**4. Induction of differentiation of dental pulp stem cells into odontoblasts**

Recently, the use of stem cells in molecular and cell biology has led to new therapeutic strategies for regenerating damaged oral tissue. It is well known that the pulp is rich in adult mesenchymal stem cells (MSCs), and the stem cells isolated from the pulp have high proliferative potential and may be able to differentiate into hard tissueforming cells. Additionally, dental pulp stem cells play an important role in regenerative medicine for both oral and non-oral areas because of their high proliferation rate, pluripotency, and ease of collection [47]. Therefore, pulp stem cells are a promising source of MSCs used in a variety of clinical applications, such as bone formation, tooth tissue engineering, and nerve tissue regeneration [48]. We established a stable dental pulp cell line derived from GFP transgenic rats. It has the characteristics of dental pulp stem cells and exhibits stable odontoblast differentiation both *in vitro* and *in vivo*. To date, there are no reports of established cells showing stable osteoblastic

and stem cell-like properties over time, both *in vitro* and *in vivo*. However, this dental pulp cell line forms dentin-like hard tissue *in vivo* but does not lead to the induction of polar odontoblasts. A scaffold is an integral part of tissue engineering. Various artificial biomaterials have been developed as scaffolds and are widely applied clinically. In recent years, some studies have focused on the geometry of biomaterials. This is because scaffold composition and optimal geometry are believed to be important for inducing cell proliferation and differentiation. Focusing on this, we have already succeeded in developing a new biomaterial, honeycomb tricalcium phosphate (TCP), which contains holes of various diameters. Previous studies have reported that the surface properties of TCP due to different sintering temperatures affect hard tissue inducibility and biocompatibility [49]. Furthermore, cartilage and bone formation can be controlled by changing the diameter of the through holes in the honeycomb TCP. In a skull defect rat model, active bone tissue formation was observed in honeycomb TCP containing a through hole with a diameter of 300 μm, suggesting its clinical applicability [50]. These findings indicate that this honeycomb TCP can potentially act as a bioactive carrier and reproduce the interaction between progenitor cells and the extracellular matrix microenvironment. Additionally, we successfully differentiated polar odontoblasts from dental pulp stem cells using honeycomb TCP.

Gronthos et al. reported the isolation and characterization of pulp stem cells from wisdom tooth pulp tissue of impacted teeth, and reported that pulp stem cells have higher cell proliferation and tissue regeneration capacity than bone marrowderived mesenchymal stem cells [51]. Since then, many researchers have reported that dental pulp stem cells differentiate into a variety of cells, including nerve cells, adipocytes, chondrocytes, and bone [52, 53].

The TGC we created is a pulp-derived stem cell that can differentiate into functional odontoblast-like cells both *in vitro* and *in vivo*. *In vitro*, the bone-forming medium resulted in increased ALP activity of TGC and formation of calcium deposits (**Figure 7**).

Since transforming growth factor (TGF)-β is involved in dentin repair and dentin formation [54], we also investigated the effect of TGF-β on TGC.

*(a) TGC showed a fibroblast-like shape (left), expression of green fluorescent protein (GFP) (right). (b) Alizarin red staining of TGC exposed to osteogenic medium from 0 to 14 days. (c) Alkaline phosphatase (ALP) activity of TGC cultivated with osteogenic medium or TGF-*β*. the ALP activity from osteogenic medium became significantly higher than that from TGF-*β*. \* p < 0.05.*

Addition of TGF-β to TGC yielded results similar to those obtained with boneforming medium. However, the increase in ALP activity by TGF-β was weaker than that of the bone-forming medium. One hypothesis that explains this weak stimulus is that TGC is already exposed to TGF-β because it can express the TGF family of proteins.

Regeneration using dental pulp cells and scaffolds has been reported. Ceramics, such as polylactic acid, poly (α-hydroxyl) acids, polylactic-co-glycolic acid, and TCP or hydroxyapatite have been used as scaffolds for dentin regeneration [55]. These artificial biomaterials have already been confirmed to be highly biocompatible and are used as scaffolds for odontoblast differentiation and bone induction. To date, there have been several reports of dentin-like hard tissue formation in experiments combining various scaffolds with dental pulp stem cells [56]. Among these artificial biomaterials, TCP has been reported to be highly biocompatible, and when transplanted into a living body, it is absorbed over time and self-assembled. However, previous studies using TCP and pulp cells to induce odontoblast differentiation have not led to the regeneration of polar dentin [6].

Many studies have used artificial biomaterials that are suitable for inducing differentiation into odontoblasts. However, no studies have effectively induced the differentiation into odontoblasts by changing the geometric structure of artificial biomaterials. We have shown by histological observation that changing the pore diameter of honeycomb TCP with multiple through-holes changes the type and amount of hard tissue formed in the pores. *In vivo* TGC transplantation experiments showed bone-like hard tissue formation at a pore diameter of 75 μm TCP and 500 μm TCP. However, for 300 μm TCP, hard tissue formation was observed to be added to the TCP surface, and the induced cells were dentin sialoprotein (DSP)-positive odontoblasts (**Figure 8**). In addition, these cells had a polar sequence and exhibited an odontoblast-like structure, which was present in the pulp cavity. Since the pore diameter of 300 μm resembles the width of the pulp cavity [57], it is considered that the 300 μm honeycomb TCP reproduces the dental pulp environment in the living body and differentiates polar odontoblasts to form dentin.

#### **Figure 8.**

*The cells forming hard tissues in the TCP pores were GFP positive. Dentin Sialprotein (DSP) was not expressed in the cytoplasm of cells forming hard tissues in 75TCP and 500TCP. In contrast, in 300TCP, DSP was expressed in the cytoplasm of cells that were arranged with polarity on the TCP wall.*
