Osteogenic capacity and cytotherapeutic potential of periodontal ligament cells for periodontal regeneration in vitro and in vivo

Abstract

Background: The periodontal ligament cells (PDLCs) contain heterogeneous cell populations and possess stem-cell-like properties. PDLCs have attracted considerable attention as an option for periodontal regeneration. However, the osteogenic differentiation of PDLCs remains obscure owing to variable osteo-inductive methods and whether PDLCs could be directly used for periodontal regeneration without stem cell enrichment is uncertain. The aim of the present study was to clarify the osteogenic differentiation capacity of PDLCs and test PDLCs as an alternative to stem cells for periodontal regeneration.

Methods: We tested the performance of human PDLCs in osteo-inductive culture and transplantation in vivo while taking human bone marrow derived mesenchymal stem cells (hMSCs) as positive control. Proliferation of PDLCs and hMSCs in osteo-inductive condition were examined by MTT assay and colony formation assay. The osteogenic differentiations of PDLCs and hMSCs were assessed by Alkaline phosphatase (ALP) activity measurement, von Kossa staining, Alizarin red S staining and quantitative RT-PCR of osteogenic marker gene including RUNX2, ALP, OCN, Col I, BSP, OPN. We transplanted osteo-inductive PDLCs and hMSCs with hydroxyapatite/tricalcium phosphate (HA/TCP) scaffolds to immunodeficient mice to explore their biological behaviors in vivo by histological staining and immunohistochemical evaluation.

Results: After 14 days of osteo-induction, PDLCs exhibited significantly higher proliferation rate but lower colony-forming ability comparing with hMSCs. PDLCs demonstrated lower ALP activity and generated fewer mineralized nodules than hMSCs. PDLCs showed overall up-regulated expression of RUNX2, ALP, OCN, Col I, BSP, OPN after osteo-induction. Col I level of PDLCs in osteo-inductive group was significantly higher while RUNX2, ALP, OCN were lower than that of hMSCs. Massive fiber bundles were produced linking or circling the scaffold while the bone-like structures were limited in the PDLCs-loaded HA/TCP samples. The fiber bundles displayed strong positive Col I, but weak OCN and OPN staining. The in vivo results were consistent with the in vitro data, which confirmed strong collagen forming ability and considerable osteogenic potential of PDLCs.

Conclusion: It is encouraging to find that PDLCs exhibit higher proliferation, stronger collagen fiber formation capacity, but lower osteogenic differentiation ability in comparison with hMSCs. This characteristic is essential for the successful periodontal reconstruction which is based on the synchronization of fiber formation and bone deposition. Moreover, PDLCs have advantages such as good accessibility, abundant source, vigorous proliferation and evident osteogenic differentiation capacity when triggered properly. They can independently form PDL-like structure in vivo without specific stem cell enrichment procedure. The application of PDLCs may offer a novel cytotherapeutic option for future clinical periodontal reconstruction.

Keywords: Human bone marrow derived mesenchymal stem cells (hMSCs); Human periodontal ligament cells (PDLCs); Osteogenic differentiation; Periodontal regeneration.

Figure 1. Morphology and proliferation of PDLCs and hMSCs in osteo-inductive condition.

(A) PDLCs’ morphology after 14 days of osteo-inductive culture. (B) hMSCs’ morphology after 14 days of osteo-inductive culture. (C) Proliferation curve of PDLCs and hMSCs under osteo-inductive culture. (D) Colony formation assay of PDLCs and hMSCs after 14 days of osteo-inductive culture. (E) Quantitative analysis of colony formation. Scale bars, 100 μm. (Student’s t-test, n ≥ 3; *P < 0.05, **P < 0.01).

Figure 2. Osteogenic differentiation of PDLCs and hMSCs after 14 days of osteo-inductive incubation in vitro.

(A) ALP staining. The PDLCs and hMSCs kept in normal growth medium were used as control, while the PDLCs and hMSCs kept in osteo-inductive medium were assessed as test groups. (B) ALP activity assay. (C) von Kossa staining of PDLCs and hMSCs after 14 days of osteo-inductive incubation. (D) Quantitative analysis of von Kossa staining. (E) Alizarin red S staining of PDLCs and hMSCs after 14 days of osteo-inductive incubation. (F) Quantitative analysis of Alizarin red S staining. Scale bars, 100 μm. (Student’s t-test, n ≥ 3; *P < 0.05, **P < 0.01).

Figure 3. Osteogenic marker genes expression in PDLCs and hMSCs after 14 days of osteo-inductive incubation in vitro.

The PDLCs and hMSCs kept in normal growth medium were used as control, while the PDLCs and hMSCs kept in osteo-inductive medium were assessed as test groups. The expression of RUNX2(A), ALP(B), OCN(C), Col I(D), BSP(E), OPN(F) were analyzed by quantitative real-time PCR. (Student’s t-test, n ≥ 3; *P < 0.05, **P < 0.01).

Figure 4. Transplantation of osteo-inductive PDLCs and hMSCs with HA/TCP scaffolds to SCID mice.

After 12 weeks, the transplants were taken out and performed HE staining. (A) Cell-free HA/TCP control. No evident bone formed or fiber cluster appeared. (B) In hMSCs-loaded HA/TCP samples, ectopic bone formation was found, but no obvious fiber bundles appeared. (C–D) In the PDLCs-loaded HA/TCP samples, massive fiber bundles linked or circled the scaffold (C, arrow) were found while the bone-like structure was limited (D, arrows). Scale bars, 100 μm.

Figure 5. Osteogenic differentiation of PDLCs and hMSCs in vivo.

The expression of Col I, OCN and OPN in the transplants were evaluated by immunohistochemical staining. (A–D) In the PDLCs-loaded HA/TCP samples, the fiber bundles displayed strong positive Col I(B), but weak OCN(C) and OPN(D) staining. (E–H) In the hMSCs-loaded HA/TCP samples, the ectopic bone structure exhibited positive staining of Col I(F), OCN(G) and OPN(H). This result confirmed strong collagen forming ability and considerable osteogenic potential of PDLCs in vivo. Scale bars, 100 μm.