Human dental pulps were isolated from 8 donors of 6∼15 years old of children’s supernumerary teeth according to the protocol approved by Pusan National University Dental Hospital’s Institutional Review Board (IRB No. PNUDH-2020-003) for the experiments involving human tissue. All donors or their guardians gave their informed consent. The isolated supernumerary teeth were washed twice with phosphate-buffered saline (PBS) and split half with a hammer and forceps, followed by a gentle retrieval of the intact pulp tissue. The firm grip on the tooth when striking it with a hammer was important. Pulp tissues were cultured using a modified sandwich culture method without DPSCs isolation (17, 18). Briefly, freshly separated pulp tissue was placed on pre-coated 48-well plate (Cat No. 30048; SPL Life Sciences) using 75 μL 50% Matrigel (Cat No. 354230, GFR Matrigel; Corning Life Sciences) and sandwiched with 200 μL of 25% Matrigel to form the upper layer, in which pulp tissue was eventually submerged. Matrigel was diluted with StemMACS MSC Expansion Media Kit XF (Xeno-Free) (Cat No. 130-104-182; Miltenyi Biotec B.V. & Co. KG) containing 100 U/mL penicillin and 100 μg/mL streptomycin. StemMACS medium 800 μL was added on top of the solidified Matrigel and replaced every three day.

Three-dimensional explant cultures of human dental pulp tissue were harvested using cold PBS and fixed in 4% paraformaldehyde (Cat No. PC2205-100-74; Biosesang) at 4℃ for 24 hours. The fixed tissue was embedded into a paraffin block and sectioned into 5 μm thick slices using a microtome (Leica RM2255; Leica Biosystems). The section samples were deparaffinized and stained with H&E (Cat No. ab245880; Abcam) for histological observation. Staining images were obtained with an EVOS FL Auto2 Imaging System (Thermo Fisher Scientific).

Three-dimensional explant cultures of human dental pulp tissues were examined at the Korea Basic Science Institute’s Electron Microscopy Research Center for ultrastructural investigation according to the standard procedure. Samples were fixed immediately in 2% glutaralde-hyde and 2% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4℃ for 2 hours. Following three washes in phosphate buffer, the tissues were post-fixed with 1% osmium tetroxide on ice for 2 hours and washed three times, all in phosphate buffer. Samples were then embedded in pure epon 812 mixture after dehydration in ethanol and following infiltration in a mixture of propylene oxide and epon. Polymerization was conducted with pure resin at 70℃ for 24 hours. Sections were obtained with a model MT-X ultramicrotome (RMC) and collected on 100 mesh copper grids. After staining with 2% uranyl acetate and lead citrate, the sections were visualized by cryogenic TEM (JEM-1400 Plus, 120kV; JEOL).

Three-dimensional explant culture of human dental pulp tissues were washed with PBS and fixed with pre-warmed 4% paraformaldehyde at room temperature (RT) for 15 minutes. These samples were placed in an embedding medium (Tissue-Tek; Sakura Finetek USA, Inc.) and sections were prepared with a cryostat (thickness: 5 μm). Fixed samples were then incubated for 30 minutes at RT in 0.1 M PBS blocking buffer containing 1% bovine serum albumin and 0.1% Triton X-100. The samples were then washed and incubated with anti-CD45 antibody (100 μg/mL, Cat No. ab10558; Abcam), anti-CD31 antibody (0.5 μg/mL, Cat No. 102402; BioLegend), anti-Nestin antibody (200 μg/mL, Cat No. sc-23927; Santa Cruz Biotechnology), anti-PHEX antibody (100 μg/mL, Cat No. orb158144; Biorbyt), anti-Osterix antibody (0.2 mg/mL, Cat No. sc-393325; Santa Cruz Biotechnology), anti-DMP1 antibody (200 μg/mL, Cat No. sc-73633, Santa Cruz Biotechnology), or anti-LGR5 antibody (1 mg/mL, Cat No. ab75732; Abcam) overnight at 4℃. For immunofluorescence, samples were incubated with secondary antibody conjugated with Alexa Fluor 568 (2 mg/mL, Cat No. A-11004; Thermo Fisher Scie-ntific) or Alexa Fluor 488 (2 mg/mL, Cat No. A-11008, Thermo Fisher Scientific) at RT for 1 hour. Samples were washed with 0.1 M PBS blocking buffer and incubated in 0.1 μg/mL 4’,6-diamidino-2-phenylindole (DAPI) for 10 minutes. Images were taken with an EVOS FL Auto2 Imaging System. The number of immunostained cells for Nestin and LGR5 measured manually by an unbiased observer using ImageJ software (National Institutes of Health, https://imagej.net/ij/).

Three-dimensional culture of intact tissue can provide the advantage in understanding the biology of damage repair. Intact dental pulp tissues were isolated from supernumerary teeth of juvenile patients after hammer cracking and placed between 50% (lower) and 25% (upper) Matrigel with addition of xeno-free maintenance media on the top (Fig. 1A, 1B). Cellular outgrowth was observed from the outermost odontoblast area (Fig. 1C, Supplementary Fig. 1, Supplementary Video 1). Outgrowth layer was easily observable with brightfield microscope as pulp tissue was shown dark. Depending on the quality of pulp tissue, outgrowth was random or homogenous and continued to extend for a month but tended to wither after a month. Occasional accumulation of dentin was observed near the end of one month culture (Fig. 1D). These results suggest that 3D explant culture of human dental pulp tissue was established for the monitoring odontoblast outgrowth and dentin formation.

To understand cellular and molecular biology, 3D explant culture of human dental pulp tissue was subjected to TEM analysis. In the pulp cores, not many cells were detected except a few pulp cells and deposition of collagen matrix (Fig. 2A). In the pre-odontoblast area where reticular outgrowth of cells from odontoblast layer extended, frequent junctions of two or three cells with polarized location of nucleus were easily observed (Fig. 2B). Inside pre-odontoblasts, vast expansion of presumable lumen of endoplasmic reticulum or Golgi apparatus surrounded mitochondria with thin cytoplasm, which indicated the status of cells prior to process extension. In the area of mineral deposition, von Korff fibers, which can be found between odontoblasts and dentin and the first sign of dentin formation in the developing tooth, were detected along with secretory granules, which may contain calcium (Fig. 2C). In addition, the cross section of dentin revealed dentinal tubules and the boundary of predentin and deintin. These results indicate that outgrow cells are pre-odontoblasts, though the reticular growth could be in vitro artifact, and the orthodentin formation can be monitored in 3D explant culture of dental pulp tissue.

To characterize the outgrowth cells, 3D explant culture were subjected to immunohistochemistry analysis after 2-week culture. When the boundary of the preexisting pulp and the newly-formed outgrowth area were analyzed, cells in the dental pulp mostly migrated toward the boundary according to DAPI staining. Cells inside the pulp and in the outgrowth area were negative for hematopoietic marker CD45 or endothelial marker CD31 (Fig. 3). Cells near the boundary were positive for stem cell marker Nestin and odontoblast markers such as PHEX and DMP1 with the stronger expression of PHEX and DMP1 in the outgrowth area. Osterix expression, which is necessary for the differentiation of preosteoblasts to osteoblasts, was mostly observed at the leading edge of the outgrowth area. DPSCs cultures with the conventional 2D method did not show the expression of CD31, PHEX, LGR5, Ostreix or DMP-1, except Nestin (Supplementary Fig. 2). When Nestin was co-stained with LGR5, the regulator of Wnt signaling, double-positive cells were detected in the outgrowth area (Fig. 4A). Among DAPI-positive cells, Nestin-positive cells were 34.8%, LGR5-positive cells were 38.3%, and the double-positive cells were 18.7% (Fig. 4B). These results suggest that 3D explant culture of human dental pulp tissue provides the platform for the study of outgrowing odontoblasts.