Abstract

Materials and Methods

Transplantation.

Approximately 2.0 × 10⁶ SHED were mixed with 40 mg of hydroxyapatite/tricalcium phosphate (HA/TCP) ceramic powder (Zimmer, Warsaw, IN) and then transplanted s.c. into immunocompromised mice (NIH-bg-nu-xid, Harlan–Sprague–Dawley) as described (19).

SHED were injected into the brain of immunocompromised mice according to specifications of an approved small-animal protocol (National Institute of Dental and Craniofacial Research 01-185). Coordinates for the target sites were determined by referencing a murine brain atlas (ref. 20; see Fig. ​Fig.55A). The anteroposterior, mediolateral, and dorsoventral coordinates were computed relative to Bregma. Ex vivo-expanded SHED (10,000 cells per ml) were infused to the dentate gyrus of the hippocampus (21, 22). Cells (0.5 ml per side) were infused to the coordinates (anteroposterior, mediolateral, and dorsoventral, respectively, −1.5 mm, ±0.8 mm, and −2.0 mm) by using a 1-ml Hamilton Syringe.

Open in a separate window

Figure 5

Transplantation of SHED into the brain. (A) Diagram indicating injection of SHED into the dentate gyrus of the hippocampus. (B) SHED were cultured in the neural-differentiation medium as described in Materials and Methods for 1 week, after which 5,000 cells in 0.5 μl of PBS were injected into the dentate gyrus of the hippocampus of immunocompromised mice. After 10 days, the brain was fixed and prepared for immunofluorescence staining with NFM and human-specific anti-mitochondrial antibody. The anti-mitochondrial antibody immunostaining showed human SHED (arrows, green) in the dentate gyrus of the hippocampus with coexpression of NFM (arrows, red). In merged images, coexpression of human mitochondria and NFM showed colocalization of antigen expression as indicated by yellow fluorescence (arrows). (Magnification, ×20.)

Results

Here we demonstrate that the remaining crown of exfoliated deciduous teeth contains a living pulp remnant comprised of a normal dental pulp including connective tissue, blood vessels, and odontoblasts (Fig. ​(Fig.11 A–C). To isolate stem cells, single-cell suspensions were derived from the remnant pulp and placed at low density in liquid culture. Approximately 12–20 cells from each exfoliated incisor were capable of forming adherent colonies (Fig. ​(Fig.11D), characteristic of other stromal stem-cell populations (7). Interestingly, when cultured either under a neuronal-differentiation condition or in 3% rat serum with B27 supplement, these cells formed sphere-like clusters (Fig. ​(Fig.11E) in which highly proliferative cells aggregated together in clusters that either adhered to the culture dish or floated freely in the culture medium. After separating the sphere-like clusters, the cells were able to grow as individual fibroblastic cells (Fig. ​(Fig.11F). When compared with adult bone marrow stromal stem cells (BMSSCs) and dental pulp stem cells (DPSCs), SHED showed a higher proliferation rate (Fig. ​(Fig.11G) and a higher number of population doublings (Fig. ​(Fig.11H).

Open in a separate window

Figure 1

Isolation of SHED. (A) The exfoliated primary incisor contained dental pulp as shown (black triangles). The dashed line shows the occlusion edge of the incisor. (B and C) Hematoxylin/eosin staining indicated dentin (D) and pulp of exfoliated deciduous teeth. The pulp contained odontoblasts (arrows), blood vessels (open arrows), and connective tissues. The straight and curved dashed lines in B represent the occlusion and resorbed root surfaces, respectively. (D) Single colonies were formed after SHED were plated at low density and cultured for 2 weeks. (E) SHED were capable of forming sphere-like clusters when cultured with the conditions described in Materials and Methods. (F) The sphere-like clusters could be dissociated by passage through needles and subsequently grew on 0.1% gelatin-coated dishes. (G) The proliferation rates of SHED, BMSSCs, and DPSCs were assessed by BrdUrd (BrdU) incorporation for 12 h. SHED showed a significantly higher proliferation rate in comparison with BMSSCs and DPSCs (*, P < 0.05, Student's t test). (H) SHED were able to proliferate to >140 population doublings, which was significantly higher (*, P < 0.05, Student's t test) than BMSSCs and DPSCs.

Ex vivo-expanded SHED were found to express the cell-surface molecules STRO-1 and CD146 (MUC18), two early mesenchymal stem-cell markers previously found to be present in BMSSCs and DPSCs (Fig. ​(Fig.22 D and E). STRO-1- and CD146-positive cells were found to be located around blood vessels of the remnant pulp by immunohistochemical staining (Fig. ​(Fig.22 A and B), implying that SHED may have originated from a perivascular microenvironment. A minor proportion (9%) of ex vivo-expanded SHED stained positive for the STRO-1 antibody with fluorescence-activated cell-sorting analysis (Fig. ​(Fig.22C). Further immunohistotypic analysis demonstrated that cultured SHED expressed stromal- and vascular-related markers ALP, MEPE, bFGF, and endostatin (Fig. ​(Fig.22 F–I).

Open in a separate window

Figure 2

SHED possessed stem-cell characteristics. (A–E) The remnant pulp showed STRO-1 (open arrows in A) and CD146 (open arrows in B) immunopositive staining for cells in perivascular areas. Fluorescence-activated cell-sorting analysis showed that ex vivo-expanded SHED contained ≈9% STRO-1-positive cells (C). SHED expressed STRO-1 (D) and CD146 (E) (arrows). (F–I) SHED expressed the osteogenic and angiogenic markers ALP, MEPE, bFGF, and endostatin. (J and K) SHED were either cultured with regular medium (J) or l-ascorbate-2-phosphate, dexamethasone, and inorganic phosphate for 4 weeks (K). Alizarin red staining showed mineralized nodule formation in the induction (K). (L) Western blot analysis showed an up-regulated expression of CBFA1, ALP, MEPE, bone sialoprotein (BSP), and DSPP after the induction as described above. HSP90 was used to assess the amount of protein loaded per sample. (M) Human recombinant BMP-4 (300 ng/ml, 24 h) was added to induce a significant up-regulation of CBFA1, Osterix, and Osteocalcin (OC) in SHED as detected by semiquantitative PCR.

To investigate the potential of SHED to differentiate into mineralized tissue, established secondary SHED cultures were supplemented with l-ascorbate-2-phosphate, dexamethasone, and inorganic phosphate as described (7). Alizarin red-positive nodules formed in the SHED cultures after 4 weeks of induction (Fig. ​(Fig.22 J and K), indicating calcium accumulation in vitro. Accordingly, Western blot analysis revealed that various bone markers CBFA1, ALP, MEPE, and bone sialoprotein were up-regulated under the induction (Fig. ​(Fig.22L). In addition, DSPP was induced by the mineralizing induction (Fig. ​(Fig.22L). Furthermore, BMP-4 treatment was capable of inducing an up-regulated expression of CBFA1, Osterix, and Osteocalcin by semiquantitative RT-PCR (Fig. ​(Fig.22M). These data indicated that SHED possessed the ability to differentiate into functional odontoblast-like cells in vitro.

To validate the capacity of SHED to form odontoblasts, ex vivo-expanded SHED were transplanted into immunocompromised mice (7, 23). The transplants yielded human-specific alu-positive odontoblasts directly associated with a dentin-like structure (Fig. ​(Fig.33 A and B). Importantly, the regenerated dentin was immunoreactive to dentin-specific DSPP antibody (Fig. ​(Fig.33C). These findings indicated that human SHED satisfies one important stem-cell attribute: the ability to differentiate into odontoblasts in vivo. However, SHED were unable to regenerate a complete dentin–pulp-like complex as do DPSCs in vivo (Fig. ​(Fig.33 A and E). In addition, SHED were capable of inducing recipient murine cells to differentiate into bone-forming cells as noted by murine-specific pf1 in situ hybridization (Fig. ​(Fig.33L) and lacked DSPP expression (Fig. ​(Fig.33D). Importantly, skin fibroblasts were never capable of inducing bone formation after in vivo transplantation (data not shown). These data indicated that SHED are distinctively different from DPSCs in respect to the odontogenic differentiation and osteogenic induction.

Open in a separate window

Figure 3

Transplanted SHED into immunocompromised mice. (A and B) After 8 weeks of transplantation, SHED were able to differentiate into odontoblasts (open arrows) that were responsible for the dentin-like structure (D) formation on the surfaces of HA (A). The same field is shown for human-specific alu in situ hybridization, indicating the human origin of odontoblasts (open arrows, B). The black dashed line represents interface between newly formed dentin (D) and HA/TCP (HA). (C) Immunohistochemical staining of anti-DSPP antibody shows a positive staining on the regenerated dentin (black arrows). (D) In contrast to DPSC transplants, newly generated bone (B) by host cells in the same SHED transplant shows no reactivity to the DSPP antibody. (E) Of 12 selected single-colony-derived SHED strains, only 3 (25%) were capable of generating dentin in vivo. Newly formed dentin (arrows) was found to be adjacent to the surfaces of HA/TCP carrier (HA) and associated with connective tissue (CT). (F) Human-specific alu in situ hybridization showed that human cells (open arrows) were associated with dentin formation (D) and were residing within the connective-tissue compartment (CT). (G) The remaining 75% (9 of 12) single-colony-derived SHED strains were unable to generate dentin in vivo. (H) In situ hybridization demonstrated that alu-positive human cells survived in the connective-tissue compartment (CT) in the transplants in which there was no odontogenesis. Human cells were also found to surround the blood vessels (arrows). (I) Seven of 12 (58.4%) single-colony-derived SHED lines induced a very limited amount of bone formation (B) on the surface of HA/TCP (HA). (J) Of 12 single-colony-derived SHED lines, 5 (41.6%) were able to induce a significant amount of bone formation (B) on the surfaces of HA/TCP (HA). (K) The alu in situ hybridization showed human cells (arrows) attached to the surfaces of HA/TCP (HA) at the initial site of bone formation (B). The black dashed lines represent the interface between newly formed bone (B) and HA/TCP (HA). (L) In situ hybridization studies showed the murine-specific pf1 DNA probe reacting with osteoblasts and osteocytes (arrows) associated with the new bone formation (B).

We next examined the characteristics of clonal cell strains, each originating from a single cell of deciduous pulp. When 12 single-colony-derived SHED clones were transplanted into immunocompromised mice, only one fourth (3 of 12) of the clones demonstrated a potential to generate ectopic dentin-like tissue on the HA/TCP carrier equivalent to that generated by multicolony-derived SHED (Fig. ​(Fig.33 E and G). SHED from either single or multiple colonies were found to form dentin-like tissue (Fig. ​(Fig.33F) and to survive in the fibrous tissue within the transplants (Fig. ​(Fig.33H) as demonstrated by human-specific alu in situ hybridization. These results infer that SHED may contain subpopulations of cells, either differentiating into odontoblasts or residing in the connective-tissue compartments. Surprisingly, all transplanted single-colony-derived SHED clones were capable of inducing bone formation in immunocompromised mice. Approximately 40% of the clonal cell strains (5 of 12) induced a significant amount of new bone, whereas the remaining 60% (7 of 12) induced a limited amount of bone (Fig. ​(Fig.33 I and J). SHED were found to be located on the surfaces of HA/TCP but did not participate in bone formation as indicated by human-specific alu in situ hybridization (Fig. ​(Fig.33K). In contrast, murine host cells were found to differentiate into osteoblasts and osteocytes as shown by reactivity to murine-specific pf1 in situ hybridization (Fig. ​(Fig.33L).

Next we studied the potential of SHED to develop into other cell lineages such as neural cells and adipocytes. To elucidate the neural-differentiation potential of SHED, we examined the expression of neural markers in SHED. We found that cultured SHED expressed a variety of neural cell markers including nestin, βIII-tubulin, GAD, NeuN, GFAP, NFM, and CNPase as measured by immunocytochemical staining (Fig. ​(Fig.44 A–H) and Western blot analysis (Fig. ​(Fig.44I). After 4 weeks of neural inductive culture, expression levels of neuronal markers including βIII-tubulin, GAD, and NeuN were increased, whereas the levels of nestin, GFAP, NFM, and CNPase remained unchanged (Fig. ​(Fig.44I). When cultured under these conditions, SHED lost their fibroblastic morphology and developed multicytoplasmic processes correlating with either βIII-tubulin/GAD or βIII-tubulin/NFM expression (Fig. ​(Fig.44 J–O). The long cellular processes could be viewed best after toluidine blue staining and were immunoreactive to MAP2 and Tau antibodies (Fig. ​(Fig.44 P–S). After the neural inductive culture, SHED continued to express glial cell makers such as nestin, CNPase, GFAP, and NFM (Fig. ​(Fig.44 T–W).

Open in a separate window

Figure 4

Neural differentiation of SHED. (A–H) Immunocytochemical staining depicts SHED expressing nestin, GFAP, NFM, CNPase, βIII-tubulin, GAD, and NeuN. (I) Western blot analysis confirmed that SHED expressed neural markers as described above. After 4 weeks of culture in the presence of B27 supplement, bFGF (40 ng/ml), and epidermal growth factor (20 ng/ml) (Neural Diff. +), expression levels of βIII-tubulin, GAD, and NeuN were up-regulated when compared with regular culture conditions as described in Materials and Methods (Neural Diff. −). However, expression levels of nestin, GFAP, CNPase, and NFM remained the same after the treatment. (J–O) SHED may coexpress neuronal markers including βIII-tubulin (green)/GAD (red) and βIII-tubulin (green)/NFM (red). The morphology of SHED showed elongated cell-cytoplasmic processes that sometimes coexpress neural markers (triangles) or only express individual neural marker (open arrows). (P–S) Toluidine blue (0.1%) staining depicting the altered morphology of SHED after induction with neural culture medium (P and Q, arrows). Immunopositive staining for anti-MAP2 and anti-Tau antibodies on dendrites and axon (R and S, arrows), respectively. Double-staining experiments showing βIII-tubulin-positive cells were also detected in the same field (R, triangle, green). (T–W) SHED continued to express glial cell markers including nestin (red), CNPase (red), GFAP (red), and NFM (green) by immunocytostaining.

Neural developmental potential was studied further in vivo by injecting SHED into the dentate gyrus of the hippocampus of immunocompromised mice (Fig. ​(Fig.55A). Histological examination showed that SHED survived for >10 days inside the mouse brain microenvironment as noted by human-specific antimitochondrial antibody staining and continued to express neural markers such as NFM (Fig. ​(Fig.55B).

After 5 weeks of culture with an adipogenic inductive mixture, ≈5% of cultured SHED were found to possess the potential to develop into Oil red O-positive lipid-laden fat cells (Fig. ​(Fig.66A). This correlated with an up-regulation in the expression of two adipocyte-specific transcripts, peroxisome proliferator-activated receptor-γ2 and lipoprotein lipase, as detected by semiquantitative RT-PCR (Fig. ​(Fig.66B).

Open in a separate window

Figure 6

Adipogenic differentiation of SHED. (A) Cultured SHED formed Oil red O-positive lipid clusters after 5 weeks of induction in the presence of 0.5 mM isobutylmethylxanthine, 0.5 μM hydrocortisone, and 60 μM indomethacin. (B) A significant up-regulation of peroxisome proliferator-activated receptor-γ2 (PPARγ2) and lipoprotein lipase (LPL) was observed in the group induced with the adipogenic mixture (Adip) as compared with the control group (Cont) by RT-PCR.

Discussion

Here we provide evidence that remnant dental pulp derived from exfoliated deciduous teeth contains a multipotent stem-cell population. These stem cells can be isolated and expanded ex vivo, thereby providing a unique and accessible population of stem cells from an unexpected tissue resource. Previous experiments have shown that dental pulp tissue of adult teeth contains a population of DPSCs that are capable of differentiating into odontoblasts and adipocytes as well as expressing nestin and GFAP and form a dentin/pulp-like complex after in vivo transplantation (23). Deciduous teeth are significantly different from permanent teeth with regards to their developmental processes, tissue structure, and function. Therefore, it is not a surprise to find that SHED are distinct from DPSCs with respect to their higher proliferation rate, increased cell-population doublings, sphere-like cell-cluster formation, osteoinductive capacity in vivo, and failure to reconstitute a dentin–pulp-like complex. SHED apparently represent a population of multipotent stem cells that are perhaps more immature than previously examined postnatal stromal stem-cell populations.

The mechanisms controlling the growth and replacement of teeth are largely unknown (17), in particular with respect to how craniofacial components including bone and soft tissues surrounding teeth participate in the process of tooth development. SHED demonstrated a strong capacity to induce recipient cell-mediated bone formation in vivo. According to our investigations, SHED could not differentiate directly into osteoblasts but did induce new bone formation by forming an osteoinductive template to recruit murine host osteogenic cells. These data imply that deciduous teeth may not only provide guidance for the eruption of permanent teeth, as generally assumed, but may also be involved in inducing bone formation during the eruption of permanent teeth.

It is notable that SHED expressed neuronal and glial cell markers, which may be related to the neural crest-cell origin of the dental pulp (25). Neural crest cells play a pivotal role in embryonic development, giving rise to a variety of cell types such as neural cells, pigment cells, smooth muscle, craniofacial cartilage, and bone (26). Previous studies demonstrated that BMSSCs were also capable of differentiating into neural-like cells after in vivo transplantation (27). Dental pulp cells are known to produce neurotrophic factors and even rescue motoneurons after spinal cord injury (28). Moreover, neural progenitors were identified recently in mammalian dermal skin layers (6). These evidences support the notion that stem cells of nonneural tissue may be capable of differentiating into neural cells.

Our study provides evidence that SHED represent a population of postnatal stem cells capable of extensive proliferation and multipotential differentiation. Deciduous teeth therefore may be an ideal resource of stem cells to repair damaged tooth structures, induce bone regeneration, and possibly to treat neural tissue injury or degenerative diseases. However, the biological significance of the existence of SHED remains to be determined. This study provides a description of a stem-cell population residing in exfoliated human deciduous teeth and establishes the foundation for further studies to determine the efficacy of using SHED in cellular-based therapies.

Acknowledgments

Abbreviations

Footnotes

References