Bone Repair Using Periodontal Ligament Progenitor Cell-seeded Constructs

PDLC Isolation and Culture

Mandibles were dissected from 6 adult Sprague-Dawley rats and washed twice in phosphate-buffered saline solution (DPBS; GIBCO, Grand Island, NY, USA) supplemented with 100 U/mL penicillin, 100 µg/mL streptomycin (GIBCO), and 0.25 µg/mL of amphotericin B (GIBCO). The first and second molars were extracted from each mandible, and PDLCs were obtained by the explant method. The explants were cultured in alpha-minimum essential medium (alpha-MEM; GIBCO) supplemented with 10% heat-inactivated fetal bovine serum (FBS; GIBCO), 100 U/mL penicillin, and 100 µg/mL streptomycin at 37°C in a humidified atmosphere of 5% CO₂. Outgrowth cells were seeded in T-75 flasks (GIBCO) in the same media and incubation conditions. The PDLCs at passage 4 were used for the experiments.

Additionally, PDLCs expressing green fluorescent protein (GFP) were used for in vivo tracking experiments (see Appendix Materials & Methods).

Phenotypic Characterization of PDLCs

Viable PDLCs at passage 4 were labeled with respective antibodies (Appendix Table 1), washed, and analyzed on a FACScan flow cytometer (BD Biosciences, San Jose, CA, USA) according to standard procedures (Moll et al., 2011).

In vitro Mineralization Assay

PDLCs were seeded into 24-well plates (Costar, Corning Inc., Corning, NY, USA; 4 × 10³ cells/cm²) and cultured in standard osteogenic media [alpha-MEM supplemented with 10% FBS, 10 mM β-glycerophosphate, 50 µg/mL ascorbic acid, and 10^–8 M dexamethasone (all from Sigma-Aldrich, St. Louis, MO, USA)] for 21 days. The cell monolayer was washed with DPBS, and the mineralization was visualized by staining with 2% solution of Alizarin Red S (Merck, Darmstadt, Germany), pH 4.2, for 5 min.

PDLC Seeding

For alkaline phosphatase activity (ALP), MTT, and gene expression assays, the HA microparticles or HA-ECM constructs were placed into 96-well culture plates (Costar, Corning Inc.; 10 mg HA/well), and PDLCs were seeded onto the scaffolds (1.5 × 10⁴ cells/well) in alpha-MEM culture media for 24 hrs, then switched to standard osteogenic media in a humidified atmosphere of 5% CO₂ at 37°C. The experiments were performed as 2 independent runs, with cell seeding performed on 2 different occasions. Bone marrow stromal cells were used as control (see Appendix).

For the in vivo experiments, HA-ECM constructs were placed into 24-well culture plates (30 mg HA/well); a 400-µL quantity of PDLC suspension in alpha-MEM was added to each construct at an initial density of 2 × 10⁵ viable cells per well. After 30 min of incubation in a humidified atmosphere of 5% CO₂ at 37°C, alpha-MEM was added to a final volume of 1 mL per well. After 24-hour incubation, the PDLCs were cultured in standard osteogenic media for 3 or 14 days.

In vitro PDLC Viability

The MTT colorimetric assay was used to assess PDLC proliferation and viability according to the manufacturer’s instructions (Roche, Indianapolis, IN, USA) as previously described (Tour et al., 2011). The MTT assay was performed in triplicate at days 1, 3, 7, and 14 of culture.

ALP Assay

The assay was performed on cell lysates in triplicate at days 1, 3, 7, and 14 of culture as previously described (Castano-Izquierdo et al., 2007). To normalize ALP activity, we measured total protein content using the Pierce 660-nm protein assay according to the manufacturer’s instructions (Thermo Fisher Scientific, Rockford, IL, USA).

RNA Isolation and cDNA Synthesis

The cell-seeded HA microparticles and HA-ECM constructs were rinsed twice with sterile DPBS, then a 100-µL quantity of cell lysis buffer (Qiagen, Valencia, CA, USA) was added to each sample. The contents of the wells were transferred to a QIAshredder column (Qiagen) for homogenization, followed by total RNA extraction according to the manufacturer’s protocol (RNeasy Mini Kit, Qiagen). The cDNA synthesis was carried out with the iScript cDNA synthesis kit according to the manufacturer’s protocol (BioRad, Hercules, CA, USA) in a T1 Thermocycler (Whatman Biometra, Göttingen, Germany) followed by spectrophotometric quantification and quality assessment with the NanoVue spectrophotometer (GE Healthcare, Buckinghamshire, UK).

TaqMan Gene Expression Assays

The list of bone-related genes and TaqMan assay IDs are presented in the Appendix Table. Fold differences were calculated by the delta-delta C _T method (Schmittgen and Livak, 2008), with GAPDH as the endogenous housekeeping gene. The results were calibrated to PDLCs seeded on HA prior to the osteogenic treatment (day 1). Three samples for each set-up (substrate; time-point) were used. The experiment was repeated twice with the same batch of cells and a different batch of biomimetic constructs. The presence of proteins encoded by OPN and BMP2 genes in the HA-ECM constructs seeded with PDLCs was assessed by immunocytochemistry (see Appendix).

In vivo Experimental Design

The in vivo experiment was approved by the Research Ethics Committee of Karolinska University Huddinge Hospital in accordance with the policy on human care and use of laboratory animals, and the ARRIVE Guidelines. The critical-sized (8-mm) calvarial defect model was used (Tcacencu and Wendel, 2008). Twenty-four adult Sprague-Dawley male rats (weight ~ 350 g each) were randomly and equally divided into 4 groups (n = 6 animals each) with regard to the type of local treatment: HA alone, HA-ECM, HA-ECM+PDLC3, or HA-ECM+PDLC14 (HA-ECM constructs seeded with the PDLCs and cultured in standard osteogenic media for 3 or 14 days, respectively).

Tissue Preparation, Histology, and Histomorphometry

The rats were sacrificed by CO₂ inhalation at 12 wks post-operatively. The calvarial bone was surgically retrieved and histologically processed as previously described (Tcacencu and Wendel, 2008). Five-µm serial sections were prepared parallel to the sagittal line and stained with Masson’s trichrome for the assessment of the general morphology, new bone formation, connective tissue ingrowth, and the residual scaffold area. The amount of the newly formed bone was expressed as the percentage of total newly formed bone area (NFB) relative to the total possible area (defect area without HA scaffold) for new bone ingrowth = NFB/(total defect area – scaffold area) × 100%.

In vitro Characterization of PDLCs

The isolated PDLCs maintained typical spindle-shaped fibroblast-like morphology and a high proliferation rate throughout passages. Flow cytometry phenotyping of undifferentiated PDLCs revealed positivity for typical mesenchymal markers CD90, CD29, and CD73 and negativity for representative lymphoid (CD3 and CD45), myeloid (CD11b), and endothelial markers (CD31) (Fig. 1B). Upon osteogenic stimulation of PDLCs, mineral depositions were detected as early as day 10, followed by a robust osteoblast-like differentiation by day 21 in culture (Fig. 1A).

OPEN IN VIEWER

Figure 1.

In vitro characterization assays of the PDLCs. Mineralization assay; cells cultured for 3 wks in osteoinductive (+) or basal (–) media; appearance of the PDLC culture stained with Alizarin Red S, phase-contrast micrograph (4x, on the right) (A). Flow cytometry analysis of the PDLCs at passage 4 (B). ALP activity of PDLCs on HA alone (blue), HA-ECM constructs (red), or tissue culture plastic (grey) normalized to total cell protein amount (C). Metabolic activity (MTT assay) of PDLCs seeded on HA alone, HA-ECM, or tissue culture plastic (D). Data represent means ± standard deviation (n = 5, five replicates per time-point for each experimental condition), *p < 0.05, two-tailed unpaired t test.

The PDLCs seeded on the biomimetic constructs proliferated and actively secreted ECM, resulting in a dense pellet-like structure (Fig. 3B). Immunocytochemistry revealed the presence of OPN and BMP-2 in the ECM of the cell-seeded constructs at both 3 and 14 days of culture (Appendix Fig. 2). The rate of PDLC proliferation was already significantly higher on HA-ECM constructs, compared with HA or plastic, at day 3 in culture (Fig. 1D). No significant difference in ALP activity was found between the 2 substrates at day 3. However, ALP activity gradually increased during culture, exhibiting significantly higher values on HA-ECM after 7 days in osteogenic culture, with the highest value on day 14 in both groups (Fig. 1C; Appendix Fig. 1).

RT-qPCR revealed that most of the osteoblast-related genes were strongly up-regulated on both substrates, with the exception of Runx2 on HA (Fig. 2). OC and ALP genes were highly up-regulated on HA-ECM at all time intervals, reaching a 300- and 100-fold increase on day 14, respectively. Col1 was continuously expressed on both substrates, displaying higher expression levels on HA-ECM than HA at days 3 and 7, followed by an activity decrease at day 14. There was a significant OPN up-regulation in PDLCs seeded on HA at days 3 and 7, with no difference compared with HA-ECM at day 14. BMP-2, BMP-4, IGF-1, and FGF-1 had similar expression patterns, featuring higher levels on the HA at day 3 followed by a significant up-regulation on HA-ECM at both days 7 and 14. FGF-2 and VEGF expression levels were characterized by a temporal increase at day 3, returning to baseline after 7 days of osteogenic treatment. PDGF was continuously up-regulated on both substrates, with significantly higher levels detected at day 3 only. BMP-7 was down-regulated in all groups, with the lowest level detected at day 14.

OPEN IN VIEWER

Figure 2.

Profiles of bone-related genes expressed in PDLCs at different time intervals in osteogenic culture after seeding onto HA alone (blue) or HA-ECM constructs (red), as measured by RT-qPCR. The data represent the mean fold difference in expression of a total of n = 3 replicates per time-point for each experimental condition, normalized to GAPDH mRNA expression and calibrated to HA day 1 (undifferentiated cells prior to osteogenic treatment) baseline level, as indicated by the dashed grey line. Heat map reflecting the log-scale temporal expression of investigated genes (bottom).

In vivo Assessment

All rats tolerated the surgery well. The histology revealed large amounts of newly formed bone covering the defects at 12 wks post-operatively (Fig. 3D). The interfaces between biomimetic constructs, newly formed bone, or host bone exhibited 2 distinct morphological types: direct bone-to-bone contact or connective tissue fibers inserted into host bone tissue, resembling the morphology of a native periodontal ligament interface (Fig. 3C). Histomorphometric analysis demonstrated significantly improved bone repair elicited by HA-ECM constructs seeded with PDLCs and cultured for 14 days compared with those cultured for 3 days (p < 0.05) (Fig. 3E). The results for residual scaffold content and connective tissue ingrowth are shown in Appendix Fig. 3.

Tracking of Implanted PDLCs within the Wound Sites

The GFP⁺ PDLCs were detected at 1 and 4 wks after surgery. The GFP⁺ cell number decreased significantly over time. The PDLCs were located mainly within connective tissue on the periosteal side, with no GFP⁺ cells present in bone tissue (Fig. 4).

OPEN IN VIEWER

Figure 3.

PDLC expansion in the explant model (A). Cell outgrowth from dental root explants (marked with asterisks) at day 5 (A, top) and after reaching confluence at day 21 (A, bottom); bar = 200 µm. Macroview of PDLC-seeded HA-ECM (B), Bar = 5 mm. Representative histology of the interface between host bone and implant (right) or newly formed bone tissue (left); connective tissue fiber insertion (black arrows) and direct bone-bone contact (white arrows); HA microparticles integrated within the newly formed bone, marked with #; bar = 50 µm (C). The histology of calvarial bone defects at 12 wks after surgery (D). Defect margins are outlined with vertical lines; Masson’s trichrome staining, bar = 500 µm. Histomorphometric analysis of the newly formed bone induced by HA alone, HA-ECM, or HA-ECM seeded with PDLCs in osteogenic culture for 3 or 14 days (E). Box-plots represent median, upper and lower quartiles, and minimum and maximum values (n = 6 animals each); *p < 0.05, **p < 0.005, Mann-Whitney U test, Kruskal-Wallis test.

OPEN IN VIEWER

Figure 4.

GFP-positive periodontal ligament progenitor cells prior to implantation (A). GFP detection by fluorescent microscopy (left) and immunocytochemistry (middle); neg. ctrl = section with primary antibody omitted (right); bar = 100 µm. Superimposed image of GFP⁺ PDLC distribution on HA-ECM constructs after 14 days of in vitro culture, demonstrated by green fluorescent signals detected by fluorescent microscopy (B). Semi-opaque HA microparticles (marked with *) are visualized with a few GFP⁺ cells out of focus due to uneven three-dimensional surface topography of the HA-ECM substrate; bar = 100 µm. Metabolic activity of GFP⁺ PDLCs on HA-ECM at different times of in vitro culture (C, top); formazan crystal visualization at 14 days (C, bottom); bar = 100 µm. Immunohistochemistry at 1 and 4 wks following in vivo PDLC implantation (D) reveals GFP expression in cells located within the connective tissue of the defect area, scattered among HA microparticles (marked with *); bar = 50 µm; Wright-Giemsa counterstain; negative control (nc): anti-GFP immunostaining of defect treated with the wild-type PDLC-loaded implant. GFP⁺ PDLC density within a defect site at different intervals post-implantation, represented as cell number per square mm (D, right); the GFP⁺ cell number in the defect area decreased significantly after 4 wks, with single (1-2 cells/defect) or no positive cells detected after 10 wks.

In vitro Osteogenic Differentiation of PDLCs Seeded on Biomimetic Constructs

Cell-delivery substrates capable of mimicking the extracellular environment of bone tissue are considered to be essential for the outcome of bone tissue-engineering therapy (Mauney et al., 2006; Pham et al., 2008). Furthermore, various experimental settings have demonstrated the importance of osteogenic induction of mesenchymal stem cells in culture prior to (Castano-Izquierdo et al., 2007; Washio et al., 2010) or after (Sikavitsas et al., 2003; Uchida et al., 2009) seeding onto the scaffolds.

In the current study, we aimed to investigate the effects of in vitro–generated biomimetic constructs on PDLC osteogenic differentiation by evaluating temporal gene expression and ALP activity. The PDLCs were cultured in standard osteogenic media starting directly from day 1 in culture. Our results demonstrated that the biomimetic constructs significantly enhanced PDLC osteogenic differentiation (increased ALP production and up-regulation of bone-related genes). Furthermore, the gene activity heat map facilitated the screening of the constructs for in vivo experiments.

Interestingly, the OPN gene was expressed to a larger extent on HA alone compared with HA-ECM. We may speculate that OPN mediates PDLC adhesion to HA by means of a cell-binding domain (Siebers et al., 2005). Since the PDLCs seeded on HA alone were more exposed to the mineral surfaces compared with PDLCs seeded on the HA-ECM construct, we may think that the higher OPN up-regulation is induced by cell adhesion to HA. Second, the rate of OPN gene expression may be influenced by other cytokines and growth factors stored in the ECM component, and may be due to different mechano-inductive forces elicited at cell-adhesion sites (Sodek et al., 2000).

We believe that seeding PDLCs onto HA-ECM constructs and culturing them in osteogenic media for 14 days prior to implantation might facilitate the accumulation of osteogenic growth factors within the biomimetic construct and improve cell adaptation to the substrate (Yoshikawa et al., 1996; Uemura et al., 2003).

The MTT values and the typical morphology of the seeded PDLCs demonstrated that both HA microparticles and HA-ECM constructs supported PDLCs attachment and exhibited high cellular compatibility.

The Effects of PDLC-seeded HA-ECM Constructs on Bone Repair

Another objective of our study was to evaluate how the in vitro osteogenic pre-culture period of PDLCs on HA-ECM constructs influences the cells’ ability to regenerate bone in vivo. Based on our in vitro data, 2 types of PDLC-seeded constructs (HA-ECM with PDLCs cultured for 3 days and 14 days) were selected for implantation into calvarial bone defects in rats. We have demonstrated that the 14-day culture period of PDLCs on HA-ECM constructs in osteogenic media is more optimal for calvarial bone regeneration compared with the 3-day culture period. In contrast to our results, Castano-Izquierdo et al. (2007) showed that rat bone-marrow-derived mesenchymal stem cells stimulated with osteogenic media for 4 days induced the largest new bone formation in vivo. However, they used a different substrate (titanium fiber mesh) and a different cell-culture set-up prior to implantation in vivo. Those findings may indicate that culture protocols must be adjusted for each type of scaffold and progenitor cell to achieve the best outcomes in vivo.

We also evaluated the PDLC survival post-implantation in vivo. The PDLC number decreased significantly over time, and no PDLCs were integrated into the newly formed bone. However, the significant increase in new bone formation induced by treatment with the PDLC-loaded scaffolds suggests that the long-term survival and engraftment of the donor PDLCs may not be required for the enhanced bone repair. It is likely that the implanted PDLCs reveal pro-regenerative effects by means of soluble factors and direct interactions with host cells, without generating new bone tissue themselves.

Altogether, our findings demonstrated that in vitro–generated ECM deposited on HA significantly enhances the in vitro osteogenic differentiation of PDLCs. Furthermore, HA-ECM constructs seeded with PDLCs in osteogenic culture for 14 days generated larger amounts of newly formed bone in vivo compared with the constructs cultured for 3 days. Moreover, HA-ECM constructs are easier to handle and have better mechanical and hemostatic properties compared with HA microparticles alone. Our results highlight the PDLC-seeded HA-ECM constructs as a promising tool for craniofacial bone regeneration.