Genetic evidence for the vital function of Osterix in cementogenesis

Abstract

To date, attempts to regenerate a complete tooth, including the critical periodontal tissues associated with the tooth root, have not been successful. Controversy still exists regarding the origin of the cell source for cellular cementum (epithelial or mesenchymal). This disagreement may be partially due to a lack of understanding of the events leading to the initiation and development of the tooth roots and supportive tissues, such as the cementum. Osterix (OSX) is a transcriptional factor essential for osteogenesis, but its role in cementogenesis has not been addressed. In the present study, we first documented a close relationship between the temporal- and spatial-expression pattern of Osx and the formation of cellular cementum. We then generated 3.6-kilobase (kb) collagen type I (3.6-kb Col 1)-Osx transgenic mice, which displayed accelerated cementum formation versus wild-type (WT) controls. Importantly, the conditional deletion of Osx in the mesenchymal cells with two different Cre systems (the 2.3-kb Col 1 and an inducible CAG-Cre estrogen receptor [CreER]) led to a sharp reduction in cellular cementum formation (including the cementum mass and mineral deposition rate) and gene expression of dentin matrix protein 1 (DMP1) by cementocytes. However, the deletion of the Osx gene after cellular cementum formed did not alter the properties of the mature cementum as evaluated by backscattered scanning electron microscopy (SEM) and resin-casted SEM. Transient transfection of Osx in the cementoblasts in vitro significantly inhibited cell proliferation and increased cell differentiation and mineralization. Taken together, these data support: (1) the mesenchymal origin of cellular cementum (from periodontal ligament [PDL] progenitor cells); (2) the vital role of OSX in controlling the formation of cellular cementum; and (3) the limited remodeling of cellular cementum in adult mice.

Fig 1. OSX expression profile correlates with cellular cementum formation during postnatal development

a. Representative H&E staining images show a very thin layer of cellular cementum at 3 weeks (left panel); an expanded cementum layer at 6 weeks (middle panel), and a dramatically larger cementum layer at 6 months (right panel, inset); b. OSX immunohistochemistry stained images show a weak signal in PDL and cementoblast cells at 3 weeks (left panel, inset); a strong signal in PDL and cementoblast cells at 6 weeks (middle panel, inset) and no signal at 6 months (right panel, inset); c. DMP1 signal is apparent in the cemental extracellular matrix at 3 weeks (left panel, inset) and 6 weeks (middle panel, inset) but not at 6 months (right panel, inset); d. 4-wk-old lower jaws obtained from OSX-lacZ knock-in and Dmp1-lacZ knock-in mice, where lacZ is used to trace the endogenous gene (OSX, left panel; Dmp1, right panel) expression pattern, were processed for x-gal stain. OSX-lacZ is expressed in progenitor cells in the PDL region, cementoblasts and cementocytes (left panel). Dmp1 is expressed only in cementocytes and alveolar bone osteocytes (AB, right panel); e. X-gal stain reveals a OSX-lacZ signal not only in cementoblasts/cementocytes but also in progenitor cells in the PDL region with an expression gradient pattern (6-wk-old, left panel; red arrow); the FITC confocal image (6-wk-old, right panel) shows numerous cells in PDLs (likely progenitors for cementoblasts) that surround the cementum and appear to correlate with lacZ positive-stained cells on the left panel (white balls). C, cementum; D, dentin; Cb, cementoblast; Cmc, cementocyte; Od, odontoblasts; PDL, periodontal ligament.

Fig 2. Overexpression of OSX accelerates cementum formation

a, Representative immunohistochemistry assay shows more OSX signals in the 2.3 Col 1-OSX transgenic (Tg) cementoblast, cementocyte and PDL cells at 4 weeks (right panel) compared to the age-matched control (WT, left panel); b. H&E stain image shows a much larger cellular cementum layer in the 2.3 Col 1-OSX Tg (right panel) compared to the age-matched control (left panel); c. The images of toluidine stain (right panel) further confirm this observation. Note that the Tg epical foramen is much narrower than that of the WT. D, dentin.

Fig 3. OSX is expressed in follicle cells but not in HERS cells, and the conditional deletion (CKO) of OSX leads to a sharp reduction in cellular cementum

The first mandibular molars were obtained from mice harboring a floxed allele of OSX and the 2.3 kb Col 1-Cre (activated from embryonic stage) (18) or the CAG-CreER transgene, which permits temporal regulation of the Cre-mediated recombination by tamoxifen (TM) (19). TM was injected (i.p.) for 3 weeks starting at 3 weeks of age, and tissues were harvested at 6 weeks. a. Representative immunohistochemistry assay shows a strong OSX signal in the wild type (WT) cementoblast (CB) and odontoblast cells (Od,) with a weak but reproducible signal in the follicle cells. Note that there is no OSX expression in the HERS cells at two weeks of age (upper panel). There was no detectable OSX signal in the CKO cementoblasts, odontoblasts, or follicle cells (lower panel); b. Toluidine stain shows a sharp reduction of the cellular cementum layer and cementoblasts, supporting cells in the CKO mice at 6 weeks of age (right panels) compared to the age-matched control (left panels); c. Immunohistochemistry assay shows a strong OSX signal in the WT cementoblasts at 6 weeks of age (left panel). There was no detectable OSX signal in the CKO cementoblasts and cementocytes (right panel); c–d. The representative images of backscattered SEM (d) and H&E stain (e) show a sharp reduction in the cellular cementum layer in the 6-week-old CKO mice crossed with CAG-CreER (right panels). D, dentin.

Fig 4. Deletion of OSX gene leads to a reduction of the cementocyte number, dendrites, and mineralization rate

a. Representative resin-casted SEM images show a reduction in cementocyte number (right upper panel) and dendrite number (right lower panel) in the CKO mouse at 4 weeks of age compared to age-matched controls (left panels); b. fluorochrome-labeled sections of the first lower molar from WT control (left panel) and OSX CKO mice (right panel) reveal a reduced mineralization rate in the CKO cementum. The distance between the two injections reflects the cementum formation rate. The area reflects cementum mass; c–d. Quantitative data show a 4-fold reduction in the mineralization rate in the CKO cementum (c), and a significant reduction in the cellular cementum area calculated based on the fluorescence labeled area (d). Data are mean ± SEM, n=4–6, **P<0.01, ***P<0.001. C, cementum; P, pulp; and D, dentin.

Fig 5. Deletion of Osx gene changes molecular markers which are important for cementogenesis

a, Representative in situ hybridization assay shows a sharp reduction in of BSP (right panel) in Osx CKO cementoblasts and cementocytes compared to age matched control (left panel); b, Representative immunohistochemistry assay shows a sharp reduction of DMP1 (right panels) in Osx CKO′ cementoblast (CB) and cementocyte (CMC) cells compared to the age matched control (left panels). BSP, bone sialoprotein; DMP1, dentin matrix protein 1; D, dentin; and B, bone.

Fig 6. Conditional deletion of OSX by tamoxifen (TM) injections after cementum has fully formed has little effect on cementum formation but dramatically alters bone remodeling

a. Representative photomicrographs of three maxillary molars (1^st on the top and 2^nd/3^rd on the bottom) show no apparent difference between the WT (left panel) and the CKO (right panel) mice that received TM injections (two times a week for 3 weeks starting at 6 months; sacrifice was at 7 months); b. The backscattered SEM images obtained from the mandibles show dramatic alveolar bone loss and defects in osteocyte morphology in the CKO mice (right and lower images); c–d. The images of backscattered SEM (c) and resin-casted cementocytes (d) revealed no apparent difference between the WT (left panels) and the CKO cellular cementum mass and cementocyte morphology (right panels). (D=dentin).

Fig 7. Overexpression of OSX in cementoblasts inhibited cell proliferation, induced the formation of mineralized nodules and stimulated differentiation

a. The MTT assay shows a significant inhibition of cementoblast proliferation at days 3 and 5 after transient transfection with PEX3-Osterix plasmids; b. Representative photo images of Alizarin red assay reveals acceleration of mineralization after OSX transfection; c. Overexpression of OSX significantly increases expressions of BSP, OCN, and ALP, differentiation markers for cementoblast. Levels of different differentiation markers were normalized with those of the loading control GAPDH in three independent experiments. *, p <0.05; d. working hypothesis: OSX is initially expressed at low levels during cementogenesis in dental follicle cells, which increases as cellular cementum formation is being initiated. OSX controls cellular cementum formation by inhibiting proliferation of cells in the local region (PDL and follicle cells), while accelerating differentiation of follicle and PDL cells toward a cementoblast (Cb)/cementocyte (CMC) phenotype. Similar to dentinogenesis, but in contrast to osteogenesis, there appears to be a lack of remodeling in cementum. PDL, periodontal ligament; DF, dental follicle; DP, dental papilla; HERS, Hertwig’s epithelial root sheath cells.