide (DMF) and tetrahydrofuran (THF) solution (50/50 mixture). Synthetic nanopowder HAp (< 200 nm) was purchased from Sigma-Aldrich (Atlanta, GA, USA).

Scanning electron microscopy (SEM) and energy-dispersive x-ray fluorescence (ED-XRF) was performed on the PDLA/10%HAp scaffolds to qualitatively and quantitatively assess the presence of HAp in the scaffold. Samples (n = 3) were dried overnight and sputter coated with gold and palladium for SEM imaging.

Scaffold Mineralization

PDLA and PLLA with 10% gelatin (PDLA_PLLA/10%Gel) scaffolds without HAp were used for this study. The gelatin-water mixture was added into the PLLA solution prior to electrospinning. The scaffolds were prepared with or without the use of salt as a porogen, and sintered and leached as previously described [10, 22]. In this previous work, the pore diameters were between 5 and 200 μm, which is desirable for cellular infiltration and nutrient transport [22]. Sintering was performed by placing the scaffolds in an oven set at 54 °C for 45 min. The SBF was prepared according to Jalota et al. [23, 24].

In a preliminary study, we tested the effectiveness of the electrodeposition mineralization on electrospun PDLA/PLLA/10%Gel mats. The mats were cut into strips and 4, 6 or 10 strips were sintered together to form scaffolds. These scaffolds were then mineralized via electrodeposition to evaluate the ability of the technique to mineralize a thick substrate. In this first study, scaffolds were mineralized for 1 h between platinum wires at 5 V in SBF maintained at 60 °C. Following this study, we explored the role porosity plays in the mineralization process. Macropores were created by depositing salt during the electrospinning process and leaching out the salt after the mats were formed [10]. Scaffolds with 10 layers and with or without pores were mineralized for 4 or 24 h, stained with alizarin red, embedded in Tissue-Tek® O.C.T. Compound, and sliced using a cryotome.

Following these studies, we varied the electrodeposition parameters (voltage, time and direction). The electrodeposition setup is shown in Fig. 1. A platinum wire was placed through the center of a cylindrical scaffold and a silicon electrode was placed around the scaffold. The scaffold was completely submerged in 40 mL of SBF with a pH of 4.4. The positive lead was attached to the silicon electrode and the negative lead was attached to the platinum wire. The positive and negative leads were switched to change the direction of the gradient and force the minerals towards the outside of the scaffold after 1 hour. The scaffolds were mineralized for 2 hours at 0.3 A and voltages of 5 V, 10 V or 15 V. The degree of mineralization of these samples was compared to scaffolds mineralized under standard (static) conditions.

Fig. 1.

Schematic of the electrodeposition setup. The conductive silicon rubber with silver-plated aluminum filler electrode is cradling the cylindrical scaffold resting in simulated body fluid. A wire platinum electrode is threaded through the center of the scaffold

In this study, scaffold mineral content was quantitatively evaluated using an alizarin red stain on 100 μm thick cross-sectional slices of the scaffolds (pH = 4.23). Cetylpyridinium chloride (CPC) (pH of 7.0) was used to extract the alizarin red stain from the samples and read at 540 nm on a spectrophotometer. The weight percent of inorganic content in the scaffold was assessed by placing the scaffolds in a furnace set to 700 °C for 24 h to remove any organic matter. In a follow-up study, scaffolds were mineralized via electrodeposition for 6 h at 4 V to determine if longer mineralization times would result in greater mineral content. The samples from this group were assessed for percent of inorganic content and compression mechanical testing (1 mm/min).

In Vitro Assessment of Biocompatibility and Osteogenic Potential

Two independent in vitro studies were performed with varying scaffold groups to determine the effect of the scaffold design on cellular biocompatibility and osteogenic potential. Scaffolds were cut into 10 mm diameter circles and adhered to the bottom surface of the well in a 48-well plate for all the in vitro studies. In the first study, the goal was to investigate the effect of HAp and length of static mineralization (in days) on cellular metabolic activity and calcium production. Human bone marrow-derived stromal cells (hBMSCs) were harvested from human fetal long bone and expanded by Samineni et al. [25]. The hBMSCs (passage 8) were maintained and expanded in Alpha-Modified Minimum Essential Media (α-MEM) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin streptomycin for 1 week. The use of hBMSCs at a high passage number was not a concern for this study since differentiation was not assessed before the study. The groups for this study are outlined in Table 1. The negative control group was cells cultured directly on the tissue culture polystyrene and positive control scaffold, NoMin_NoHAp_NoSalt, was non-mineralized and did not have HAp embedded into the polymeric network or pore induced by salt leaching. The experimental groups were scaffolds, previously described, with 10% HAp and salt-leached (‘Salt’) pores and were either mineralized via static mineralization for 3 (Min3days_HAp_Salt) or 7 days (Min7days_HAp_Salt). The hBMSCs were seeded onto the appropriate substrate at 10,000 cells per substrate to yield a seeding density of 10,000 cells per cm². The cells were seeded at a low density to prevent cell contact inhibition at later timepoints. All substrates were sterilized with 70% ethanol and UV radiation and pre-conditioned with media for 24 h. For all groups, the media was switched to osteogenic media, alpha MEM supplemented with 0.1 M ascorbic acid, 0.01 M dexamethasone and 0.01 M Beta-glycerophosphate, after 1 week of culture. PrestoBlue® assay was performed to evaluate cellular metabolic activity (normalized to cell number), and alizarin red stain was used to assess calcium deposition (normalized to metabolic activity). The groups were compared to evaluate the effect of the scaffold’s overall composition on cellular behavior.

Table 1.

Experimental groups

Group	Mineralized	HAp in scaffold	Salt-leached pores
TCP (control)	X	X	X
NoMin_NoHAp_NoSalt	X	X	X
Min3days_HAp_SL	3 days	✓	✓
Min7days_HAp_SL	7 days	✓	✓

In the second independent study, the goal was to evaluate the osteogenic potential of the PDLA/PLLA_10%Gel scaffolds, with and without 10% HAp, over 10 days using a different cell line. Human marrow stromal cells (hMSCs) (passage 1) donated from a male aged 22 years old and obtained from Texas A&M Health Science Center and used for this study. The cells were maintained in α-MEM supplemented with 10% fetal bovine serum (FBS), 1% penicillin and 1% 2 mM L-glutamine. In order to better measure the degree of osteoblastic activity of the hMSCs, MC3T3-E1 pre-osteoblasts were also cultured simultaneously in a separate well plate and examined for osteocalcin secretion as a comparison. The cells were seeded at 50,000 cells/substrate onto TCP (control), scaffolds without HAp (PDLA/PLLA_10%Gel), and scaffolds with HAp (PDLA/PLA_10%Gel/10%HAp). Scaffolds from both experimental groups were mineralized via static mineralization for 6 h (different methods of mineralization were evaluated in a separate part of this study). Cellular viability was assessed using PrestoBlue® assay and normalized to cell number. The supernatant from each group was collected at 3, 7 and 10 days and evaluated for osteogenic protein secretion using a quantitative ELISA for intact human osteocalcin (OC) (novex by Life Technologies). Alkaline phosphatase (ALP), an early marker of osteogenesis, was also measured with an Alkaline Phosphatase Staining Kit II (Stemgent, Inc., USA). Images were taken at 3, 7 and 10 days with colored light microscopy for qualitative analysis and quantified using Image J, Fig. 2.

Fig. 2.

a ALP stained images (× 10) were processed in ImageJ to calculate area and percent area of ALP expression. The images were converted to grayscale (b) and then segmented to isolate the red-stained areas (c). Area size and percent area was computed and compared between groups. Scale bar = 200 μm

Statistical Analysis

Statistical analysis was performed using KaleidaGraph Synergy Software. All data was subjected to a one-way analysis of variance (ANOVA) with post hoc analysis (Tukey’s test) to determine the statistical significance of differences between groups, p value < 0.05. Two-way ANOVA with post hoc analysis (Tukey’s test) was performed to compared each group over time, p value < 0.05.

Results

Characterization of Scaffold with HAp (SEM & ED-XRF)

Qualitative and quantitative analysis of the PDLA/10% HAp and PDLA scaffolds was performed using SEM and ED-XRF to visualize the HAp in the nanofibrous network. Figure 3 (left) shows the HAp aggregates through the nanofibrous structure. Furthermore, the two primary elements in HAp, calcium and phosphorus, were detected in the PDLA/10% HAp scaffold at statistically higher level than the control PDLA scaffold.

Fig. 3.

(Left) SEM image of the PDLA/10%HAp electrospun scaffold (yellow arrows indicate HAp aggregates. (Scale bar = 10 μm). (Right) ED-XRF measurements of PDLA/10%HAp and PDLA scaffold. Statistical analysis: # denotes ANOVA Tukey’s test (post hoc) p < 0.05. Data presented as mean ± standard deviation

Mineralization of Layered Scaffolds

Scaffolds of varying layers were mineralized via electrodeposition and stained with alizarin red to evaluate the depth of mineralization at varying times. The cross-sectional images of the stained scaffolds are shown in Fig. 4. Figure 4a–c are scaffolds mineralized for 1 h at 5 V. Qualitatively, there appears to be a larger amount of alizarin red stain present throughout the 4-layer scaffold, Fig. 4a. In a follow-up study, the mineralization time and voltage was increased (up to 24 h and 15 V respectively) in an attempt to mineralize the full thickness of a 10-layered scaffold with and without salt. Although not significant, there appears to be more mineral deposited at the surface and throughout the scaffold with salt pores, Fig. 4f and g. The percent mineral content in the scaffold mineralized at varying voltages is shown in Fig. 4h. The average for the samples mineralized via static mineralization is 2.91 ± 1.54%. The average percent mineralization for scaffolds mineralized by electrodeposition at 5 V, 10 V and 15 V are 8.51 ± 4.20%, 4.37 ± 2.15% and 7.63 ± 1.60%, respectively.

Fig. 4.

a–c Cross-sectional images of alizarin red stained scaffolds with 4, 6 and 10 layers mineralized for 1 h at 5 V in 60 °C SBF. d–g) Cross-sectional images of 10 layered alizarin red stained scaffolds with and without salt pores mineralized for 4 and 24 h at 5 V in 60 °C SBF. The solid white material surrounding the samples is the embedding agent. h The ash weight results of scaffold mineralized via electrodeposition mineralization. No statistical significance observed. Data presented as mean ± standard deviation

In Vitro Viability Analysis of hBMSCs on Trabecular Scaffold

Metabolic activity was measured using PrestoBlue® assay as an indirect measurement of cellular viability over 28 days, Fig. 5a. The data was normalized to cell number. There was a significant increase in metabolic activity at day 21 for the hBMSCs seeded on TCP, NoMin_NoHAp_NoSalt, scaffolds and the salt-leached scaffolds mineralized for 7 days, Min7Day_HAp_Salt in comparison to day 7 and day 28 within the same groups.

Fig. 5.

a Normalized metabolic activity analysis of hBMSCs on TCP, non-mineralized scaffold without HAp (NoMin_NoHAp_NoSalt) and scaffolds with HAp mineralized for 3 (Min3Days_HAp_Salt) or 7 days (Min7Days_HAp_Salt) evaluated over 28 days (y-axis is in relative fluorescence units). * = p < 0.05 compared to day 7 and day 28 within the same group. Data presented as mean ± standard deviation * b Normalized cellular calcium deposition of hBMSCs evaluated using alizarin red stain over 28 days (same groups as above). TCP data not shown. *p < 0.05 compared to positive control group, NoMin_NoHAp, NoSalt. Data presented as mean ± standard deviation

The normalized absorbance readings of the alizarin red stain are shown in Fig. 5b. The absorbance reading of the associated substrate without cells was subtracted from the absorbance reading of the substrate with the cells; therefore the reading is directly proportional to the amount of calcium deposited by the hBMSCs in response to the scaffold’s composition. The assay measured a negligible amount of calcium deposited by the cells seeded on TCP over 28 days (data not shown), at day 7 for all the scaffold groups, and at day 28 for the NoMin_NoHAp_NoSalt and Min3Day_HAp_Salt groups, Fig. 5b. The amount of calcium deposited by the cells seeded on the scaffolds mineralized for 3 days, Min3Days_HAp_Salt was significantly greater than the amount deposited by cells in the other groups at day 14. At day 21, the calcium deposited by the cells on the mineralized scaffolds was significantly greater than the amount deposited on the non-mineralized control scaffold.

In Vitro Analysis of Osteogenic Potential of HAp Trabecular Scaffold

The metabolic activity of the hMSCs (obtained from Texas A&M Health Science Center) seeded on TCP and PDLA scaffolds with HAp (PDLA-HAp _hMSCs) and without HAp (PDLA_hMSCs) was measured to assess biocompatibility. There were no statistically significant differences between groups over 10 days, as shown in Fig. 6a. This suggests the scaffold is biocompatible and the cells attached to the scaffolds similarly to the TCP. The decrease in cellular viability on day 10 seen with the hMSCs seeded on the TCP (control) could be due to high confluency, whereas the hMSCS seeded on the 3D scaffold have more surface area for proliferation and migration.

Fig. 6.

a Metabolic activity indicates no statistically significant differences in cellular activity between groups up to 10 days. b Osteocalcin secretion in serum measured at days 3, 7 and 10 (N = 6). Statistical analysis via ANOVA Tukey’s test (post hoc): # p < 0.05 compared samples from other groups at the same timepoint, *p < 0.05 compared to hMSCs on TCP within the same time point, + p < 0.05 compared to sample from the same group at a different time point. Data presented as mean ± standard deviation

Osteocalcin protein secretion was measured quantitatively using an ELISA kit. Media was also collected from a negative control scaffold (scaffold without cells) to verify the osteocalcin being measured was secreted from the cells and not leached HAp. There was an increase in osteocalcin secretion from the hMSCs that were seeded on the control scaffold without Hap and the scaffold with HAp in comparison to the hMSCs on TCP, Fig. 6b. At day 10, there were no significant differences in the OC secretion levels of the hMSCs seeded on the scaffold with HAp and the pre-osteoblast cells seeded on TCP.

To further determine the level of osteoblastic differentiation by the hMSC cells from each group were stained for also stained for alkaline phosphatase (ALP) ALP activity over 10 days. The intensity of the red color is directly proportional to ALP activity. There is an increase in the ALP activity from the cells seeded on the PDLA_HAp scaffold on day 3, Fig. 7a.

Fig. 7.

a Representative images of ALP stain (red) of hMSCS seeded on TCP (hMSCs_TCP), PDLA scaffold without HAp (hMSCS_PDLA) and PDLA with HAp (hMSCs_PDLA_HAp) (scale bar = 200 μm). b Normalized ALP area (um² per substrate). There was a statistically significant increase in ALP from the hMSCs seeded on the scaffold groups, hMSCs_PDLA and hMSCs_PDLA_HAp at day 7 and day 10 in comparison to the control. Statistical analysis: # denotes ANOVA Tukey’s test (post hoc) p < 0.05 from the control (hMSCs_TCP). Data presented as mean ± standard deviation

The increase in ALP was compared quantitatively and depicted in Fig. 7b. There was no significant difference in the ALP activity between the two scaffold groups, but a statistically significant difference in ALP activity between the hMSCs seeded on the scaffolds and seeded on TCP. The red hue of the stain is decreased at Day 10.

Discussion

The overall objective of the studies mentioned was to characterize a gelatin-synthetic porous osteoinductive scaffold for bone regeneration. Qualitative and quantitative analysis exhibited the ability to successfully incorporate HAp aggregates into the synthetic polymeric scaffold. HAp, Ca₅(PO₄)₃(OH), is a naturally occurring mineral form of calcium apatite. Calcium and phosphorus, two elements in HAp, expressed a statistically significant higher intensity in the PDLA/10%HAp scaffold in comparison to the control PDLA electrospun scaffold as measured using ED-XRF. The intensity reading of the phosphorus and calcium in the control scaffold can be contributed to artifacts or detection of the PDLA compounds. Electrospinning the HAp directly in the scaffold was more desirable than previous methods explored in our laboratory because this method led to higher retention of the HAp particles post leaching. Further investigation will need to be performed for homogeneous distribution of the HAp through the nanofibrous network and around the individual fibers.

Scaffolds of varying thicknesses were mineralized via electrodeposition to evaluate the novel mineralization technique for three-dimensional structures. The thickest scaffold with 10 layers had the highest mineral content when mineralized for 4 h with the presence of salt pores. It is believed that the salt pores promoted the influx of the ions from the SBF solution during mineralization. Nonetheless, the electrodeposition mineralization at 5 V did not successfully mineralize the full thickness of the scaffold. Therefore, scaffolds were then mineralized via electrodeposition mineralization for 2 h at 5 V, 10 V and 15 V and compared to scaffolds mineralized under static conditions. Although not statistically significant, the scaffolds mineralized at 5 V had higher mineral content than samples mineralized via static and electrodeposition at 10 V and 15 V. This could be due to the silicone rubber losing conductivity at higher voltages. The conductivity of the material decreases after repeated use or exposure to higher voltages. A decrease in conductivity would decrease the amount of current flowing through the electrode at the same voltage. In this case, the decrease in current created a weaker gradient of minerals forced through the scaffold. The first electrodeposition samples were mineralized at 5 V, but the silicone electrode was not as conductive when samples were mineralized at 10 V or 15 V. Based on the results, electrodeposition mineralization at 5 V for 2 h will be utilized for future studies as the technique to actively integrate mineral ions throughout the scaffold.

The biocompatibility and bioactivity of scaffolds mineralized for 3 and 7 days was assessed and compared to non-mineralized scaffolds. The increased metabolic activity of the cells seeded on TCP in comparison to the cells seeded on the scaffolds was expected due to lower initial seeding efficiency on three-dimensional nanofibrous scaffolds. Decreased metabolic activity was seen in all the groups except Min3Days_HAP_Salt group and could be due to an over-confluent cell population. The multipotency and proliferation of MSCs can be altered due to high cell density [26–28]. There was a similar trend of decrease in calcium deposition for all groups on day 28 as seen in the metabolic activity results indicating a direct relationship between metabolic activity and calcium deposition. The significant increase in alizarin red stain from day 7 to day 14 seen with the scaffolds mineralized for 3 days, Min3Days_HAp_Salt, is possibly due to the change to osteogenic media at day 7. Overall, the metabolic activity of the cells was not adversely affected by the composition and surface morphology of the mineralized porous scaffolds. The mineral deposited on the bioactive scaffold via static mineralization appeared to promote the hBMSCs to deposit more calcium than the cells seeded on TCP and the non-mineralized scaffold. The osteogenic potential of the scaffold with HAp was evaluated quantitatively and qualitatively by detecting the secretion and presence of bone-specific enzyme alkaline phosphatase, bone matrix component calcium and bone protein osteocalcin (OC). Additional, serum osteocalcin (OC), a non-collageneous protein marker for osteoblast activity, was measured using an ELISA [29]. The unexpected increase in the OC activity for the hMSCs seeded on the control scaffold may be due to mechanotransduction in response to the scaffold’s stiffness and the mineral content deposited during the mineralization process. Furthermore, the addition of HAp in the scaffold led to a statistically significant increase in OC secretion from the hMSCs at day 10 in comparison to the cells seeded on the control scaffold. Many researchers have concluded OC binds to HAp crystal through gamma-carboxylation of three residues and has a high affinity for hydroxyapatite [29]. Therefore the cells exhibited osteoblastic-like activity in response to the mineral content and/or HAp in the scaffold in vitro. Additionally, ALP is an early marker for osteogenesis and is highly expressed within 3–7 days in vitro from cells seeded on ceramic-based substrates [11, 30]. This explains the decrease in ALP activity on day 10, as seen in the normalized area and percent area of the ALP stain (Fig. 7b).

Conclusion

A bioengineered replacement for trabecular bone must be porous for cellular infiltration and bioactive to promote the formation of new bone tissue. In this work, we investigated various mineralization techniques to increase the mineral content of full-thickness nanofibrous scaffolds. Electrodeposition mineralization, a common technique used in the metal industry, deposited three times more mineral content than the static and flow mineralization techniques. The addition of the calcium through direct integration of HAp and mineralization promoted osteogenic differentiation of hMSCs as shown by the detection of alkaline phosphatase, calcium and osteocalcin. The end goal of this work is to develop a three-dimensional scaffold for large bone defects. Future work will need to be performed to evaluate the scaffold in vivo.