Small molecules direct the generation of ameloblast-like cells from human embryonic stem cells

Abstract

Background: Ameloblasts present a promising avenue for the investigation of enamel and tooth regeneration. Previous protocols for directing the differentiation of human embryonic stem cells (hESCs) into dental epithelial (DE) cells involving the need for additional cells, conditional medium, and the use of costly cytokines. Importantly, ameloblasts have not been generated from hESCs in previous studies. Hence, we aimed to identify defined differentiation conditions that would solely utilize small molecules to achieve the production of ameloblasts.

Methods: We developed a three-step strategy entailing the progression of hESCs through non-neural ectoderm (NNE) and DE to generate functional ameloblasts in vitro. Initially, the NNE fate was induced from hESCs using a 6-day differentiation protocol with 1 µmol/L Retinoic acid (RA). Subsequently, the NNE lineage was differentiated into DE by employing a combination of 1 µmol/L LDN193189 (a BMP signaling inhibitor) and 1 µmol/L XAV939 (a WNT signaling inhibitor). In the final phase, 3 µmol/L CHIR99021 (a WNT signaling activator) and 2 µmol/L DAPT (a NOTCH signaling inhibitor) were utilized to achieve the fate of ameloblasts from DE cells. Three-dimensional cultures were investigated to enhance the ameloblast differentiation ability of the induced DE cells. Quantitative reverse transcription polymerase chain reaction (qRT-PCR) and immunofluorescence were conducted to assess the expression of lineage-specific markers. Alizarin Red S (ARS) staining was performed to evaluate the formation of mineralization nodules.

Results: The application of RA facilitated the efficient generation of NNE within a six-day period. Subsequently, upon stimulation with LDN193189 and XAV939, a notable emergence of DE cells was observed on the eighth days. By the tenth day, ameloblast-like cells derived from hESCs were generated. Upon cultivation in spheroids, these cells exhibited elevated levels of ameloblast markers AMBN and AMELX expression, suggesting that spheroid culture augments the differentiation of ameloblasts.

Conclusion: We established an efficient small molecule-based method to differentiate hESCs into ameloblast-like cells through the concerted modulation of RA, BMP, WNT, and NOTCH signaling pathways, potentially advancing research in enamel and tooth regeneration.

Keywords: Ameloblasts; Human embryonic stem cell; Small molecules; Spheroids.

Fig. 1

Selection of induction conditions for acquisition of NNE fate. (A) Schematic flow chart showing the three NNE differentiation protocols. (B) qRT-PCR analysis showing expression of NNE markers TFAP2A, DLX5 and DLX3, stemness, surface ectoderm, neural ectoderm, PPE, mesoderm and endoderm markers OCT4, K18, SOX1, SIX1, Brachyury, FOXA2 on day 4 of differentiation. (C) Brightfield and Immunostaining of TFAP2A (red), OCT4 (red) on day 4 of differentiation. Ca. Brightfield indicated the change of cell density. Scale bar, 10 μm, original magnification ×100. Cb-c. Immunostaining of TFAP2A (red), OCT4 (red). Scale bar, 100 μm, original magnification ×400. Human embryonic stem cells (hESCs). Non-neural ectoderm (NNE). SB431542 (SB). SU5402 (SU). Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate). *, p < 0.05, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001 and n.s, not significant

Fig. 2

Optimization of the cell density and timing of RA to enrich NNE differentiation. (A) Brightfield indicated the change of cell density. Scale bar, 50 μm. (B) qRT-PCR analysis showing expression of NNE markers DLX5, DLX3 and TFAP2A, NE, mesoderm and endoderm markers SOX1, BRACHYURY, FOXA2. (C) Immunostaining of TFAP2A (red), on day 4 of differentiation. Scale bar, 100 μm, original magnification ×400. (D) qRT-PCR analysis showing expression of NNE markers DLX5, DLX3, TFAP2A, and PPE makers SIX1 with the prolonged induction. (E) Immunostaining of TFAP2A (a) and SIX (b) with the prolonged induction. Scale bar, 100 μm, original magnification ×400. Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate). *, p < 0.05, ***, p < 0.001, ****, p < 0.0001 and n.s, not significant

Fig. 3

Combination of two small molecules synergistically promoted the PPE(DE) formation. A. Schematic representation of the PPE (DE) differentiation protocol. B. qRT-PCR analysis showing expression of PPE markers SIX1 and EYA1 by modulating BMP and WNT signaling. The statistic analysis was performed by comparing the DMSO control. C. Immunostaining of SIX1 (green) on day 8 of differentiation. Scale bar, 100 μm, original magnification ×400. D. qRT-PCR analysis showing expression of SIX1, EYA1, and DE markers PITX2 and PITX1 with the prolonged induction. E. Immunostaining of SIX1 (a) PITX2 (b) and PITX1 (c) with the prolonged induction. Scale bar, 100 μm, original magnification ×400; The percentage of PITX2⁺ and PITX1⁺ cells on day 8 of differentiation (d). F. ARS staining of hESCs-derived DE after induction of ameloblasts differentiation medium. Scale bar, 100 μm, original magnification ×200. G. ARS staining of hDPSCs after coculturing with hESCs-derived DE by transwell system. Scale bar, 100 μm, original magnification ×200. H. The percentage of SOX2⁺, ZO-1⁺ and Vimentin⁺ cells on day 8 of differentiation (a); Immunostaining of SOX2, ZO-1 and Vimentin on day 8 of differentiation (b). Scale bar, 100 μm, original magnification ×400; LDN193189 (LDN). XAV939 (XAV). Pre-placodal ectoderm (PPE). Dental epithelial (DE). Ameloblasts differentiation medium (OM). hESCs-derived DE (hESCs-DE). Human dental pulp stem cells (hDPSCs). Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate). *, p < 0.05, **, p < 0.01, ***, p< 0.001, ****, p < 0.0001

Fig. 4

Optimization conditions for ameloblast-lke cells generation. A. Schematic representation of the ameloblast differentiation protocol. B. Cell morphology changes in the differentiation process. Scale bar, 100 μm, original magnification ×100. C. qRT-PCR analysis showing expression of DE markers PITX2 and ameloblast markers AMBN, SP6 by modulating WNT, SHH and BMP signaling. The results were normalized to the DMSO control (a-f) and the CHIR control (g-i). D. Immunostaining of AMBN (red) on day 8 of differentiation. Scale bar, 100 μm, original magnification ×400. E. qRT-PCR analysis showing the expression of SP6, DLX2 and AMBN with time. F. Immunostaining of AMBN (red) with time. Scale bar, 100 μm, original magnification ×400. Mouse ameloblasts (mABs). CHIR99021 (CHIR). Purmorphamine (Purm). Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate). *, p < 0.05, **, p < 0.01, ***, p < 0.001, ****, p < 0.0001 and n.s, not significant

Fig. 5

Effect of 3D culture on ameloblast differentiation. A. Schematic representation of the cells in suspension culture. B. qRT-PCR analysis showing expression of ameloblast markers SP6, DLX2, AMBN, AMELX and ENAMELIN with time under 3D culture. C. Immunostaining of the ameloblast markers of AMBN (a) and AMELX (b) within the prolonged induction. Boxed region were examined under higher magnification. Scale bar, 50 μm, original magnification ×400. Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate)

Fig. 6

Comparison of ameloblast capacity between 2D and 3D culture. A. Schematic representation of the ameloblast differentiation protocol. B. qRT-PCR analysis showing expression of ameloblast markers SP6, DLX2, AMBN and AMELX under different culture conditions. C. Immunostaining of the AMELX. Scale bar, 100 μm, original magnification ×400. Data shown are mean ± SEM of three independent experiments (three technical replicates per biological replicate). **, p < 0.01, ***, p < 0.001