Differential osteogenicity of multiple donor-derived human mesenchymal stem cells and osteoblasts in monolayer, scaffold-based 3D culture and in vivo

Cell culture

Human primary osteoblasts (hOBs) were isolated from nonsclerotic, trabecular bone taken from the tibial plateau as described previously [50] (three healthy male donors, age 62–67 years). The procedure was approved by the Ethics Committee of the Queensland University of Technology and the Prince Charles Hospital (Approval number 0600000232) and conducted according to the World Medical Association Declaration of Helsinki. Trabecular bone-derived osteoblastic cells have extensively been characterized regarding their morphological and functional criteria [22, 41, 50, 54, 60]. Cells were expanded to passage 3 for subsequent experiments.

Human primary mesenchymal stem cells (hMSCs) were purchased from the Tulane Center for the Preparation and Distribution of Adult Stem Cells (http://www.som.tulane.edu/gene_therapy/distribute.shtml, three healthy male donors, age 35–42). The multipotent differentiation potential and cell surface marker profile of these cells have been described previously [13, 34, 58]. After pooling, passage 3 hOBs and hMSCs were maintained in α-minimal essential medium (α-MEM, Invitrogen, Melbourne, Victoria, Australia) supplemented with 10% fetal bovine serum (FBS, Invitrogen, Melbourne, Victoria, Australia), 100 IU/ml penicillin (Invitrogen, Melbourne, Victoria, Australia), and 0.1 mg/ml streptomycin (Invitrogen, Melbourne, Victoria, Australia).

2D culture

HOBs and hMSCs were seeded at a density of 3000 cells/cm² into 6-well polystyrene tissue culture plates (Nunc, ThermoFisher Scientific, Scoresby, Victoria, Australia). At confluency, mineralization was induced as described previously [50] over a period of 28 days. Controls were cultured in normal culture medium.

To assess the level of mineralization after osteogenic induction, alizarin red s staining and a Wako HRII calcium (Wako, Osaka, Japan) assay were performed as described previously [50].

Alkaline phosphatase activity

Two-dimensional cultures were washed and incubated with 1 ml phenol red- and serum-free α-MEM for 24 h. Then, 100 μl of media supernatant was transferred in triplicates to a 96-well plate (Nunc). After 3 h of incubation with 100 μl of 1 mg/ml p-nitrophenylphosphate/0.2 M Tris(hydroxymethyl)-aminomethane (Tris) buffer (Sigma-Aldrich, Castle Hill, New South Wales, Australia), optical density at 405 nm was measured using a plate reader (Polar Star Optima, BMG Labtech, Offenburg, Germany). Alkaline phosphatase (ALP) activity was normalized against the sample deoxyribonucleic acid (DNA) content determined using a Quant-iT PicoGreen dsDNA assay kit (Invitrogen, Melbourne, Victoria, Australia).

Scaffold fabrication

Medical-grade polycaprolactone tricalcium phosphate (mPCL-TCP) scaffolds produced by fused deposition modeling (FDM) were purchased from Osteopore International (www.osteopore.com.sg). The scaffold properties have been characterized as described previously [51].

Scaffold seeding and 3D culture

Prior to cell seeding, scaffolds were prepared as described by Reichert et al. [51]. The 150,000 hOBs or hMSCs suspended in 50 μl media were equally distributed onto the scaffolds and incubated for 2 h at 37°C, 5% CO₂ to allow cell attachment. One milliliter of α-MEM/10% FBS was then added carefully. Cells were cultured in osteogenic media over a period of 4 weeks with semi-weekly media changes.

Seeding efficiency and proliferation in 3D culture

To assess seeding efficiency and proliferation of hMSCs and hOBs on the 3D scaffolds, total DNA was quantified using the Quant-iT PicoGreen™ proliferation assay (Invitrogen, Melbourne, Victoria, Australia) according to the manufacturer’s instructions. Fluorescence of triplicates was measured using a plate reader (Polar Star Optima, BMG Labtech, Offenburg, Germany) at 485 nm/520 nm excitation/emission wavelength.

Cell viability in 3D culture

After 28 days of culture, seeded scaffolds were incubated with 2 μg/ml fluorescein diacetate (FDA, Invitrogen, Melbourne, Victoria, Australia) and 10 μg/ml propidium iodide (PI, Invitrogen, Melbourne, Victoria, Australia) in phenol red-free medium (α-MEM, Invitrogen, Melbourne, Victoria, Australia) at 37°C for 15 min. The specimens were then rinsed with phosphate-buffered saline (PBS, Invitrogen, Melbourne, Australia), and images of the hydrated samples were captured using a Leica SP5 laser scanning confocal microscope (Wetzlar, Germany).

Phalloidin staining

The 3D cultures were fixed with 4% paraformaldehyde/PBS (PFA) for 20 min and permeabilized using 0.2% Triton X-100/PBS for 5 min. After washing with PBS, samples were blocked for 10 min in 2% bovine serum albumin (BSA, Sigma-Aldrich, Castle Hill, New South Wales, Australia) in PBS. Samples were then incubated with 3 μl/ml PicoGreen™ (Invitrogen, Melbourne, Victoria, Australia) and 8 U/ml rhodamine-conjugated phalloidin (Invitrogen, Melbourne, Victoria, Australia) in 2% BSA/PBS for 1 h. After rinsing with PBS, the hydrated samples were imaged using a Leica SP5 laser scanning confocal microscope.

Scanning electron microscopy (SEM)

On day 28, 3D cultures were fixed with 3% (v/v) glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.3) overnight at 4°C. After washing in 0.1 M sodium cacodylate buffer, they were dehydrated through a graded series of ethanol. Finally, samples were incubated in hexamethyldisilazane (HMDS, Prositech, Thuringowa, Queensland, Australia) for 60 min and air dried. Specimens were mounted and gold coated (SC500, Bio-Rad) prior to visualization using a Quanta 200 scanning electron microscope (FEI).

RNA isolation, primer design, and qRT-PCR

Total ribonucleic acid (RNA) was harvested in triplicate from control as well as differentiated samples at day 28. Samples were isolated using Trizol (Invitrogen, Melbourne, Victoria, Australia) according to the manufacturer’s instructions. cDNA was synthesized from 1 μg of total RNA using the SuperScript III kit (Invitrogen, Melbourne, Victoria, Australia). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed on a 7900HT FAST Real Time PCR system (Applied Biosystems, Melbourne, Victoria, Australia) using a 384-well plate layout. Templates and reagents were aliquoted using an Eppendorf 5075 epMotion pipetting robot (Quantum Scientific, Brisbane, Queensland, Australia). The following volumes were used per reaction (total volume 10 μl): 5 μl of 2× SybrGreen reaction mix (Applied Biosystems, Melbourne, Victoria, Australia), 1 μl of forward and reverse primer (1 μM, Table 1), 2 μl cDNA template dilution (1:10 from stock), and 1 μl RNAse and DNAse free water (Invitrogen, Melbourne, Victoria, Australia). The thermocycling conditions were 1 cycle of 10 min at 95°C, 40 cycles of a 10-s step at 95°C followed by a 1 min step at 60°C. Levels of genes of interest were normalized to the geometrical mean of three housekeeping genes (18s, GAPDH, and β-actin).

Scaffold implantation

Cell scaffold constructs were implanted subcutaneously in dorsal pockets of 7-week-old female non-obese diabetic/severe combined immunodeficiency (NOD/SCID) mice (Australian Research Council, Perth, Australia). Approved by the Ethics Committee of Queensland University of Technology (Approval number 0900000915) and conducted according to the World Medical Association Declaration of Helsinki. Experimental groups included (1) scaffold+hMSCs; (2) scaffold+hOBs; (3) scaffold+hMSCs+recombinant human bone morphogenetic protein 7 (rhBMP-7); (4) scaffold+hOBs+rhBMP-7. To administer rhBMP-7 (5 μg/scaffold; Olympus Biotech, Mt Waverley, Victoria, Australia) to the inner duct of the scaffold intraoperatively, 50.5 μl of fibrin glue (Tisseel, Baxter, Old Toongabbie, New South Wales, Australia) were used [51]. Eight weeks after transplantation, specimens were retrieved and fixed with 4% PFA for 24 h followed by 48 h in 2% PFA. Samples were stored in 70% ethanol until analysis.

Calcein labeling and live imaging

Five days prior to termination, a fluorescent calcein solution was prepared according to the manufacturer’s instructions (Sigma-Aldrich, Castle Hill, New South Wales, Australia) (1 mg/ml, pH 7.4) and injected intraperitoneally (10 mg/kg body weight). Before euthanasia, mice were imaged (day 5) under isoflurane anesthesia using a Xenogen IVIS^® imaging system (Glen Waverley, Victoria, Australia) to detect calcein deposits within the newly formed tissue-engineered bone.

μCT analysis and biomechanical testing

Microcomputed tomography (μCT) analysis and biomechanical testing was performed as described by Reichert et al. [51].

Histochemistry and immunohistochemistry

Following paraffin embedding [51], 5-μm sections were deparaffinized, rehydrated, rinsed with distilled water, and immersed in 0.2 M Tris-hydrochloric acid (Tris-HCl) buffer (pH 7.4). Sections were stained with Meyer’s Haematoxylin and Eosin using a standard protocol [23] or processed for immunohistochemistry. For immunohistochemistry, sections were incubated with 3% hydrogen peroxide (H₂O₂) for 20 min at room temperature (RT). After three washes with 0.2 M Tris-HCl (pH 7.4,) sections were incubated for 10 min with a ready-to-use Proteinase K solution (Dako, Noble Park, Victoria, Australia) or in sodium citrate buffer at 95°C for 4 min. After blocking with 2% BSA/PBS for 20 min, samples were incubated with 2 mg/ml dilution primary mouse anti-human osteocalcin (Abcam, ab13418, Waterloo, New South Wales, Australia) or 1 mg/ml rabbit anti-human von Willebrand factor (vWF) antibody (Milipore, AB7356, Kilsyth, Victoria, Australia) or human specific nuclear mitotic apparatus protein 1 (NuMA) antibody (Abcam, S2825, Waterloo, New South Wales, Australia) in 2% BSA/PBS overnight at 4°C in a humidified chamber (Table 2). Controls were incubated with mouse or rabbit IgG (Invitrogen, Melbourne, Victoria, Australia). After washing with PBS, sections were incubated for 30 min with peroxidase-labeled goat anti-mouse or anti-rabbit antibodies (Dako EnVison+ Dual Link System Peroxidase, Noble Park, Victoria, Australia), respectively. Sections were incubated with 3,3-diaminobenzidine (DAB) substrate (Dako, Noble Park, Victoria, Australia) until a brown staining was evident, and counterstained with hematoxylin. After dehydration in an ethanol series (50%, 70%, 95%) and xylene, sections were mounted with Eukitt^® mounting medium (Sigma-Aldrich, Castle Hill, New South Wales, Australia), and images were captured using a Zeiss Axio Imager A.1 (Jena, Germany).

Blood vessel counting

Blood vessel counting was performed on 15 representative vWF-stained slides for each experimental condition. Herein, five randomly selected regions of interest were photographed at a magnification of 60× using a Carl Zeiss Axio Imager A1 microscope (North Ryde, New South Wales, Australia). Blood vessels were then counted using the AxioVision 3.1 software package, and the mean numbers of vessels/mm² was calculated [7].

Tartrate-resistant acid phosphatase (TRAP) staining

The presence of osteoclasts was assessed using TRAP staining as previously published [51].

Undecalcified poly-methyl-methacrylate sections

Samples were dehydrated through a graded series of ethanol (30 min in 70%,1 h in 90%, 95% and 100% ethanol) and processed through xylene for 40 min three times. Samples were then infiltrated with methyl methacrylate (MMA, Sigma-Aldrich, Castle Hill, New South Wales, Australia) for 4 h (vacuum conditions) and subsequently incubated with MMA containing 3% polyethylene glycol (PEG) (three changes, 24 h each, vacuum conditions). To initiate polymerization, the catalyst Perkadox (0.3%) was added to the MMA/PEG solution, which was then incubated at 50°C for 10 min. Specimens were put to rest until fully polymerized. Embedded samples were cut into 5-μm sections using an automated sledge microtome (Reichert-Jung, Polycut SM 2500), stretched with 70% ethanol and mounted onto gelatin-coated microscope slides. Sections were overlaid with a plastic film, and slides were clamped together and dried for 12 h at 60°C. Sections were deplastified in xylene, rehydrated, and stained using a combined von Kossa/van Gieson staining procedure [59].

Histomorphometry

To determine the amount of mineralization on von Kossa/van Gieson-stained sections, semiquantitative analysis was performed using Image J software as described by Reichert et al. [51].

Scanning electron microscopy for osteocytes in mineralized matrix

A mirror finish polish was achieved on the MMA-embedded sections by sequential wet sanding with 400, 600, and 1200 grit sandpapers, and a final polish with 1.0 α alumina powder polishing using a soft cloth polishing wheel. These polished faces were etched with 37% phosphoric acid for 2–10 s, and washed in sodium hypochlorite (bleach) for 5 min, followed by a brief rinse in dH₂O (distilled water). After drying, the blocks were gold sputtered coated (Leica EM SCD005) and then examined with a FEI Quanta 200 Environmental scanning electron microscope (SEM) operating at 10 kV.

To calculate the osteocyte density within the mineralized matrix, 10 fields of view were examined at 1000× magnification, and the area of the mineralized matrix present in each field was calculated with Image J software.

Statistical analysis

Statistical analysis was performed using analysis of variance (ANOVA, SPSS 19), and p<0.05 were considered significant, the Tukey post hoc test was performed following ANOVA. All experiments were performed in triplicate. Results are presented as the mean±standard error of the mean.

2D hMSC cultures show increased matrix mineralization compared to hOB cultures

The potential of hMSCs and hOBs to deposit a mineralized extracellular matrix was investigated by culturing these cells in osteogenic differentiation medium. After 4 weeks of osteogenic induction in monolayer culture, extensive amounts of alizarin red-positive mineral deposits were seen in both hMSC and hOB cultures (Figure 1A–D). In contrast, control cultures grown under non-osteogenic conditions showed no significant alizarin red staining. This was in line with the significantly lower calcium levels measured utilizing the Wako HR II calcium assay (d14: hOB 404.66 ng±63.73 vs. 524.67 ng±44.51; hMSC 1694.81 ng±260.99 vs. 712.80 ng±177.33; d28: hOB 944.76 ng±40.63 vs. 492.70 ng±167.62; hMSC 4007.62 ng±949.84 vs. 888.27 ng±160) (Figure 1E). Compared to hMSCs, hOBs formed significantly fewer mineralized nodules (p<0.05), which was in accordance with the lower amounts of calcium detected in hOB cultures after 14 and 28 days (Figure 1B). We further assessed alkaline phosphatase (ALP) activity, a marker of osteogenic differentiation, at day 21. ALP activity was significantly increased in osteogenically induced hOBs (0.28±0.05) and hMSCs (0.34±0.02) (p<0.05) compared to non-osteogenic controls (hOB 0.18±0.01; hMSC 0.23±0.01) (Figure 1F). To compare hOB and hMSC proliferation, DNA contents were quantified over a culture period of 14 days using a PicoGreen assay (Figure 1G). Starting from day 5, significantly higher cell numbers were detected for hOBs (683.24 ng±23.82) compared to hMSCs (376.34 ng±9.21) (p<0.05) indicating a higher proliferative potential of hOBs.

Figure 1:

Assessment of mineralization, alkaline phosphatase activity and proliferation. Under osteogenic conditions extracellular matrix calcium concentrations, visualized by alizarin red staining (A–D) and quantified using a Ca2+ assay (E), increased significantly on days 14 and 28 also stimulating alkaline phosphatase activity (F). In 2D, hOBs showed a higher proliferation rate compared to hMSCs (G) (*statistical significance).

In 3D culture, hOBs show a higher proliferation potential than hMSCs

Three-dimensional cell culture systems have become increasingly popular in cell biology as they are believed to reflect native tissue conditions much closer compared to conventional 2D cell cultures [44]. Moreover, scaffolds utilized for 3D in vitro cultures can also be used in analogous in vivo experiment consecutively allowing for a better comparison of obtained in vitro and in vivo data. Thus, hOBs and hMSCs were cultured on mPCL-TCP scaffolds. After 28 days, light microscopy, confocal laser scanning microscopy, and SEM analysis revealed elongated, spindle-shaped, osteoblast-like cell morphologies for both cell types. Cells formed a dense, interconnected 3D network throughout the pores of the mPCL-TCP scaffolds (Figure 2). FDA/PI staining indicated a high cell viability of >90% (Figure 2C and H). Furthermore, cell proliferation of hOBs and hMSCs in 3D culture was compared. Similar to the findings in 2D cultures, hOBs (d28: 2381.12 ng±100.37) proliferated at a higher rate when compared to hMSCs (d28: 1613.6 ng±86.34) as indicated by the higher DNA contents detected after 14 and 28 days (Figure 2K); furthermore, higher calcium levels were detected in 3D hMSC cultures (0.15 mg±0.03) compared to hOB cultures (0.08 mg±0.03).

Figure 2:

Morphology, viability and proliferation of 3D cultures. Cultures of hMSCs and hOBs on mPCL-TCP scaffolds on day 28. Both cell types showed high rates of survival (C, H), displayed a spindle-shaped morphology (B, G), proliferated well, and synthesized mineralized extracellular matrix forming dense interconnected networks on the scaffold struts and within its pores (D, E, I, J).

Osteogenic markers are expressed differentially in 2D and 3D cultures

To investigate the influence of 2D vs. 3D culture conditions on osteogenic marker gene expression levels, hOBs and hMSCs were grown for 28 days in monolayer and on mPCL-TCP scaffolds. Under osteogenic conditions, hMSCs and hOBs upregulated mRNA levels of ALP both in 2D and 3D culture. For osteogenically induced hMSCs, higher relative expression levels of ALP were detected in 2D rather than in 3D culture. In contrast, expression levels in hOB cultures were found to be significantly higher in 3D than in monolayer culture.

Osteocalcin (OC), a late osteogenic expression marker, was expressed at relatively low levels in hMSC and hOB controls. Human OBs exhibited increased levels of OC expression levels under osteogenic stimulation in 2D, while significantly higher OC levels were found in 3D cultures compared to 2D cultures. In hMSCs, low levels of OC expression were found for both 2D and 3D cultures.

Moreover, osteonectin (ON) expression levels were decreased in osteogenic hOB and hMSC cultures compared to controls. Significantly increased levels were found for hOBs in 3D cultures, opposed to hMSCs. Generally, in 3D cultures PTHrP, OPG, RANKL, SOST, and DMP1 levels were remarkably higher than in 2D cultures. At the same time, PTHrP, RANKL, and DMP-1 expression decreased under osteogenic stimulation, while expression levels of OPG and SOST increased significantly. Relative expression of transforming growth factor-β (TGF-β) decreased under osteogenic conditions for both cell types in 2D and 3D cultures. However, TGF-β expression levels were significantly higher in 2D cultures. Taken together, both cell types expressed osteogenic markers when stimulated with osteogenic media, while significant differences in gene expression were observed between 2D and 3D cultures (Figure 3).

Figure 3:

Expression of osteogenic marker genes of hMSCs and hOBs cultured in 2D and 3D culture. For hOBs, higher levels of expression for ALP (A), osteonectin (B), osteocalcin (C), and PTHrP (D), osteoprotegerin (G), and RANKL (H) were found in 3D culture, while collagen type I (E) and TGF-β (F) were expressed to a lower extent. #Significant difference between 2D and 3D at day 28 within one group. *Significant difference between control and osteogenically induced samples at day 28 either in 2D or in 3D.

In vivo, hOBs deposit more bone in the presence of rhBMP-7 when compared to hMSCs

To analyze the potential of hMSCs and hOBs to induce bone formation in vivo, scaffolds with both cell types were implanted subcutaneously into NOD/SCID mice with or without rhBMP-7. Prior to euthanasia at 8 weeks, calcein was injected to visualize the tissue-engineered bone constructs in vivo. A seemingly higher fluorescent signal was detected for scaffolds implanted in combination with rhBMP-7 (Figure S1 A). When explanting the scaffolds, all constructs appeared well vascularized (Figure S1 D, E). MicroCT revealed little and scattered bone formation in scaffolds seeded with cells only. More deposited bone was observed when cell-seeded scaffolds were combined with rhBMP-7. The highest amount of mineralized tissue had formed in tissue-engineered constructs consisting of hOBs and rhBMP-7 (5.85 mm³), followed by hMSCs combined with rhBMP-7 (4.79 mm³), hMSCs (3.33 mm³), and hOBs (2.67 mm³) (Figure 4V). A similar bone mineral density was found for all groups with values ranging from 619.98 to 652.71 mg HA/cm³. Concomitantly, with an increased amount of mineralized tissue, higher compressive stiffness values were measured (Figure 4U, hOB 32.26 N/mm±1.42, hOB-BMP 45.85 N/mm±3.66, hMSC 33.76 N/mm±1.7, hMSC-BMP 34.54 N/mm±1.0). In H&E staining, newly formed bone (NB), adipose tissue (AT), fibrous connective tissue (CT), and blood vessel formation (Figure 4D, E, I, J, S, T, asterisks) could be observed. In any group, bone formation occurred preferably in peripheral regions of the scaffold, but was constrained by the scaffold boundaries. The formation of heterotopic bone marrow (BM) was only observed in groups including BMP-7 (Figure S2). Von Kossa/van Gieson staining confirmed the highest amount of mineralization found in the hOB-BMP-7 group (Figure 4 W, hOB 4.83%±0.35, hOB-BMP 10.52%±0.79, hMSC 5.75%±0.49, hMSC-BMP 7.87%±0.53). TRAP staining indicated high osteoclastic activity in all groups with rhBMP-7, while no osteoclasts were detected in the other experimental groups (Figure 5A–D). Osteoclasts were lining the outer boundary of areas of new bone formation in close proximity to the surrounding fibrous connective tissue. Furthermore, high levels of OC were detected in the bone matrix, in bone-lining osteoblasts, and in osteocytes (arrows). Multiple blood vessels were observed in both the periphery and center of the transplanted constructs as analysed by v. Willebrand factor (vWF) stainings. A significantly higher blood vessel density was determined for hMSC-BMP-7 constructs (19.47/mm²±2.61) and hOB-BMP-7 samples (18.95/mm²±1.82) compared to scaffolds that had been implanted in the absence of BMP-7 (11.91/mm²±5.42 for hMSC and 12.42/mm²±3.93 for hOB seeded scaffolds).

Figure 4:

MicroCT 3D reconstructions, von Kossa/van Gieson and H&E-stained sections, quantification of mineralization, and biomechanical testing. MicroCT 3D reconstructioins (first column), von Kossa/van Gieson (second and third columns), and H&E (fourth and fifth column) staining showed the highest amount of mineralized tissue in combination with rhBMP-7.

Figure 5:

TRAP, OC, vWF, and NuMA staining on paraffin-embedded specimens. High osteoclastic activity was detected in all rhBMP-7-groups but not in control groups (A–D). In the case of bone formation, strong extracellular matrix-associated and cytoplasmatic OC expression was detected (arrows). Multiple areas positive for vWF were observed (I–L). NuMA staining visualized the presence of human cells in the tissue-engineered bone constructs (M–P).

To investigate if the seeded hMSC and hOB cells survived the implantation and contributed to bone formation 8 weeks after transplantation, immunohistochemical staining specific to human, not mouse, cell nuclei was performed. A positive staining was observed in the hOB, hMSC-BMP-7, and the hOB-BMP-7 group (Figure 5N–P) whereas absent in hMSC samples (Figure 5M). NuMA-positive cells were mainly detected in the fibrous tissue surrounding the scaffold, as well as in areas of increased bone formation in the hMSC-BMP-7 and the hOB-BMP-7 groups. NuMA-positively stained cells were absent within the mineralized tissue as well as in the negative controls.

Higher osteocyte density in hOB-rhBMP-7 group than in the hMSC-rhBMP-7 group

Resin-cast SEM technique was used to image and characterize resin-embedded undecalicifed tissue-engineered bone constructs. It has previously been demonstrated [18] that after MMA infiltration of bone, a standard histological method for mineralized samples, acid etching the surface allows for the removal of the mineralized matrix, exposing a replica morphology of the lacuna-canalicular system at the bone-scaffold interface. An abundance of well-preserved, morphologically intact osteocytes with a high degree of interconnecting processes, suggesting a functional syncytium (Figure 6A–F) could be shown in the hOB+rhBMP-7 and hMSCs+rhBMP-7 group. In some areas, osteocytes were observed interacting with blood vessels. Empty lacunae were occasionally visualized for both these samples. Significantly more osteocytes per unit area were observed in the hOBs+rhBMP-7 (33.87/50 mm²±5.63) group (Figure 6G) compared to hMSCs+rhBMP-7 (18.9/50 mm²±3.61).

Figure 6:

SEM of osteocytes and the lacuno-canalicular network and osteocyte quantification. Scanning electron micrographs of resin-cast sections demonstrating the degree of connectivity between osteocytes, and osteocytes with blood vessels (bv) in the mineralized matrix in the hMSC+rhBMP-7 (A–C) and hOB+rhBMP-7 (D–F) samples. A higher amount of osteocytes was found in the hOB+rhBMP-7 group (G).